diff --git a/CHANGELOG.md b/CHANGELOG.md index f3dc99275..3aca132f8 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,9 +1,97 @@ -## v0.1.7 [unreleased] +## v0.1.10 [unreleased] + +### Release Notes +- The -test flag will now only output 2 collections for plugins that need it ### Features +- [#205](https://github.com/influxdb/telegraf/issues/205): Include per-db redis keyspace info +- [#226](https://github.com/influxdb/telegraf/pull/226): Add timestamps to points in Kafka/AMQP outputs. Thanks @ekini +- [#90](https://github.com/influxdb/telegraf/issues/90): Add Docker labels to tags in docker plugin +- [#223](https://github.com/influxdb/telegraf/pull/223): Add port tag to nginx plugin. Thanks @neezgee! +- [#227](https://github.com/influxdb/telegraf/pull/227): Add command intervals to exec plugin. Thanks @jpalay! -### Bigfixes +### Bugfixes +- [#228](https://github.com/influxdb/telegraf/pull/228): New version of package will replace old one. Thanks @ekini! + +## v0.1.9 [2015-09-22] + +### Release Notes +- InfluxDB output config change: `url` is now `urls`, and is a list. Config files +will still be backwards compatible if only `url` is specified. +- The -test flag will now output two metric collections +- Support for filtering telegraf outputs on the CLI -- Telegraf will now +allow filtering of output sinks on the command-line using the `-outputfilter` +flag, much like how the `-filter` flag works for plugins. +- Support for filtering on config-file creation -- Telegraf now supports +filtering to -sample-config command. You can now run +`telegraf -sample-config -filter cpu -outputfilter influxdb` to get a config +file with only the cpu plugin defined, and the influxdb output defined. +- **Breaking Change**: The CPU collection plugin has been refactored to fix some +bugs and outdated dependency issues. At the same time, I also decided to fix +a naming consistency issue, so cpu_percentageIdle will become cpu_usage_idle. +Also, all CPU time measurements now have it indicated in their name, so cpu_idle +will become cpu_time_idle. Additionally, cpu_time measurements are going to be +dropped in the default config. +- **Breaking Change**: The memory plugin has been refactored and some measurements +have been renamed for consistency. Some measurements have also been removed from being outputted. They are still being collected by gopsutil, and could easily be +re-added in a "verbose" mode if there is demand for it. + +### Features +- [#143](https://github.com/influxdb/telegraf/issues/143): InfluxDB clustering support +- [#181](https://github.com/influxdb/telegraf/issues/181): Makefile GOBIN support. Thanks @Vye! +- [#203](https://github.com/influxdb/telegraf/pull/200): AMQP output. Thanks @ekini! +- [#182](https://github.com/influxdb/telegraf/pull/182): OpenTSDB output. Thanks @rplessl! +- [#187](https://github.com/influxdb/telegraf/pull/187): Retry output sink connections on startup. +- [#220](https://github.com/influxdb/telegraf/pull/220): Add port tag to apache plugin. Thanks @neezgee! +- [#217](https://github.com/influxdb/telegraf/pull/217): Add filtering for output sinks +and filtering when specifying a config file. + +### Bugfixes +- [#170](https://github.com/influxdb/telegraf/issues/170): Systemd support +- [#175](https://github.com/influxdb/telegraf/issues/175): Set write precision before gathering metrics +- [#178](https://github.com/influxdb/telegraf/issues/178): redis plugin, multiple server thread hang bug +- Fix net plugin on darwin +- [#84](https://github.com/influxdb/telegraf/issues/84): Fix docker plugin on CentOS. Thanks @neezgee! +- [#189](https://github.com/influxdb/telegraf/pull/189): Fix mem_used_perc. Thanks @mced! +- [#192](https://github.com/influxdb/telegraf/issues/192): Increase compatibility of postgresql plugin. Now supports versions 8.1+ +- [#203](https://github.com/influxdb/telegraf/issues/203): EL5 rpm support. Thanks @ekini! +- [#206](https://github.com/influxdb/telegraf/issues/206): CPU steal/guest values wrong on linux. +- [#212](https://github.com/influxdb/telegraf/issues/212): Add hashbang to postinstall script. Thanks @ekini! +- [#212](https://github.com/influxdb/telegraf/issues/212): Fix makefile warning. Thanks @ekini! + +## v0.1.8 [2015-09-04] + +### Release Notes +- Telegraf will now write data in UTC at second precision by default +- Now using Go 1.5 to build telegraf + +### Features +- [#150](https://github.com/influxdb/telegraf/pull/150): Add Host Uptime metric to system plugin +- [#158](https://github.com/influxdb/telegraf/pull/158): Apache Plugin. Thanks @KPACHbIuLLIAnO4 +- [#159](https://github.com/influxdb/telegraf/pull/159): Use second precision for InfluxDB writes +- [#165](https://github.com/influxdb/telegraf/pull/165): Add additional metrics to mysql plugin. Thanks @nickscript0 +- [#162](https://github.com/influxdb/telegraf/pull/162): Write UTC by default, provide option +- [#166](https://github.com/influxdb/telegraf/pull/166): Upload binaries to S3 +- [#169](https://github.com/influxdb/telegraf/pull/169): Ping plugin + +### Bugfixes + +## v0.1.7 [2015-08-28] + +### Features +- [#38](https://github.com/influxdb/telegraf/pull/38): Kafka output producer. +- [#133](https://github.com/influxdb/telegraf/pull/133): Add plugin.Gather error logging. Thanks @nickscript0! +- [#136](https://github.com/influxdb/telegraf/issues/136): Add a -usage flag for printing usage of a single plugin. +- [#137](https://github.com/influxdb/telegraf/issues/137): Memcached: fix when a value contains a space +- [#138](https://github.com/influxdb/telegraf/issues/138): MySQL server address tag. +- [#142](https://github.com/influxdb/telegraf/pull/142): Add Description and SampleConfig funcs to output interface +- Indent the toml config file for readability + +### Bugfixes +- [#128](https://github.com/influxdb/telegraf/issues/128): system_load measurement missing. - [#129](https://github.com/influxdb/telegraf/issues/129): Latest pkg url fix. +- [#131](https://github.com/influxdb/telegraf/issues/131): Fix memory reporting on linux & darwin. Thanks @subhachandrachandra! +- [#140](https://github.com/influxdb/telegraf/issues/140): Memory plugin prec->perc typo fix. Thanks @brunoqc! ## v0.1.6 [2015-08-20] diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md new file mode 100644 index 000000000..e8c38456e --- /dev/null +++ b/CONTRIBUTING.md @@ -0,0 +1,163 @@ +## Sign the CLA + +Before we can merge a pull request, you will need to sign the CLA, +which can be found [on our website](http://influxdb.com/community/cla.html) + +## Plugins + +This section is for developers that want to create new collection plugins. +Telegraf is entirely plugin driven. This interface allows for operators to +pick and chose what is gathered as well as makes it easy for developers +to create new ways of generating metrics. + +Plugin authorship is kept as simple as possible to promote people to develop +and submit new plugins. + +### Plugin Guidelines + +* A plugin must conform to the `plugins.Plugin` interface. +* Telegraf promises to run each plugin's Gather function serially. This means +developers don't have to worry about thread safety within these functions. +* Each generated metric automatically has the name of the plugin that generated +it prepended. This is to keep plugins honest. +* Plugins should call `plugins.Add` in their `init` function to register themselves. +See below for a quick example. +* To be available within Telegraf itself, plugins must add themselves to the +`github.com/influxdb/telegraf/plugins/all/all.go` file. +* The `SampleConfig` function should return valid toml that describes how the +plugin can be configured. This is include in `telegraf -sample-config`. +* The `Description` function should say in one line what this plugin does. + +### Plugin interface + +```go +type Plugin interface { + SampleConfig() string + Description() string + Gather(Accumulator) error +} + +type Accumulator interface { + Add(measurement string, value interface{}, tags map[string]string) + AddFieldsWithTime(measurement string, + values map[string]interface{}, + tags map[string]string, + timestamp time.Time) +} +``` + +### Accumulator + +The way that a plugin emits metrics is by interacting with the Accumulator. + +The `Add` function takes 3 arguments: +* **measurement**: A string description of the metric. For instance `bytes_read` or `faults`. +* **value**: A value for the metric. This accepts 5 different types of value: + * **int**: The most common type. All int types are accepted but favor using `int64` + Useful for counters, etc. + * **float**: Favor `float64`, useful for gauges, percentages, etc. + * **bool**: `true` or `false`, useful to indicate the presence of a state. `light_on`, etc. + * **string**: Typically used to indicate a message, or some kind of freeform information. + * **time.Time**: Useful for indicating when a state last occurred, for instance `light_on_since`. +* **tags**: This is a map of strings to strings to describe the where or who +about the metric. For instance, the `net` plugin adds a tag named `"interface"` +set to the name of the network interface, like `"eth0"`. + +The `AddFieldsWithTime` allows multiple values for a point to be passed. The values +used are the same type profile as **value** above. The **timestamp** argument +allows a point to be registered as having occurred at an arbitrary time. + +Let's say you've written a plugin that emits metrics about processes on the current host. + +```go + +type Process struct { + CPUTime float64 + MemoryBytes int64 + PID int +} + +func Gather(acc plugins.Accumulator) error { + for _, process := range system.Processes() { + tags := map[string]string { + "pid": fmt.Sprintf("%d", process.Pid), + } + + acc.Add("cpu", process.CPUTime, tags) + acc.Add("memory", process.MemoryBytes, tags) + } +} +``` + +### Example + +```go +package simple + +// simple.go + +import "github.com/influxdb/telegraf/plugins" + +type Simple struct { + Ok bool +} + +func (s *Simple) Description() string { + return "a demo plugin" +} + +func (s *Simple) SampleConfig() string { + return "ok = true # indicate if everything is fine" +} + +func (s *Simple) Gather(acc plugins.Accumulator) error { + if s.Ok { + acc.Add("state", "pretty good", nil) + } else { + acc.Add("state", "not great", nil) + } + + return nil +} + +func init() { + plugins.Add("simple", func() plugins.Plugin { return &Simple{} }) +} +``` + +## Outputs + +TODO: this section will describe requirements for contributing an output + +## Unit Tests + +### Execute short tests + +execute `make test-short` + +### Execute long tests + +As Telegraf collects metrics from several third-party services it becomes a +difficult task to mock each service as some of them have complicated protocols +which would take some time to replicate. + +To overcome this situation we've decided to use docker containers to provide a +fast and reproducible environment to test those services which require it. +For other situations +(i.e: https://github.com/influxdb/telegraf/blob/master/plugins/redis/redis_test.go ) +a simple mock will suffice. + +To execute Telegraf tests follow these simple steps: + +- Install docker compose following [these](https://docs.docker.com/compose/install/) +instructions +- execute `make test` + +**OSX users**: you will need to install `boot2docker` or `docker-machine`. +The Makefile will assume that you have a `docker-machine` box called `default` to +get the IP address. + +### Unit test troubleshooting + +Try cleaning up your test environment by executing `make test-cleanup` and +re-running diff --git a/Godeps/Godeps.json b/Godeps/Godeps.json index 9400bc682..447d70320 100644 --- a/Godeps/Godeps.json +++ b/Godeps/Godeps.json @@ -1,6 +1,6 @@ { "ImportPath": "github.com/influxdb/telegraf", - "GoVersion": "go1.4.2", + "GoVersion": "go1.5", "Packages": [ "./..." ], @@ -28,6 +28,11 @@ "ImportPath": "github.com/cenkalti/backoff", "Rev": "4dc77674aceaabba2c7e3da25d4c823edfb73f99" }, + { + "ImportPath": "github.com/cloudfoundry/gosigar", + "Comment": "scotty_09012012-27-g3ed7c74", + "Rev": "3ed7c74352dae6dc00bdc8c74045375352e3ec05" + }, { "ImportPath": "github.com/dancannon/gorethink/encoding", "Comment": "v1.x.x-1-g786f12a", @@ -55,7 +60,7 @@ }, { "ImportPath": "github.com/fsouza/go-dockerclient", - "Rev": "42d06e2b125654477366c320dcea99107a86e9c2" + "Rev": "af9789bbd78acf3e279274caa54682185eb7ed33" }, { "ImportPath": "github.com/go-sql-driver/mysql", @@ -91,34 +96,9 @@ "Rev": "d1e82c1ec3f15ee991f7cc7ffd5b67ff6f5bbaee" }, { - "ImportPath": "github.com/influxdb/influxdb/client", - "Comment": "v0.9.3-rc1", - "Rev": "f4077764b2bb2b03241452d88e9db321c62bb560" - }, - { - "ImportPath": "github.com/influxdb/influxdb/influxql", - "Comment": "v0.9.3-rc1", - "Rev": "f4077764b2bb2b03241452d88e9db321c62bb560" - }, - { - "ImportPath": "github.com/influxdb/influxdb/meta", - "Comment": "v0.9.3-rc1", - "Rev": "f4077764b2bb2b03241452d88e9db321c62bb560" - }, - { - "ImportPath": "github.com/influxdb/influxdb/snapshot", - "Comment": "v0.9.3-rc1", - "Rev": "f4077764b2bb2b03241452d88e9db321c62bb560" - }, - { - "ImportPath": "github.com/influxdb/influxdb/toml", - "Comment": "v0.9.3-rc1", - "Rev": "f4077764b2bb2b03241452d88e9db321c62bb560" - }, - { - "ImportPath": "github.com/influxdb/influxdb/tsdb", - "Comment": "v0.9.3-rc1", - "Rev": "f4077764b2bb2b03241452d88e9db321c62bb560" + "ImportPath": "github.com/influxdb/influxdb", + "Comment": "v0.9.3", + "Rev": "5d42b212fca8facfe9db0c83822f09b88be643ec" }, { "ImportPath": "github.com/lib/pq", @@ -161,6 +141,45 @@ "ImportPath": "github.com/samuel/go-zookeeper/zk", "Rev": "5bb5cfc093ad18a28148c578f8632cfdb4d802e4" }, + { + "ImportPath": "github.com/shirou/gopsutil/common", + "Comment": "1.0.0-158-g0fd612e", + "Rev": "0fd612ec7b9079dc624ae4815acadf1903d82011" + }, + { + "ImportPath": "github.com/shirou/gopsutil/cpu", + "Comment": "1.0.0-158-g0fd612e", + "Rev": "0fd612ec7b9079dc624ae4815acadf1903d82011" + }, + { + "ImportPath": "github.com/shirou/gopsutil/disk", + "Comment": "1.0.0-158-g0fd612e", + "Rev": "0fd612ec7b9079dc624ae4815acadf1903d82011" + }, + { + "ImportPath": "github.com/shirou/gopsutil/docker", + "Comment": "1.0.0-158-g0fd612e", + "Rev": "0fd612ec7b9079dc624ae4815acadf1903d82011" + }, + { + "ImportPath": "github.com/shirou/gopsutil/load", + "Comment": "1.0.0-158-g0fd612e", + "Rev": "0fd612ec7b9079dc624ae4815acadf1903d82011" + }, + { + "ImportPath": "github.com/shirou/gopsutil/mem", + "Comment": "1.0.0-158-g0fd612e", + "Rev": "0fd612ec7b9079dc624ae4815acadf1903d82011" + }, + { + "ImportPath": "github.com/shirou/gopsutil/net", + "Comment": "1.0.0-158-g0fd612e", + "Rev": "0fd612ec7b9079dc624ae4815acadf1903d82011" + }, + { + "ImportPath": "github.com/streadway/amqp", + "Rev": "f4879ba28fffbb576743b03622a9ff20461826b2" + }, { "ImportPath": "github.com/stretchr/objx", "Rev": "cbeaeb16a013161a98496fad62933b1d21786672" diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/.gitignore b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/.gitignore new file mode 100644 index 000000000..8000dd9db --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/.gitignore @@ -0,0 +1 @@ +.vagrant diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/.travis.yml b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/.travis.yml new file mode 100644 index 000000000..2a9c5d0c7 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/.travis.yml @@ -0,0 +1,8 @@ +language: go + +go: + - 1.2 + +install: + - 'go install github.com/onsi/ginkgo/ginkgo' +script: 'ginkgo -r' diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/LICENSE b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/LICENSE new file mode 100644 index 000000000..11069edd7 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/LICENSE @@ -0,0 +1,201 @@ + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + +TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + +1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + +2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + +3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + +4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + +5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + +6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + +7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + +8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + +9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + +END OF TERMS AND CONDITIONS + +APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + +Copyright [yyyy] [name of copyright owner] + +Licensed under the Apache License, Version 2.0 (the "License"); +you may not use this file except in compliance with the License. +You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + +Unless required by applicable law or agreed to in writing, software +distributed under the License is distributed on an "AS IS" BASIS, +WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +See the License for the specific language governing permissions and +limitations under the License. diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/NOTICE b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/NOTICE new file mode 100644 index 000000000..fda553b5c --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/NOTICE @@ -0,0 +1,9 @@ +Copyright (c) [2009-2011] VMware, Inc. All Rights Reserved. + +This product is licensed to you under the Apache License, Version 2.0 (the "License"). +You may not use this product except in compliance with the License. + +This product includes a number of subcomponents with +separate copyright notices and license terms. Your use of these +subcomponents is subject to the terms and conditions of the +subcomponent's license, as noted in the LICENSE file. \ No newline at end of file diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/README.md b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/README.md new file mode 100644 index 000000000..90d51f9b1 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/README.md @@ -0,0 +1,22 @@ +# Go sigar + +## Overview + +Go sigar is a golang implementation of the +[sigar API](https://github.com/hyperic/sigar). The Go version of +sigar has a very similar interface, but is being written from scratch +in pure go/cgo, rather than cgo bindings for libsigar. + +## Test drive + + $ go get github.com/cloudfoundry/gosigar + $ cd $GOPATH/src/github.com/cloudfoundry/gosigar/examples + $ go run uptime.go + +## Supported platforms + +Currently targeting modern flavors of darwin and linux. + +## License + +Apache 2.0 diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/Vagrantfile b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/Vagrantfile new file mode 100644 index 000000000..6fd990c14 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/Vagrantfile @@ -0,0 +1,25 @@ +# Vagrantfile API/syntax version. Don't touch unless you know what you're doing! +VAGRANTFILE_API_VERSION = "2" + +Vagrant.configure(VAGRANTFILE_API_VERSION) do |config| + config.vm.box = "hashicorp/precise64" + config.vm.provision "shell", inline: "mkdir -p /home/vagrant/go" + config.vm.synced_folder ".", "/home/vagrant/go/src/github.com/cloudfoundry/gosigar" + config.vm.provision "shell", inline: "chown -R vagrant:vagrant /home/vagrant/go" + install_go = <<-BASH + set -e + +if [ ! -d "/usr/local/go" ]; then + cd /tmp && wget https://storage.googleapis.com/golang/go1.3.3.linux-amd64.tar.gz + cd /usr/local + tar xvzf /tmp/go1.3.3.linux-amd64.tar.gz + echo 'export GOPATH=/home/vagrant/go; export PATH=/usr/local/go/bin:$PATH:$GOPATH/bin' >> /home/vagrant/.bashrc +fi +export GOPATH=/home/vagrant/go +export PATH=/usr/local/go/bin:$PATH:$GOPATH/bin +/usr/local/go/bin/go get -u github.com/onsi/ginkgo/ginkgo +/usr/local/go/bin/go get -u github.com/onsi/gomega; +BASH + config.vm.provision "shell", inline: 'apt-get install -y git-core' + config.vm.provision "shell", inline: install_go +end diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/concrete_sigar.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/concrete_sigar.go new file mode 100644 index 000000000..0e80aa4b9 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/concrete_sigar.go @@ -0,0 +1,69 @@ +package sigar + +import ( + "time" +) + +type ConcreteSigar struct{} + +func (c *ConcreteSigar) CollectCpuStats(collectionInterval time.Duration) (<-chan Cpu, chan<- struct{}) { + // samplesCh is buffered to 1 value to immediately return first CPU sample + samplesCh := make(chan Cpu, 1) + + stopCh := make(chan struct{}) + + go func() { + var cpuUsage Cpu + + // Immediately provide non-delta value. + // samplesCh is buffered to 1 value, so it will not block. + cpuUsage.Get() + samplesCh <- cpuUsage + + ticker := time.NewTicker(collectionInterval) + + for { + select { + case <-ticker.C: + previousCpuUsage := cpuUsage + + cpuUsage.Get() + + select { + case samplesCh <- cpuUsage.Delta(previousCpuUsage): + default: + // Include default to avoid channel blocking + } + + case <-stopCh: + return + } + } + }() + + return samplesCh, stopCh +} + +func (c *ConcreteSigar) GetLoadAverage() (LoadAverage, error) { + l := LoadAverage{} + err := l.Get() + return l, err +} + +func (c *ConcreteSigar) GetMem() (Mem, error) { + m := Mem{} + err := m.Get() + return m, err +} + +func (c *ConcreteSigar) GetSwap() (Swap, error) { + s := Swap{} + err := s.Get() + return s, err +} + +func (c *ConcreteSigar) GetFileSystemUsage(path string) (FileSystemUsage, error) { + f := FileSystemUsage{} + err := f.Get(path) + return f, err +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/concrete_sigar_test.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/concrete_sigar_test.go new file mode 100644 index 000000000..ec51811c4 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/concrete_sigar_test.go @@ -0,0 +1,85 @@ +package sigar_test + +import ( + "time" + + . "github.com/onsi/ginkgo" + . "github.com/onsi/gomega" + + sigar "github.com/cloudfoundry/gosigar" +) + +var _ = Describe("ConcreteSigar", func() { + var concreteSigar *sigar.ConcreteSigar + + BeforeEach(func() { + concreteSigar = &sigar.ConcreteSigar{} + }) + + Describe("CollectCpuStats", func() { + It("immediately makes first CPU usage available even though it's not very accurate", func() { + samplesCh, stop := concreteSigar.CollectCpuStats(500 * time.Millisecond) + + firstValue := <-samplesCh + Expect(firstValue.User).To(BeNumerically(">", 0)) + + stop <- struct{}{} + }) + + It("makes CPU usage delta values available", func() { + samplesCh, stop := concreteSigar.CollectCpuStats(500 * time.Millisecond) + + firstValue := <-samplesCh + + secondValue := <-samplesCh + Expect(secondValue.User).To(BeNumerically("<", firstValue.User)) + + stop <- struct{}{} + }) + + It("does not block", func() { + _, stop := concreteSigar.CollectCpuStats(10 * time.Millisecond) + + // Sleep long enough for samplesCh to fill at least 2 values + time.Sleep(20 * time.Millisecond) + + stop <- struct{}{} + + // If CollectCpuStats blocks it will never get here + Expect(true).To(BeTrue()) + }) + }) + + It("GetLoadAverage", func() { + avg, err := concreteSigar.GetLoadAverage() + Expect(avg.One).ToNot(BeNil()) + Expect(avg.Five).ToNot(BeNil()) + Expect(avg.Fifteen).ToNot(BeNil()) + + Expect(err).ToNot(HaveOccurred()) + }) + + It("GetMem", func() { + mem, err := concreteSigar.GetMem() + Expect(err).ToNot(HaveOccurred()) + + Expect(mem.Total).To(BeNumerically(">", 0)) + Expect(mem.Used + mem.Free).To(BeNumerically("<=", mem.Total)) + }) + + It("GetSwap", func() { + swap, err := concreteSigar.GetSwap() + Expect(err).ToNot(HaveOccurred()) + Expect(swap.Used + swap.Free).To(BeNumerically("<=", swap.Total)) + }) + + It("GetSwap", func() { + fsusage, err := concreteSigar.GetFileSystemUsage("/") + Expect(err).ToNot(HaveOccurred()) + Expect(fsusage.Total).ToNot(BeNil()) + + fsusage, err = concreteSigar.GetFileSystemUsage("T O T A L L Y B O G U S") + Expect(err).To(HaveOccurred()) + Expect(fsusage.Total).To(Equal(uint64(0))) + }) +}) diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/cputimes.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/cputimes.go new file mode 100644 index 000000000..cdfcd2f11 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/cputimes.go @@ -0,0 +1,52 @@ +package main + +import ( + "fmt" + "time" + + "github.com/cloudfoundry/gosigar" +) + +func main() { + cpus := sigar.CpuList{} + cpus.Get() + tcpu := getOverallCpu(cpus) + + for i, cpu := range cpus.List { + fmt.Printf("CPU%d Ticks: %d\n", i, cpu.Total()) + } + + fmt.Printf("Total CPU Ticks: %d\n", tcpu.Total()) + fmt.Printf("Total CPU Time: %d\n", tcpu.Total()/128) + fmt.Printf("User CPU Time: %d\n", tcpu.User/128) + + time.Sleep(1 * time.Second) + tcpu2 := sigar.Cpu{} + tcpu2.Get() + + dcpu := tcpu2.Delta(tcpu) + tcpuDelta := tcpu2.Total() - tcpu.Total() + iPercentage := 100.0 * float64(dcpu.Idle) / float64(tcpuDelta) + fmt.Printf("Idle percentage: %f\n", iPercentage) + bPercentage := 100.0 * float64(busy(tcpu2)-busy(tcpu)) / float64(tcpuDelta) + fmt.Printf("Busy percentage: %f\n", bPercentage) +} + +func busy(c sigar.Cpu) uint64 { + return c.Total() - c.Idle +} + +func getOverallCpu(cl sigar.CpuList) sigar.Cpu { + var overallCpu sigar.Cpu + for _, c := range cl.List { + overallCpu.User += c.User + overallCpu.Nice += c.Nice + overallCpu.Sys += c.Sys + overallCpu.Idle += c.Idle + overallCpu.Wait += c.Wait + overallCpu.Irq += c.Irq + overallCpu.SoftIrq += c.SoftIrq + overallCpu.Stolen += c.Stolen + } + return overallCpu +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/df.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/df.go new file mode 100644 index 000000000..96c92f41d --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/df.go @@ -0,0 +1,39 @@ +// Copyright (c) 2012 VMware, Inc. + +package main + +import ( + "fmt" + "github.com/cloudfoundry/gosigar" + "os" +) + +const output_format = "%-15s %4s %4s %5s %4s %-15s\n" + +func formatSize(size uint64) string { + return sigar.FormatSize(size * 1024) +} + +func main() { + fslist := sigar.FileSystemList{} + fslist.Get() + + fmt.Fprintf(os.Stdout, output_format, + "Filesystem", "Size", "Used", "Avail", "Use%", "Mounted on") + + for _, fs := range fslist.List { + dir_name := fs.DirName + + usage := sigar.FileSystemUsage{} + + usage.Get(dir_name) + + fmt.Fprintf(os.Stdout, output_format, + fs.DevName, + formatSize(usage.Total), + formatSize(usage.Used), + formatSize(usage.Avail), + sigar.FormatPercent(usage.UsePercent()), + dir_name) + } +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/free.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/free.go new file mode 100644 index 000000000..9bf9d3db3 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/free.go @@ -0,0 +1,33 @@ +// Copyright (c) 2012 VMware, Inc. + +package main + +import ( + "fmt" + "github.com/cloudfoundry/gosigar" + "os" +) + +func format(val uint64) uint64 { + return val / 1024 +} + +func main() { + mem := sigar.Mem{} + swap := sigar.Swap{} + + mem.Get() + swap.Get() + + fmt.Fprintf(os.Stdout, "%18s %10s %10s\n", + "total", "used", "free") + + fmt.Fprintf(os.Stdout, "Mem: %10d %10d %10d\n", + format(mem.Total), format(mem.Used), format(mem.Free)) + + fmt.Fprintf(os.Stdout, "-/+ buffers/cache: %10d %10d\n", + format(mem.ActualUsed), format(mem.ActualFree)) + + fmt.Fprintf(os.Stdout, "Swap: %10d %10d %10d\n", + format(swap.Total), format(swap.Used), format(swap.Free)) +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/ps.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/ps.go new file mode 100644 index 000000000..e3cc2281f --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/ps.go @@ -0,0 +1,37 @@ +// Copyright (c) 2012 VMware, Inc. + +package main + +import ( + "fmt" + "github.com/cloudfoundry/gosigar" +) + +func main() { + pids := sigar.ProcList{} + pids.Get() + + // ps -eo pid,ppid,stime,time,rss,state,comm + fmt.Print(" PID PPID STIME TIME RSS S COMMAND\n") + + for _, pid := range pids.List { + state := sigar.ProcState{} + mem := sigar.ProcMem{} + time := sigar.ProcTime{} + + if err := state.Get(pid); err != nil { + continue + } + if err := mem.Get(pid); err != nil { + continue + } + if err := time.Get(pid); err != nil { + continue + } + + fmt.Printf("%5d %5d %s %s %6d %c %s\n", + pid, state.Ppid, + time.FormatStartTime(), time.FormatTotal(), + mem.Resident/1024, state.State, state.Name) + } +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/uptime.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/uptime.go new file mode 100644 index 000000000..337a9b01a --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/examples/uptime.go @@ -0,0 +1,27 @@ +// Copyright (c) 2012 VMware, Inc. + +package main + +import ( + "fmt" + "github.com/cloudfoundry/gosigar" + "os" + "time" +) + +func main() { + concreteSigar := sigar.ConcreteSigar{} + + uptime := sigar.Uptime{} + uptime.Get() + avg, err := concreteSigar.GetLoadAverage() + if err != nil { + fmt.Printf("Failed to get load average") + return + } + + fmt.Fprintf(os.Stdout, " %s up %s load average: %.2f, %.2f, %.2f\n", + time.Now().Format("15:04:05"), + uptime.Format(), + avg.One, avg.Five, avg.Fifteen) +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/fakes/fake_sigar.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/fakes/fake_sigar.go new file mode 100644 index 000000000..6fb77417c --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/fakes/fake_sigar.go @@ -0,0 +1,72 @@ +package fakes + +import ( + "time" + + sigar "github.com/cloudfoundry/gosigar" +) + +type FakeSigar struct { + LoadAverage sigar.LoadAverage + LoadAverageErr error + + Mem sigar.Mem + MemErr error + + Swap sigar.Swap + SwapErr error + + FileSystemUsage sigar.FileSystemUsage + FileSystemUsageErr error + FileSystemUsagePath string + + CollectCpuStatsCpuCh chan sigar.Cpu + CollectCpuStatsStopCh chan struct{} +} + +func NewFakeSigar() *FakeSigar { + return &FakeSigar{ + CollectCpuStatsCpuCh: make(chan sigar.Cpu, 1), + CollectCpuStatsStopCh: make(chan struct{}), + } +} + +func (f *FakeSigar) CollectCpuStats(collectionInterval time.Duration) (<-chan sigar.Cpu, chan<- struct{}) { + samplesCh := make(chan sigar.Cpu, 1) + stopCh := make(chan struct{}) + + go func() { + for { + select { + case cpuStat := <-f.CollectCpuStatsCpuCh: + select { + case samplesCh <- cpuStat: + default: + // Include default to avoid channel blocking + } + + case <-f.CollectCpuStatsStopCh: + return + } + } + }() + + return samplesCh, stopCh +} + +func (f *FakeSigar) GetLoadAverage() (sigar.LoadAverage, error) { + return f.LoadAverage, f.LoadAverageErr +} + +func (f *FakeSigar) GetMem() (sigar.Mem, error) { + return f.Mem, f.MemErr +} + +func (f *FakeSigar) GetSwap() (sigar.Swap, error) { + return f.Swap, f.SwapErr +} + +func (f *FakeSigar) GetFileSystemUsage(path string) (sigar.FileSystemUsage, error) { + f.FileSystemUsagePath = path + return f.FileSystemUsage, f.FileSystemUsageErr +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/README.md b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/README.md new file mode 100644 index 000000000..dd34ebcfb --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/README.md @@ -0,0 +1,50 @@ +# Process notifications for Go + +## Overview + +The psnotify package captures process events from the kernel via +kqueue on Darwin/BSD and the netlink connector on Linux. + +The psnotify API is similar to the +[fsnotify](https://github.com/howeyc/fsnotify) package. + +Example: +```go + watcher, err := psnotify.NewWatcher() + if err != nil { + log.Fatal(err) + } + + // Process events + go func() { + for { + select { + case ev := <-watcher.Fork: + log.Println("fork event:", ev) + case ev := <-watcher.Exec: + log.Println("exec event:", ev) + case ev := <-watcher.Exit: + log.Println("exit event:", ev) + case err := <-watcher.Error: + log.Println("error:", err) + } + } + }() + + err = watcher.Watch(os.Getpid(), psnotify.PROC_EVENT_ALL) + if err != nil { + log.Fatal(err) + } + + /* ... do stuff ... */ + watcher.Close() +``` + +## Supported platforms + +Currently targeting modern flavors of Darwin and Linux. +Should work on BSD, but untested. + +## License + +Apache 2.0 diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify.go new file mode 100644 index 000000000..6a69f4de2 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify.go @@ -0,0 +1,136 @@ +// Copyright (c) 2012 VMware, Inc. + +package psnotify + +import ( + "errors" + "fmt" +) + +type ProcEventFork struct { + ParentPid int // Pid of the process that called fork() + ChildPid int // Child process pid created by fork() +} + +type ProcEventExec struct { + Pid int // Pid of the process that called exec() +} + +type ProcEventExit struct { + Pid int // Pid of the process that called exit() +} + +type watch struct { + flags uint32 // Saved value of Watch() flags param +} + +type eventListener interface { + close() error // Watch.Close() closes the OS specific listener +} + +type Watcher struct { + listener eventListener // OS specifics (kqueue or netlink) + watches map[int]*watch // Map of watched process ids + Error chan error // Errors are sent on this channel + Fork chan *ProcEventFork // Fork events are sent on this channel + Exec chan *ProcEventExec // Exec events are sent on this channel + Exit chan *ProcEventExit // Exit events are sent on this channel + done chan bool // Used to stop the readEvents() goroutine + isClosed bool // Set to true when Close() is first called +} + +// Initialize event listener and channels +func NewWatcher() (*Watcher, error) { + listener, err := createListener() + + if err != nil { + return nil, err + } + + w := &Watcher{ + listener: listener, + watches: make(map[int]*watch), + Fork: make(chan *ProcEventFork), + Exec: make(chan *ProcEventExec), + Exit: make(chan *ProcEventExit), + Error: make(chan error), + done: make(chan bool, 1), + } + + go w.readEvents() + return w, nil +} + +// Close event channels when done message is received +func (w *Watcher) finish() { + close(w.Fork) + close(w.Exec) + close(w.Exit) + close(w.Error) +} + +// Closes the OS specific event listener, +// removes all watches and closes all event channels. +func (w *Watcher) Close() error { + if w.isClosed { + return nil + } + w.isClosed = true + + for pid := range w.watches { + w.RemoveWatch(pid) + } + + w.done <- true + + w.listener.close() + + return nil +} + +// Add pid to the watched process set. +// The flags param is a bitmask of process events to capture, +// must be one or more of: PROC_EVENT_FORK, PROC_EVENT_EXEC, PROC_EVENT_EXIT +func (w *Watcher) Watch(pid int, flags uint32) error { + if w.isClosed { + return errors.New("psnotify watcher is closed") + } + + watchEntry, found := w.watches[pid] + + if found { + watchEntry.flags |= flags + } else { + if err := w.register(pid, flags); err != nil { + return err + } + w.watches[pid] = &watch{flags: flags} + } + + return nil +} + +// Remove pid from the watched process set. +func (w *Watcher) RemoveWatch(pid int) error { + _, ok := w.watches[pid] + if !ok { + msg := fmt.Sprintf("watch for pid=%d does not exist", pid) + return errors.New(msg) + } + delete(w.watches, pid) + return w.unregister(pid) +} + +// Internal helper to check if there is a message on the "done" channel. +// The "done" message is sent by the Close() method; when received here, +// the Watcher.finish method is called to close all channels and return +// true - in which case the caller should break from the readEvents loop. +func (w *Watcher) isDone() bool { + var done bool + select { + case done = <-w.done: + w.finish() + default: + } + return done +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_bsd.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_bsd.go new file mode 100644 index 000000000..e147d7638 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_bsd.go @@ -0,0 +1,93 @@ +// Copyright (c) 2012 VMware, Inc. + +// +build darwin freebsd netbsd openbsd + +// Go interface to BSD kqueue process events. +package psnotify + +import ( + "syscall" +) + +const ( + // Flags (from ) + PROC_EVENT_FORK = syscall.NOTE_FORK // fork() events + PROC_EVENT_EXEC = syscall.NOTE_EXEC // exec() events + PROC_EVENT_EXIT = syscall.NOTE_EXIT // exit() events + + // Watch for all process events + PROC_EVENT_ALL = PROC_EVENT_FORK | PROC_EVENT_EXEC | PROC_EVENT_EXIT +) + +type kqueueListener struct { + kq int // The syscall.Kqueue() file descriptor + buf [1]syscall.Kevent_t // An event buffer for Add/Remove watch +} + +// Initialize bsd implementation of the eventListener interface +func createListener() (eventListener, error) { + listener := &kqueueListener{} + kq, err := syscall.Kqueue() + listener.kq = kq + return listener, err +} + +// Initialize Kevent_t fields and propagate changelist for the given pid +func (w *Watcher) kevent(pid int, fflags uint32, flags int) error { + listener, _ := w.listener.(*kqueueListener) + event := &listener.buf[0] + + syscall.SetKevent(event, pid, syscall.EVFILT_PROC, flags) + event.Fflags = fflags + + _, err := syscall.Kevent(listener.kq, listener.buf[:], nil, nil) + + return err +} + +// Delete filter for given pid from the queue +func (w *Watcher) unregister(pid int) error { + return w.kevent(pid, 0, syscall.EV_DELETE) +} + +// Add and enable filter for given pid in the queue +func (w *Watcher) register(pid int, flags uint32) error { + return w.kevent(pid, flags, syscall.EV_ADD|syscall.EV_ENABLE) +} + +// Poll the kqueue file descriptor and dispatch to the Event channels +func (w *Watcher) readEvents() { + listener, _ := w.listener.(*kqueueListener) + events := make([]syscall.Kevent_t, 10) + + for { + if w.isDone() { + return + } + + n, err := syscall.Kevent(listener.kq, nil, events, nil) + if err != nil { + w.Error <- err + continue + } + + for _, ev := range events[:n] { + pid := int(ev.Ident) + + switch ev.Fflags { + case syscall.NOTE_FORK: + w.Fork <- &ProcEventFork{ParentPid: pid} + case syscall.NOTE_EXEC: + w.Exec <- &ProcEventExec{Pid: pid} + case syscall.NOTE_EXIT: + w.RemoveWatch(pid) + w.Exit <- &ProcEventExit{Pid: pid} + } + } + } +} + +// Close our kqueue file descriptor; deletes any remaining filters +func (listener *kqueueListener) close() error { + return syscall.Close(listener.kq) +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_linux.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_linux.go new file mode 100644 index 000000000..f9154ef3d --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_linux.go @@ -0,0 +1,253 @@ +// Copyright (c) 2012 VMware, Inc. + +// Go interface to the Linux netlink process connector. +// See Documentation/connector/connector.txt in the linux kernel source tree. +package psnotify + +import ( + "bytes" + "encoding/binary" + "os" + "syscall" +) + +const ( + // internal flags (from ) + _CN_IDX_PROC = 0x1 + _CN_VAL_PROC = 0x1 + + // internal flags (from ) + _PROC_CN_MCAST_LISTEN = 1 + _PROC_CN_MCAST_IGNORE = 2 + + // Flags (from ) + PROC_EVENT_FORK = 0x00000001 // fork() events + PROC_EVENT_EXEC = 0x00000002 // exec() events + PROC_EVENT_EXIT = 0x80000000 // exit() events + + // Watch for all process events + PROC_EVENT_ALL = PROC_EVENT_FORK | PROC_EVENT_EXEC | PROC_EVENT_EXIT +) + +var ( + byteOrder = binary.LittleEndian +) + +// linux/connector.h: struct cb_id +type cbId struct { + Idx uint32 + Val uint32 +} + +// linux/connector.h: struct cb_msg +type cnMsg struct { + Id cbId + Seq uint32 + Ack uint32 + Len uint16 + Flags uint16 +} + +// linux/cn_proc.h: struct proc_event.{what,cpu,timestamp_ns} +type procEventHeader struct { + What uint32 + Cpu uint32 + Timestamp uint64 +} + +// linux/cn_proc.h: struct proc_event.fork +type forkProcEvent struct { + ParentPid uint32 + ParentTgid uint32 + ChildPid uint32 + ChildTgid uint32 +} + +// linux/cn_proc.h: struct proc_event.exec +type execProcEvent struct { + ProcessPid uint32 + ProcessTgid uint32 +} + +// linux/cn_proc.h: struct proc_event.exit +type exitProcEvent struct { + ProcessPid uint32 + ProcessTgid uint32 + ExitCode uint32 + ExitSignal uint32 +} + +// standard netlink header + connector header +type netlinkProcMessage struct { + Header syscall.NlMsghdr + Data cnMsg +} + +type netlinkListener struct { + addr *syscall.SockaddrNetlink // Netlink socket address + sock int // The syscall.Socket() file descriptor + seq uint32 // struct cn_msg.seq +} + +// Initialize linux implementation of the eventListener interface +func createListener() (eventListener, error) { + listener := &netlinkListener{} + err := listener.bind() + return listener, err +} + +// noop on linux +func (w *Watcher) unregister(pid int) error { + return nil +} + +// noop on linux +func (w *Watcher) register(pid int, flags uint32) error { + return nil +} + +// Read events from the netlink socket +func (w *Watcher) readEvents() { + buf := make([]byte, syscall.Getpagesize()) + + listener, _ := w.listener.(*netlinkListener) + + for { + if w.isDone() { + return + } + + nr, _, err := syscall.Recvfrom(listener.sock, buf, 0) + + if err != nil { + w.Error <- err + continue + } + if nr < syscall.NLMSG_HDRLEN { + w.Error <- syscall.EINVAL + continue + } + + msgs, _ := syscall.ParseNetlinkMessage(buf[:nr]) + + for _, m := range msgs { + if m.Header.Type == syscall.NLMSG_DONE { + w.handleEvent(m.Data) + } + } + } +} + +// Internal helper to check if pid && event is being watched +func (w *Watcher) isWatching(pid int, event uint32) bool { + if watch, ok := w.watches[pid]; ok { + return (watch.flags & event) == event + } + return false +} + +// Dispatch events from the netlink socket to the Event channels. +// Unlike bsd kqueue, netlink receives events for all pids, +// so we apply filtering based on the watch table via isWatching() +func (w *Watcher) handleEvent(data []byte) { + buf := bytes.NewBuffer(data) + msg := &cnMsg{} + hdr := &procEventHeader{} + + binary.Read(buf, byteOrder, msg) + binary.Read(buf, byteOrder, hdr) + + switch hdr.What { + case PROC_EVENT_FORK: + event := &forkProcEvent{} + binary.Read(buf, byteOrder, event) + ppid := int(event.ParentTgid) + pid := int(event.ChildTgid) + + if w.isWatching(ppid, PROC_EVENT_EXEC) { + // follow forks + watch, _ := w.watches[ppid] + w.Watch(pid, watch.flags) + } + + if w.isWatching(ppid, PROC_EVENT_FORK) { + w.Fork <- &ProcEventFork{ParentPid: ppid, ChildPid: pid} + } + case PROC_EVENT_EXEC: + event := &execProcEvent{} + binary.Read(buf, byteOrder, event) + pid := int(event.ProcessTgid) + + if w.isWatching(pid, PROC_EVENT_EXEC) { + w.Exec <- &ProcEventExec{Pid: pid} + } + case PROC_EVENT_EXIT: + event := &exitProcEvent{} + binary.Read(buf, byteOrder, event) + pid := int(event.ProcessTgid) + + if w.isWatching(pid, PROC_EVENT_EXIT) { + w.RemoveWatch(pid) + w.Exit <- &ProcEventExit{Pid: pid} + } + } +} + +// Bind our netlink socket and +// send a listen control message to the connector driver. +func (listener *netlinkListener) bind() error { + sock, err := syscall.Socket( + syscall.AF_NETLINK, + syscall.SOCK_DGRAM, + syscall.NETLINK_CONNECTOR) + + if err != nil { + return err + } + + listener.sock = sock + listener.addr = &syscall.SockaddrNetlink{ + Family: syscall.AF_NETLINK, + Groups: _CN_IDX_PROC, + } + + err = syscall.Bind(listener.sock, listener.addr) + + if err != nil { + return err + } + + return listener.send(_PROC_CN_MCAST_LISTEN) +} + +// Send an ignore control message to the connector driver +// and close our netlink socket. +func (listener *netlinkListener) close() error { + err := listener.send(_PROC_CN_MCAST_IGNORE) + syscall.Close(listener.sock) + return err +} + +// Generic method for sending control messages to the connector +// driver; where op is one of PROC_CN_MCAST_{LISTEN,IGNORE} +func (listener *netlinkListener) send(op uint32) error { + listener.seq++ + pr := &netlinkProcMessage{} + plen := binary.Size(pr.Data) + binary.Size(op) + pr.Header.Len = syscall.NLMSG_HDRLEN + uint32(plen) + pr.Header.Type = uint16(syscall.NLMSG_DONE) + pr.Header.Flags = 0 + pr.Header.Seq = listener.seq + pr.Header.Pid = uint32(os.Getpid()) + + pr.Data.Id.Idx = _CN_IDX_PROC + pr.Data.Id.Val = _CN_VAL_PROC + + pr.Data.Len = uint16(binary.Size(op)) + + buf := bytes.NewBuffer(make([]byte, 0, pr.Header.Len)) + binary.Write(buf, byteOrder, pr) + binary.Write(buf, byteOrder, op) + + return syscall.Sendto(listener.sock, buf.Bytes(), 0, listener.addr) +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_test.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_test.go new file mode 100644 index 000000000..28f38a8d7 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/psnotify/psnotify_test.go @@ -0,0 +1,283 @@ +// Copyright (c) 2012 VMware, Inc. + +package psnotify + +import ( + "fmt" + "os" + "os/exec" + "runtime" + "syscall" + "testing" + "time" +) + +type anyEvent struct { + exits []int + forks []int + execs []int + errors []error + done chan bool +} + +type testWatcher struct { + t *testing.T + watcher *Watcher + events *anyEvent +} + +// General purpose Watcher wrapper for all tests +func newTestWatcher(t *testing.T) *testWatcher { + watcher, err := NewWatcher() + if err != nil { + t.Fatal(err) + } + + events := &anyEvent{ + done: make(chan bool, 1), + } + + tw := &testWatcher{ + t: t, + watcher: watcher, + events: events, + } + + go func() { + for { + select { + case <-events.done: + return + case ev := <-watcher.Fork: + events.forks = append(events.forks, ev.ParentPid) + case ev := <-watcher.Exec: + events.execs = append(events.execs, ev.Pid) + case ev := <-watcher.Exit: + events.exits = append(events.exits, ev.Pid) + case err := <-watcher.Error: + events.errors = append(events.errors, err) + } + } + }() + + return tw +} + +func (tw *testWatcher) close() { + pause := 100 * time.Millisecond + time.Sleep(pause) + + tw.events.done <- true + + tw.watcher.Close() + + time.Sleep(pause) +} + +func skipTest(t *testing.T) bool { + if runtime.GOOS == "linux" && os.Getuid() != 0 { + fmt.Println("SKIP: test must be run as root on linux") + return true + } + return false +} + +func startSleepCommand(t *testing.T) *exec.Cmd { + cmd := exec.Command("sh", "-c", "sleep 100") + if err := cmd.Start(); err != nil { + t.Error(err) + } + return cmd +} + +func runCommand(t *testing.T, name string) *exec.Cmd { + cmd := exec.Command(name) + if err := cmd.Run(); err != nil { + t.Error(err) + } + return cmd +} + +func expectEvents(t *testing.T, num int, name string, pids []int) bool { + if len(pids) != num { + t.Errorf("Expected %d %s events, got=%v", num, name, pids) + return false + } + return true +} + +func expectEventPid(t *testing.T, name string, expect int, pid int) bool { + if expect != pid { + t.Errorf("Expected %s pid=%d, received=%d", name, expect, pid) + return false + } + return true +} + +func TestWatchFork(t *testing.T) { + if skipTest(t) { + return + } + + pid := os.Getpid() + + tw := newTestWatcher(t) + + // no watches added yet, so this fork event will no be captured + runCommand(t, "date") + + // watch fork events for this process + if err := tw.watcher.Watch(pid, PROC_EVENT_FORK); err != nil { + t.Error(err) + } + + // this fork event will be captured, + // the exec and exit events will not be captured + runCommand(t, "cal") + + tw.close() + + if expectEvents(t, 1, "forks", tw.events.forks) { + expectEventPid(t, "fork", pid, tw.events.forks[0]) + } + + expectEvents(t, 0, "execs", tw.events.execs) + expectEvents(t, 0, "exits", tw.events.exits) +} + +func TestWatchExit(t *testing.T) { + if skipTest(t) { + return + } + + tw := newTestWatcher(t) + + cmd := startSleepCommand(t) + + childPid := cmd.Process.Pid + + // watch for exit event of our child process + if err := tw.watcher.Watch(childPid, PROC_EVENT_EXIT); err != nil { + t.Error(err) + } + + // kill our child process, triggers exit event + syscall.Kill(childPid, syscall.SIGTERM) + + cmd.Wait() + + tw.close() + + expectEvents(t, 0, "forks", tw.events.forks) + + expectEvents(t, 0, "execs", tw.events.execs) + + if expectEvents(t, 1, "exits", tw.events.exits) { + expectEventPid(t, "exit", childPid, tw.events.exits[0]) + } +} + +// combined version of TestWatchFork() and TestWatchExit() +func TestWatchForkAndExit(t *testing.T) { + if skipTest(t) { + return + } + + pid := os.Getpid() + + tw := newTestWatcher(t) + + if err := tw.watcher.Watch(pid, PROC_EVENT_FORK); err != nil { + t.Error(err) + } + + cmd := startSleepCommand(t) + + childPid := cmd.Process.Pid + + if err := tw.watcher.Watch(childPid, PROC_EVENT_EXIT); err != nil { + t.Error(err) + } + + syscall.Kill(childPid, syscall.SIGTERM) + + cmd.Wait() + + tw.close() + + if expectEvents(t, 1, "forks", tw.events.forks) { + expectEventPid(t, "fork", pid, tw.events.forks[0]) + } + + expectEvents(t, 0, "execs", tw.events.execs) + + if expectEvents(t, 1, "exits", tw.events.exits) { + expectEventPid(t, "exit", childPid, tw.events.exits[0]) + } +} + +func TestWatchFollowFork(t *testing.T) { + if skipTest(t) { + return + } + + // Darwin is not able to follow forks, as the kqueue fork event + // does not provide the child pid. + if runtime.GOOS != "linux" { + fmt.Println("SKIP: test follow forks is linux only") + return + } + + pid := os.Getpid() + + tw := newTestWatcher(t) + + // watch for all process events related to this process + if err := tw.watcher.Watch(pid, PROC_EVENT_ALL); err != nil { + t.Error(err) + } + + commands := []string{"date", "cal"} + childPids := make([]int, len(commands)) + + // triggers fork/exec/exit events for each command + for i, name := range commands { + cmd := runCommand(t, name) + childPids[i] = cmd.Process.Pid + } + + // remove watch for this process + tw.watcher.RemoveWatch(pid) + + // run commands again to make sure we don't receive any unwanted events + for _, name := range commands { + runCommand(t, name) + } + + tw.close() + + // run commands again to make sure nothing panics after + // closing the watcher + for _, name := range commands { + runCommand(t, name) + } + + num := len(commands) + if expectEvents(t, num, "forks", tw.events.forks) { + for _, epid := range tw.events.forks { + expectEventPid(t, "fork", pid, epid) + } + } + + if expectEvents(t, num, "execs", tw.events.execs) { + for i, epid := range tw.events.execs { + expectEventPid(t, "exec", childPids[i], epid) + } + } + + if expectEvents(t, num, "exits", tw.events.exits) { + for i, epid := range tw.events.exits { + expectEventPid(t, "exit", childPids[i], epid) + } + } +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_darwin.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_darwin.go new file mode 100644 index 000000000..e3a8c4b9c --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_darwin.go @@ -0,0 +1,467 @@ +// Copyright (c) 2012 VMware, Inc. + +package sigar + +/* +#include +#include +#include +#include +#include +#include +#include +#include +#include +*/ +import "C" + +import ( + "bytes" + "encoding/binary" + "fmt" + "io" + "syscall" + "time" + "unsafe" +) + +func (self *LoadAverage) Get() error { + avg := []C.double{0, 0, 0} + + C.getloadavg(&avg[0], C.int(len(avg))) + + self.One = float64(avg[0]) + self.Five = float64(avg[1]) + self.Fifteen = float64(avg[2]) + + return nil +} + +func (self *Uptime) Get() error { + tv := syscall.Timeval32{} + + if err := sysctlbyname("kern.boottime", &tv); err != nil { + return err + } + + self.Length = time.Since(time.Unix(int64(tv.Sec), int64(tv.Usec)*1000)).Seconds() + + return nil +} + +func (self *Mem) Get() error { + var vmstat C.vm_statistics_data_t + + if err := sysctlbyname("hw.memsize", &self.Total); err != nil { + return err + } + + if err := vm_info(&vmstat); err != nil { + return err + } + + kern := uint64(vmstat.inactive_count) << 12 + self.Free = uint64(vmstat.free_count) << 12 + + self.Used = self.Total - self.Free + self.ActualFree = self.Free + kern + self.ActualUsed = self.Used - kern + + return nil +} + +type xsw_usage struct { + Total, Avail, Used uint64 +} + +func (self *Swap) Get() error { + sw_usage := xsw_usage{} + + if err := sysctlbyname("vm.swapusage", &sw_usage); err != nil { + return err + } + + self.Total = sw_usage.Total + self.Used = sw_usage.Used + self.Free = sw_usage.Avail + + return nil +} + +func (self *Cpu) Get() error { + var count C.mach_msg_type_number_t = C.HOST_CPU_LOAD_INFO_COUNT + var cpuload C.host_cpu_load_info_data_t + + status := C.host_statistics(C.host_t(C.mach_host_self()), + C.HOST_CPU_LOAD_INFO, + C.host_info_t(unsafe.Pointer(&cpuload)), + &count) + + if status != C.KERN_SUCCESS { + return fmt.Errorf("host_statistics error=%d", status) + } + + self.User = uint64(cpuload.cpu_ticks[C.CPU_STATE_USER]) + self.Sys = uint64(cpuload.cpu_ticks[C.CPU_STATE_SYSTEM]) + self.Idle = uint64(cpuload.cpu_ticks[C.CPU_STATE_IDLE]) + self.Nice = uint64(cpuload.cpu_ticks[C.CPU_STATE_NICE]) + + return nil +} + +func (self *CpuList) Get() error { + var count C.mach_msg_type_number_t + var cpuload *C.processor_cpu_load_info_data_t + var ncpu C.natural_t + + status := C.host_processor_info(C.host_t(C.mach_host_self()), + C.PROCESSOR_CPU_LOAD_INFO, + &ncpu, + (*C.processor_info_array_t)(unsafe.Pointer(&cpuload)), + &count) + + if status != C.KERN_SUCCESS { + return fmt.Errorf("host_processor_info error=%d", status) + } + + // jump through some cgo casting hoops and ensure we properly free + // the memory that cpuload points to + target := C.vm_map_t(C.mach_task_self_) + address := C.vm_address_t(uintptr(unsafe.Pointer(cpuload))) + defer C.vm_deallocate(target, address, C.vm_size_t(ncpu)) + + // the body of struct processor_cpu_load_info + // aka processor_cpu_load_info_data_t + var cpu_ticks [C.CPU_STATE_MAX]uint32 + + // copy the cpuload array to a []byte buffer + // where we can binary.Read the data + size := int(ncpu) * binary.Size(cpu_ticks) + buf := C.GoBytes(unsafe.Pointer(cpuload), C.int(size)) + + bbuf := bytes.NewBuffer(buf) + + self.List = make([]Cpu, 0, ncpu) + + for i := 0; i < int(ncpu); i++ { + cpu := Cpu{} + + err := binary.Read(bbuf, binary.LittleEndian, &cpu_ticks) + if err != nil { + return err + } + + cpu.User = uint64(cpu_ticks[C.CPU_STATE_USER]) + cpu.Sys = uint64(cpu_ticks[C.CPU_STATE_SYSTEM]) + cpu.Idle = uint64(cpu_ticks[C.CPU_STATE_IDLE]) + cpu.Nice = uint64(cpu_ticks[C.CPU_STATE_NICE]) + + self.List = append(self.List, cpu) + } + + return nil +} + +func (self *FileSystemList) Get() error { + num, err := getfsstat(nil, C.MNT_NOWAIT) + if num < 0 { + return err + } + + buf := make([]syscall.Statfs_t, num) + + num, err = getfsstat(buf, C.MNT_NOWAIT) + if err != nil { + return err + } + + fslist := make([]FileSystem, 0, num) + + for i := 0; i < num; i++ { + fs := FileSystem{} + + fs.DirName = bytePtrToString(&buf[i].Mntonname[0]) + fs.DevName = bytePtrToString(&buf[i].Mntfromname[0]) + fs.SysTypeName = bytePtrToString(&buf[i].Fstypename[0]) + + fslist = append(fslist, fs) + } + + self.List = fslist + + return err +} + +func (self *ProcList) Get() error { + n := C.proc_listpids(C.PROC_ALL_PIDS, 0, nil, 0) + if n <= 0 { + return syscall.EINVAL + } + buf := make([]byte, n) + n = C.proc_listpids(C.PROC_ALL_PIDS, 0, unsafe.Pointer(&buf[0]), n) + if n <= 0 { + return syscall.ENOMEM + } + + var pid int32 + num := int(n) / binary.Size(pid) + list := make([]int, 0, num) + bbuf := bytes.NewBuffer(buf) + + for i := 0; i < num; i++ { + if err := binary.Read(bbuf, binary.LittleEndian, &pid); err != nil { + return err + } + if pid == 0 { + continue + } + + list = append(list, int(pid)) + } + + self.List = list + + return nil +} + +func (self *ProcState) Get(pid int) error { + info := C.struct_proc_taskallinfo{} + + if err := task_info(pid, &info); err != nil { + return err + } + + self.Name = C.GoString(&info.pbsd.pbi_comm[0]) + + switch info.pbsd.pbi_status { + case C.SIDL: + self.State = RunStateIdle + case C.SRUN: + self.State = RunStateRun + case C.SSLEEP: + self.State = RunStateSleep + case C.SSTOP: + self.State = RunStateStop + case C.SZOMB: + self.State = RunStateZombie + default: + self.State = RunStateUnknown + } + + self.Ppid = int(info.pbsd.pbi_ppid) + + self.Tty = int(info.pbsd.e_tdev) + + self.Priority = int(info.ptinfo.pti_priority) + + self.Nice = int(info.pbsd.pbi_nice) + + return nil +} + +func (self *ProcMem) Get(pid int) error { + info := C.struct_proc_taskallinfo{} + + if err := task_info(pid, &info); err != nil { + return err + } + + self.Size = uint64(info.ptinfo.pti_virtual_size) + self.Resident = uint64(info.ptinfo.pti_resident_size) + self.PageFaults = uint64(info.ptinfo.pti_faults) + + return nil +} + +func (self *ProcTime) Get(pid int) error { + info := C.struct_proc_taskallinfo{} + + if err := task_info(pid, &info); err != nil { + return err + } + + self.User = + uint64(info.ptinfo.pti_total_user) / uint64(time.Millisecond) + + self.Sys = + uint64(info.ptinfo.pti_total_system) / uint64(time.Millisecond) + + self.Total = self.User + self.Sys + + self.StartTime = (uint64(info.pbsd.pbi_start_tvsec) * 1000) + + (uint64(info.pbsd.pbi_start_tvusec) / 1000) + + return nil +} + +func (self *ProcArgs) Get(pid int) error { + var args []string + + argv := func(arg string) { + args = append(args, arg) + } + + err := kern_procargs(pid, nil, argv, nil) + + self.List = args + + return err +} + +func (self *ProcExe) Get(pid int) error { + exe := func(arg string) { + self.Name = arg + } + + return kern_procargs(pid, exe, nil, nil) +} + +// wrapper around sysctl KERN_PROCARGS2 +// callbacks params are optional, +// up to the caller as to which pieces of data they want +func kern_procargs(pid int, + exe func(string), + argv func(string), + env func(string, string)) error { + + mib := []C.int{C.CTL_KERN, C.KERN_PROCARGS2, C.int(pid)} + argmax := uintptr(C.ARG_MAX) + buf := make([]byte, argmax) + err := sysctl(mib, &buf[0], &argmax, nil, 0) + if err != nil { + return nil + } + + bbuf := bytes.NewBuffer(buf) + bbuf.Truncate(int(argmax)) + + var argc int32 + binary.Read(bbuf, binary.LittleEndian, &argc) + + path, err := bbuf.ReadBytes(0) + if exe != nil { + exe(string(chop(path))) + } + + // skip trailing \0's + for { + c, _ := bbuf.ReadByte() + if c != 0 { + bbuf.UnreadByte() + break // start of argv[0] + } + } + + for i := 0; i < int(argc); i++ { + arg, err := bbuf.ReadBytes(0) + if err == io.EOF { + break + } + if argv != nil { + argv(string(chop(arg))) + } + } + + if env == nil { + return nil + } + + delim := []byte{61} // "=" + + for { + line, err := bbuf.ReadBytes(0) + if err == io.EOF || line[0] == 0 { + break + } + pair := bytes.SplitN(chop(line), delim, 2) + env(string(pair[0]), string(pair[1])) + } + + return nil +} + +// XXX copied from zsyscall_darwin_amd64.go +func sysctl(mib []C.int, old *byte, oldlen *uintptr, + new *byte, newlen uintptr) (err error) { + var p0 unsafe.Pointer + p0 = unsafe.Pointer(&mib[0]) + _, _, e1 := syscall.Syscall6(syscall.SYS___SYSCTL, uintptr(p0), + uintptr(len(mib)), + uintptr(unsafe.Pointer(old)), uintptr(unsafe.Pointer(oldlen)), + uintptr(unsafe.Pointer(new)), uintptr(newlen)) + if e1 != 0 { + err = e1 + } + return +} + +func vm_info(vmstat *C.vm_statistics_data_t) error { + var count C.mach_msg_type_number_t = C.HOST_VM_INFO_COUNT + + status := C.host_statistics( + C.host_t(C.mach_host_self()), + C.HOST_VM_INFO, + C.host_info_t(unsafe.Pointer(vmstat)), + &count) + + if status != C.KERN_SUCCESS { + return fmt.Errorf("host_statistics=%d", status) + } + + return nil +} + +// generic Sysctl buffer unmarshalling +func sysctlbyname(name string, data interface{}) (err error) { + val, err := syscall.Sysctl(name) + if err != nil { + return err + } + + buf := []byte(val) + + switch v := data.(type) { + case *uint64: + *v = *(*uint64)(unsafe.Pointer(&buf[0])) + return + } + + bbuf := bytes.NewBuffer([]byte(val)) + return binary.Read(bbuf, binary.LittleEndian, data) +} + +// syscall.Getfsstat() wrapper is broken, roll our own to workaround. +func getfsstat(buf []syscall.Statfs_t, flags int) (n int, err error) { + var ptr uintptr + var size uintptr + + if len(buf) > 0 { + ptr = uintptr(unsafe.Pointer(&buf[0])) + size = unsafe.Sizeof(buf[0]) * uintptr(len(buf)) + } else { + ptr = uintptr(0) + size = uintptr(0) + } + + trap := uintptr(syscall.SYS_GETFSSTAT64) + ret, _, errno := syscall.Syscall(trap, ptr, size, uintptr(flags)) + + n = int(ret) + if errno != 0 { + err = errno + } + + return +} + +func task_info(pid int, info *C.struct_proc_taskallinfo) error { + size := C.int(unsafe.Sizeof(*info)) + ptr := unsafe.Pointer(info) + + n := C.proc_pidinfo(C.int(pid), C.PROC_PIDTASKALLINFO, 0, ptr, size) + if n != size { + return syscall.ENOMEM + } + + return nil +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_format.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_format.go new file mode 100644 index 000000000..d80a64e88 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_format.go @@ -0,0 +1,126 @@ +// Copyright (c) 2012 VMware, Inc. + +package sigar + +import ( + "bufio" + "bytes" + "fmt" + "strconv" + "time" +) + +// Go version of apr_strfsize +func FormatSize(size uint64) string { + ord := []string{"K", "M", "G", "T", "P", "E"} + o := 0 + buf := new(bytes.Buffer) + w := bufio.NewWriter(buf) + + if size < 973 { + fmt.Fprintf(w, "%3d ", size) + w.Flush() + return buf.String() + } + + for { + remain := size & 1023 + size >>= 10 + + if size >= 973 { + o++ + continue + } + + if size < 9 || (size == 9 && remain < 973) { + remain = ((remain * 5) + 256) / 512 + if remain >= 10 { + size++ + remain = 0 + } + + fmt.Fprintf(w, "%d.%d%s", size, remain, ord[o]) + break + } + + if remain >= 512 { + size++ + } + + fmt.Fprintf(w, "%3d%s", size, ord[o]) + break + } + + w.Flush() + return buf.String() +} + +func FormatPercent(percent float64) string { + return strconv.FormatFloat(percent, 'f', -1, 64) + "%" +} + +func (self *FileSystemUsage) UsePercent() float64 { + b_used := (self.Total - self.Free) / 1024 + b_avail := self.Avail / 1024 + utotal := b_used + b_avail + used := b_used + + if utotal != 0 { + u100 := used * 100 + pct := u100 / utotal + if u100%utotal != 0 { + pct += 1 + } + return (float64(pct) / float64(100)) * 100.0 + } + + return 0.0 +} + +func (self *Uptime) Format() string { + buf := new(bytes.Buffer) + w := bufio.NewWriter(buf) + uptime := uint64(self.Length) + + days := uptime / (60 * 60 * 24) + + if days != 0 { + s := "" + if days > 1 { + s = "s" + } + fmt.Fprintf(w, "%d day%s, ", days, s) + } + + minutes := uptime / 60 + hours := minutes / 60 + hours %= 24 + minutes %= 60 + + fmt.Fprintf(w, "%2d:%02d", hours, minutes) + + w.Flush() + return buf.String() +} + +func (self *ProcTime) FormatStartTime() string { + if self.StartTime == 0 { + return "00:00" + } + start := time.Unix(int64(self.StartTime)/1000, 0) + format := "Jan02" + if time.Since(start).Seconds() < (60 * 60 * 24) { + format = "15:04" + } + return start.Format(format) +} + +func (self *ProcTime) FormatTotal() string { + t := self.Total / 1000 + ss := t % 60 + t /= 60 + mm := t % 60 + t /= 60 + hh := t % 24 + return fmt.Sprintf("%02d:%02d:%02d", hh, mm, ss) +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_interface.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_interface.go new file mode 100644 index 000000000..dd72a76b0 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_interface.go @@ -0,0 +1,141 @@ +package sigar + +import ( + "time" +) + +type Sigar interface { + CollectCpuStats(collectionInterval time.Duration) (<-chan Cpu, chan<- struct{}) + GetLoadAverage() (LoadAverage, error) + GetMem() (Mem, error) + GetSwap() (Swap, error) + GetFileSystemUsage(string) (FileSystemUsage, error) +} + +type Cpu struct { + User uint64 + Nice uint64 + Sys uint64 + Idle uint64 + Wait uint64 + Irq uint64 + SoftIrq uint64 + Stolen uint64 +} + +func (cpu *Cpu) Total() uint64 { + return cpu.User + cpu.Nice + cpu.Sys + cpu.Idle + + cpu.Wait + cpu.Irq + cpu.SoftIrq + cpu.Stolen +} + +func (cpu Cpu) Delta(other Cpu) Cpu { + return Cpu{ + User: cpu.User - other.User, + Nice: cpu.Nice - other.Nice, + Sys: cpu.Sys - other.Sys, + Idle: cpu.Idle - other.Idle, + Wait: cpu.Wait - other.Wait, + Irq: cpu.Irq - other.Irq, + SoftIrq: cpu.SoftIrq - other.SoftIrq, + Stolen: cpu.Stolen - other.Stolen, + } +} + +type LoadAverage struct { + One, Five, Fifteen float64 +} + +type Uptime struct { + Length float64 +} + +type Mem struct { + Total uint64 + Used uint64 + Free uint64 + ActualFree uint64 + ActualUsed uint64 +} + +type Swap struct { + Total uint64 + Used uint64 + Free uint64 +} + +type CpuList struct { + List []Cpu +} + +type FileSystem struct { + DirName string + DevName string + TypeName string + SysTypeName string + Options string + Flags uint32 +} + +type FileSystemList struct { + List []FileSystem +} + +type FileSystemUsage struct { + Total uint64 + Used uint64 + Free uint64 + Avail uint64 + Files uint64 + FreeFiles uint64 +} + +type ProcList struct { + List []int +} + +type RunState byte + +const ( + RunStateSleep = 'S' + RunStateRun = 'R' + RunStateStop = 'T' + RunStateZombie = 'Z' + RunStateIdle = 'D' + RunStateUnknown = '?' +) + +type ProcState struct { + Name string + State RunState + Ppid int + Tty int + Priority int + Nice int + Processor int +} + +type ProcMem struct { + Size uint64 + Resident uint64 + Share uint64 + MinorFaults uint64 + MajorFaults uint64 + PageFaults uint64 +} + +type ProcTime struct { + StartTime uint64 + User uint64 + Sys uint64 + Total uint64 +} + +type ProcArgs struct { + List []string +} + +type ProcExe struct { + Name string + Cwd string + Root string +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_interface_test.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_interface_test.go new file mode 100644 index 000000000..fe26abd1b --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_interface_test.go @@ -0,0 +1,135 @@ +package sigar_test + +import ( + "os" + "path/filepath" + "runtime" + + . "github.com/onsi/ginkgo" + . "github.com/onsi/gomega" + + . "github.com/cloudfoundry/gosigar" +) + +var _ = Describe("Sigar", func() { + var invalidPid = 666666 + + It("cpu", func() { + cpu := Cpu{} + err := cpu.Get() + Expect(err).ToNot(HaveOccurred()) + }) + + It("load average", func() { + avg := LoadAverage{} + err := avg.Get() + Expect(err).ToNot(HaveOccurred()) + }) + + It("uptime", func() { + uptime := Uptime{} + err := uptime.Get() + Expect(err).ToNot(HaveOccurred()) + Expect(uptime.Length).To(BeNumerically(">", 0)) + }) + + It("mem", func() { + mem := Mem{} + err := mem.Get() + Expect(err).ToNot(HaveOccurred()) + + Expect(mem.Total).To(BeNumerically(">", 0)) + Expect(mem.Used + mem.Free).To(BeNumerically("<=", mem.Total)) + }) + + It("swap", func() { + swap := Swap{} + err := swap.Get() + Expect(err).ToNot(HaveOccurred()) + Expect(swap.Used + swap.Free).To(BeNumerically("<=", swap.Total)) + }) + + It("cpu list", func() { + cpulist := CpuList{} + err := cpulist.Get() + Expect(err).ToNot(HaveOccurred()) + + nsigar := len(cpulist.List) + numcpu := runtime.NumCPU() + Expect(nsigar).To(Equal(numcpu)) + }) + + It("file system list", func() { + fslist := FileSystemList{} + err := fslist.Get() + Expect(err).ToNot(HaveOccurred()) + + Expect(len(fslist.List)).To(BeNumerically(">", 0)) + }) + + It("file system usage", func() { + fsusage := FileSystemUsage{} + err := fsusage.Get("/") + Expect(err).ToNot(HaveOccurred()) + + err = fsusage.Get("T O T A L L Y B O G U S") + Expect(err).To(HaveOccurred()) + }) + + It("proc list", func() { + pids := ProcList{} + err := pids.Get() + Expect(err).ToNot(HaveOccurred()) + + Expect(len(pids.List)).To(BeNumerically(">", 2)) + + err = pids.Get() + Expect(err).ToNot(HaveOccurred()) + }) + + It("proc state", func() { + state := ProcState{} + err := state.Get(os.Getppid()) + Expect(err).ToNot(HaveOccurred()) + + Expect([]RunState{RunStateRun, RunStateSleep}).To(ContainElement(state.State)) + Expect([]string{"go", "ginkgo"}).To(ContainElement(state.Name)) + + err = state.Get(invalidPid) + Expect(err).To(HaveOccurred()) + }) + + It("proc mem", func() { + mem := ProcMem{} + err := mem.Get(os.Getppid()) + Expect(err).ToNot(HaveOccurred()) + + err = mem.Get(invalidPid) + Expect(err).To(HaveOccurred()) + }) + + It("proc time", func() { + time := ProcTime{} + err := time.Get(os.Getppid()) + Expect(err).ToNot(HaveOccurred()) + + err = time.Get(invalidPid) + Expect(err).To(HaveOccurred()) + }) + + It("proc args", func() { + args := ProcArgs{} + err := args.Get(os.Getppid()) + Expect(err).ToNot(HaveOccurred()) + + Expect(len(args.List)).To(BeNumerically(">=", 2)) + }) + + It("proc exe", func() { + exe := ProcExe{} + err := exe.Get(os.Getppid()) + Expect(err).ToNot(HaveOccurred()) + + Expect([]string{"go", "ginkgo"}).To(ContainElement(filepath.Base(exe.Name))) + }) +}) diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_linux.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_linux.go new file mode 100644 index 000000000..68ffb0f9a --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_linux.go @@ -0,0 +1,386 @@ +// Copyright (c) 2012 VMware, Inc. + +package sigar + +import ( + "bufio" + "bytes" + "io" + "io/ioutil" + "os" + "strconv" + "strings" + "syscall" +) + +var system struct { + ticks uint64 + btime uint64 +} + +var Procd string + +func init() { + system.ticks = 100 // C.sysconf(C._SC_CLK_TCK) + + Procd = "/proc" + + // grab system boot time + readFile(Procd+"/stat", func(line string) bool { + if strings.HasPrefix(line, "btime") { + system.btime, _ = strtoull(line[6:]) + return false // stop reading + } + return true + }) +} + +func (self *LoadAverage) Get() error { + line, err := ioutil.ReadFile(Procd + "/loadavg") + if err != nil { + return nil + } + + fields := strings.Fields(string(line)) + + self.One, _ = strconv.ParseFloat(fields[0], 64) + self.Five, _ = strconv.ParseFloat(fields[1], 64) + self.Fifteen, _ = strconv.ParseFloat(fields[2], 64) + + return nil +} + +func (self *Uptime) Get() error { + sysinfo := syscall.Sysinfo_t{} + + if err := syscall.Sysinfo(&sysinfo); err != nil { + return err + } + + self.Length = float64(sysinfo.Uptime) + + return nil +} + +func (self *Mem) Get() error { + var buffers, cached uint64 + table := map[string]*uint64{ + "MemTotal": &self.Total, + "MemFree": &self.Free, + "Buffers": &buffers, + "Cached": &cached, + } + + if err := parseMeminfo(table); err != nil { + return err + } + + self.Used = self.Total - self.Free + kern := buffers + cached + self.ActualFree = self.Free + kern + self.ActualUsed = self.Used - kern + + return nil +} + +func (self *Swap) Get() error { + table := map[string]*uint64{ + "SwapTotal": &self.Total, + "SwapFree": &self.Free, + } + + if err := parseMeminfo(table); err != nil { + return err + } + + self.Used = self.Total - self.Free + return nil +} + +func (self *Cpu) Get() error { + return readFile(Procd+"/stat", func(line string) bool { + if len(line) > 4 && line[0:4] == "cpu " { + parseCpuStat(self, line) + return false + } + return true + + }) +} + +func (self *CpuList) Get() error { + capacity := len(self.List) + if capacity == 0 { + capacity = 4 + } + list := make([]Cpu, 0, capacity) + + err := readFile(Procd+"/stat", func(line string) bool { + if len(line) > 3 && line[0:3] == "cpu" && line[3] != ' ' { + cpu := Cpu{} + parseCpuStat(&cpu, line) + list = append(list, cpu) + } + return true + }) + + self.List = list + + return err +} + +func (self *FileSystemList) Get() error { + capacity := len(self.List) + if capacity == 0 { + capacity = 10 + } + fslist := make([]FileSystem, 0, capacity) + + err := readFile("/etc/mtab", func(line string) bool { + fields := strings.Fields(line) + + fs := FileSystem{} + fs.DevName = fields[0] + fs.DirName = fields[1] + fs.SysTypeName = fields[2] + fs.Options = fields[3] + + fslist = append(fslist, fs) + + return true + }) + + self.List = fslist + + return err +} + +func (self *ProcList) Get() error { + dir, err := os.Open(Procd) + if err != nil { + return err + } + defer dir.Close() + + const readAllDirnames = -1 // see os.File.Readdirnames doc + + names, err := dir.Readdirnames(readAllDirnames) + if err != nil { + return err + } + + capacity := len(names) + list := make([]int, 0, capacity) + + for _, name := range names { + if name[0] < '0' || name[0] > '9' { + continue + } + pid, err := strconv.Atoi(name) + if err == nil { + list = append(list, pid) + } + } + + self.List = list + + return nil +} + +func (self *ProcState) Get(pid int) error { + contents, err := readProcFile(pid, "stat") + if err != nil { + return err + } + + fields := strings.Fields(string(contents)) + + self.Name = fields[1][1 : len(fields[1])-1] // strip ()'s + + self.State = RunState(fields[2][0]) + + self.Ppid, _ = strconv.Atoi(fields[3]) + + self.Tty, _ = strconv.Atoi(fields[6]) + + self.Priority, _ = strconv.Atoi(fields[17]) + + self.Nice, _ = strconv.Atoi(fields[18]) + + self.Processor, _ = strconv.Atoi(fields[38]) + + return nil +} + +func (self *ProcMem) Get(pid int) error { + contents, err := readProcFile(pid, "statm") + if err != nil { + return err + } + + fields := strings.Fields(string(contents)) + + size, _ := strtoull(fields[0]) + self.Size = size << 12 + + rss, _ := strtoull(fields[1]) + self.Resident = rss << 12 + + share, _ := strtoull(fields[2]) + self.Share = share << 12 + + contents, err = readProcFile(pid, "stat") + if err != nil { + return err + } + + fields = strings.Fields(string(contents)) + + self.MinorFaults, _ = strtoull(fields[10]) + self.MajorFaults, _ = strtoull(fields[12]) + self.PageFaults = self.MinorFaults + self.MajorFaults + + return nil +} + +func (self *ProcTime) Get(pid int) error { + contents, err := readProcFile(pid, "stat") + if err != nil { + return err + } + + fields := strings.Fields(string(contents)) + + user, _ := strtoull(fields[13]) + sys, _ := strtoull(fields[14]) + // convert to millis + self.User = user * (1000 / system.ticks) + self.Sys = sys * (1000 / system.ticks) + self.Total = self.User + self.Sys + + // convert to millis + self.StartTime, _ = strtoull(fields[21]) + self.StartTime /= system.ticks + self.StartTime += system.btime + self.StartTime *= 1000 + + return nil +} + +func (self *ProcArgs) Get(pid int) error { + contents, err := readProcFile(pid, "cmdline") + if err != nil { + return err + } + + bbuf := bytes.NewBuffer(contents) + + var args []string + + for { + arg, err := bbuf.ReadBytes(0) + if err == io.EOF { + break + } + args = append(args, string(chop(arg))) + } + + self.List = args + + return nil +} + +func (self *ProcExe) Get(pid int) error { + fields := map[string]*string{ + "exe": &self.Name, + "cwd": &self.Cwd, + "root": &self.Root, + } + + for name, field := range fields { + val, err := os.Readlink(procFileName(pid, name)) + + if err != nil { + return err + } + + *field = val + } + + return nil +} + +func parseMeminfo(table map[string]*uint64) error { + return readFile(Procd+"/meminfo", func(line string) bool { + fields := strings.Split(line, ":") + + if ptr := table[fields[0]]; ptr != nil { + num := strings.TrimLeft(fields[1], " ") + val, err := strtoull(strings.Fields(num)[0]) + if err == nil { + *ptr = val * 1024 + } + } + + return true + }) +} + +func parseCpuStat(self *Cpu, line string) error { + fields := strings.Fields(line) + + self.User, _ = strtoull(fields[1]) + self.Nice, _ = strtoull(fields[2]) + self.Sys, _ = strtoull(fields[3]) + self.Idle, _ = strtoull(fields[4]) + self.Wait, _ = strtoull(fields[5]) + self.Irq, _ = strtoull(fields[6]) + self.SoftIrq, _ = strtoull(fields[7]) + self.Stolen, _ = strtoull(fields[8]) + + return nil +} + +func readFile(file string, handler func(string) bool) error { + contents, err := ioutil.ReadFile(file) + if err != nil { + return err + } + + reader := bufio.NewReader(bytes.NewBuffer(contents)) + + for { + line, _, err := reader.ReadLine() + if err == io.EOF { + break + } + if !handler(string(line)) { + break + } + } + + return nil +} + +func strtoull(val string) (uint64, error) { + return strconv.ParseUint(val, 10, 64) +} + +func procFileName(pid int, name string) string { + return Procd + "/" + strconv.Itoa(pid) + "/" + name +} + +func readProcFile(pid int, name string) ([]byte, error) { + path := procFileName(pid, name) + contents, err := ioutil.ReadFile(path) + + if err != nil { + if perr, ok := err.(*os.PathError); ok { + if perr.Err == syscall.ENOENT { + return nil, syscall.ESRCH + } + } + } + + return contents, err +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_linux_test.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_linux_test.go new file mode 100644 index 000000000..c5fcdbc9a --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_linux_test.go @@ -0,0 +1,225 @@ +package sigar_test + +import ( + "io/ioutil" + "time" + + . "github.com/onsi/ginkgo" + . "github.com/onsi/gomega" + + sigar "github.com/cloudfoundry/gosigar" +) + +var _ = Describe("sigarLinux", func() { + var procd string + + BeforeEach(func() { + var err error + procd, err = ioutil.TempDir("", "sigarTests") + Expect(err).ToNot(HaveOccurred()) + sigar.Procd = procd + }) + + AfterEach(func() { + sigar.Procd = "/proc" + }) + + Describe("CPU", func() { + var ( + statFile string + cpu sigar.Cpu + ) + + BeforeEach(func() { + statFile = procd + "/stat" + cpu = sigar.Cpu{} + }) + + Describe("Get", func() { + It("gets CPU usage", func() { + statContents := []byte("cpu 25 1 2 3 4 5 6 7") + err := ioutil.WriteFile(statFile, statContents, 0644) + Expect(err).ToNot(HaveOccurred()) + + err = cpu.Get() + Expect(err).ToNot(HaveOccurred()) + Expect(cpu.User).To(Equal(uint64(25))) + }) + + It("ignores empty lines", func() { + statContents := []byte("cpu ") + err := ioutil.WriteFile(statFile, statContents, 0644) + Expect(err).ToNot(HaveOccurred()) + + err = cpu.Get() + Expect(err).ToNot(HaveOccurred()) + Expect(cpu.User).To(Equal(uint64(0))) + }) + }) + + Describe("CollectCpuStats", func() { + It("collects CPU usage over time", func() { + statContents := []byte("cpu 25 1 2 3 4 5 6 7") + err := ioutil.WriteFile(statFile, statContents, 0644) + Expect(err).ToNot(HaveOccurred()) + + concreteSigar := &sigar.ConcreteSigar{} + cpuUsages, stop := concreteSigar.CollectCpuStats(500 * time.Millisecond) + + Expect(<-cpuUsages).To(Equal(sigar.Cpu{ + User: uint64(25), + Nice: uint64(1), + Sys: uint64(2), + Idle: uint64(3), + Wait: uint64(4), + Irq: uint64(5), + SoftIrq: uint64(6), + Stolen: uint64(7), + })) + + statContents = []byte("cpu 30 3 7 10 25 55 36 65") + err = ioutil.WriteFile(statFile, statContents, 0644) + Expect(err).ToNot(HaveOccurred()) + + Expect(<-cpuUsages).To(Equal(sigar.Cpu{ + User: uint64(5), + Nice: uint64(2), + Sys: uint64(5), + Idle: uint64(7), + Wait: uint64(21), + Irq: uint64(50), + SoftIrq: uint64(30), + Stolen: uint64(58), + })) + + stop <- struct{}{} + }) + }) + }) + + Describe("Mem", func() { + var meminfoFile string + BeforeEach(func() { + meminfoFile = procd + "/meminfo" + + meminfoContents := ` +MemTotal: 374256 kB +MemFree: 274460 kB +Buffers: 9764 kB +Cached: 38648 kB +SwapCached: 0 kB +Active: 33772 kB +Inactive: 31184 kB +Active(anon): 16572 kB +Inactive(anon): 552 kB +Active(file): 17200 kB +Inactive(file): 30632 kB +Unevictable: 0 kB +Mlocked: 0 kB +SwapTotal: 786428 kB +SwapFree: 786428 kB +Dirty: 0 kB +Writeback: 0 kB +AnonPages: 16564 kB +Mapped: 6612 kB +Shmem: 584 kB +Slab: 19092 kB +SReclaimable: 9128 kB +SUnreclaim: 9964 kB +KernelStack: 672 kB +PageTables: 1864 kB +NFS_Unstable: 0 kB +Bounce: 0 kB +WritebackTmp: 0 kB +CommitLimit: 973556 kB +Committed_AS: 55880 kB +VmallocTotal: 34359738367 kB +VmallocUsed: 21428 kB +VmallocChunk: 34359713596 kB +HardwareCorrupted: 0 kB +AnonHugePages: 0 kB +HugePages_Total: 0 +HugePages_Free: 0 +HugePages_Rsvd: 0 +HugePages_Surp: 0 +Hugepagesize: 2048 kB +DirectMap4k: 59328 kB +DirectMap2M: 333824 kB +` + err := ioutil.WriteFile(meminfoFile, []byte(meminfoContents), 0444) + Expect(err).ToNot(HaveOccurred()) + }) + + It("returns correct memory info", func() { + mem := sigar.Mem{} + err := mem.Get() + Expect(err).ToNot(HaveOccurred()) + + Expect(mem.Total).To(BeNumerically("==", 374256*1024)) + Expect(mem.Free).To(BeNumerically("==", 274460*1024)) + }) + }) + + Describe("Swap", func() { + var meminfoFile string + BeforeEach(func() { + meminfoFile = procd + "/meminfo" + + meminfoContents := ` +MemTotal: 374256 kB +MemFree: 274460 kB +Buffers: 9764 kB +Cached: 38648 kB +SwapCached: 0 kB +Active: 33772 kB +Inactive: 31184 kB +Active(anon): 16572 kB +Inactive(anon): 552 kB +Active(file): 17200 kB +Inactive(file): 30632 kB +Unevictable: 0 kB +Mlocked: 0 kB +SwapTotal: 786428 kB +SwapFree: 786428 kB +Dirty: 0 kB +Writeback: 0 kB +AnonPages: 16564 kB +Mapped: 6612 kB +Shmem: 584 kB +Slab: 19092 kB +SReclaimable: 9128 kB +SUnreclaim: 9964 kB +KernelStack: 672 kB +PageTables: 1864 kB +NFS_Unstable: 0 kB +Bounce: 0 kB +WritebackTmp: 0 kB +CommitLimit: 973556 kB +Committed_AS: 55880 kB +VmallocTotal: 34359738367 kB +VmallocUsed: 21428 kB +VmallocChunk: 34359713596 kB +HardwareCorrupted: 0 kB +AnonHugePages: 0 kB +HugePages_Total: 0 +HugePages_Free: 0 +HugePages_Rsvd: 0 +HugePages_Surp: 0 +Hugepagesize: 2048 kB +DirectMap4k: 59328 kB +DirectMap2M: 333824 kB +` + err := ioutil.WriteFile(meminfoFile, []byte(meminfoContents), 0444) + Expect(err).ToNot(HaveOccurred()) + }) + + It("returns correct memory info", func() { + swap := sigar.Swap{} + err := swap.Get() + Expect(err).ToNot(HaveOccurred()) + + Expect(swap.Total).To(BeNumerically("==", 786428*1024)) + Expect(swap.Free).To(BeNumerically("==", 786428*1024)) + }) + }) +}) diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_suite_test.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_suite_test.go new file mode 100644 index 000000000..44287f631 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_suite_test.go @@ -0,0 +1,13 @@ +package sigar_test + +import ( + . "github.com/onsi/ginkgo" + . "github.com/onsi/gomega" + + "testing" +) + +func TestGosigar(t *testing.T) { + RegisterFailHandler(Fail) + RunSpecs(t, "Gosigar Suite") +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_unix.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_unix.go new file mode 100644 index 000000000..39f18784b --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_unix.go @@ -0,0 +1,26 @@ +// Copyright (c) 2012 VMware, Inc. + +// +build darwin freebsd linux netbsd openbsd + +package sigar + +import "syscall" + +func (self *FileSystemUsage) Get(path string) error { + stat := syscall.Statfs_t{} + err := syscall.Statfs(path, &stat) + if err != nil { + return err + } + + bsize := stat.Bsize / 512 + + self.Total = (uint64(stat.Blocks) * uint64(bsize)) >> 1 + self.Free = (uint64(stat.Bfree) * uint64(bsize)) >> 1 + self.Avail = (uint64(stat.Bavail) * uint64(bsize)) >> 1 + self.Used = self.Total - self.Free + self.Files = stat.Files + self.FreeFiles = stat.Ffree + + return nil +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_util.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_util.go new file mode 100644 index 000000000..a02df9419 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_util.go @@ -0,0 +1,22 @@ +// Copyright (c) 2012 VMware, Inc. + +package sigar + +import ( + "unsafe" +) + +func bytePtrToString(ptr *int8) string { + bytes := (*[10000]byte)(unsafe.Pointer(ptr)) + + n := 0 + for bytes[n] != 0 { + n++ + } + + return string(bytes[0:n]) +} + +func chop(buf []byte) []byte { + return buf[0 : len(buf)-1] +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_windows.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_windows.go new file mode 100644 index 000000000..0c779d7c3 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_windows.go @@ -0,0 +1,100 @@ +// Copyright (c) 2012 VMware, Inc. + +package sigar + +// #include +// #include +import "C" + +import ( + "fmt" + "unsafe" +) + +func init() { +} + +func (self *LoadAverage) Get() error { + return nil +} + +func (self *Uptime) Get() error { + return nil +} + +func (self *Mem) Get() error { + var statex C.MEMORYSTATUSEX + statex.dwLength = C.DWORD(unsafe.Sizeof(statex)) + + succeeded := C.GlobalMemoryStatusEx(&statex) + if succeeded == C.FALSE { + lastError := C.GetLastError() + return fmt.Errorf("GlobalMemoryStatusEx failed with error: %d", int(lastError)) + } + + self.Total = uint64(statex.ullTotalPhys) + return nil +} + +func (self *Swap) Get() error { + return notImplemented() +} + +func (self *Cpu) Get() error { + return notImplemented() +} + +func (self *CpuList) Get() error { + return notImplemented() +} + +func (self *FileSystemList) Get() error { + return notImplemented() +} + +func (self *ProcList) Get() error { + return notImplemented() +} + +func (self *ProcState) Get(pid int) error { + return notImplemented() +} + +func (self *ProcMem) Get(pid int) error { + return notImplemented() +} + +func (self *ProcTime) Get(pid int) error { + return notImplemented() +} + +func (self *ProcArgs) Get(pid int) error { + return notImplemented() +} + +func (self *ProcExe) Get(pid int) error { + return notImplemented() +} + +func (self *FileSystemUsage) Get(path string) error { + var availableBytes C.ULARGE_INTEGER + var totalBytes C.ULARGE_INTEGER + var totalFreeBytes C.ULARGE_INTEGER + + pathChars := C.CString(path) + defer C.free(unsafe.Pointer(pathChars)) + + succeeded := C.GetDiskFreeSpaceEx((*C.CHAR)(pathChars), &availableBytes, &totalBytes, &totalFreeBytes) + if succeeded == C.FALSE { + lastError := C.GetLastError() + return fmt.Errorf("GetDiskFreeSpaceEx failed with error: %d", int(lastError)) + } + + self.Total = *(*uint64)(unsafe.Pointer(&totalBytes)) + return nil +} + +func notImplemented() error { + panic("Not Implemented") + return nil +} diff --git a/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_windows_test.go b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_windows_test.go new file mode 100644 index 000000000..868bdaab8 --- /dev/null +++ b/Godeps/_workspace/src/github.com/cloudfoundry/gosigar/sigar_windows_test.go @@ -0,0 +1,32 @@ +package sigar_test + +import ( + "os" + + . "github.com/onsi/ginkgo" + . "github.com/onsi/gomega" + + sigar "github.com/cloudfoundry/gosigar" +) + +var _ = Describe("SigarWindows", func() { + Describe("Memory", func() { + It("gets the total memory", func() { + mem := sigar.Mem{} + err := mem.Get() + + Ω(err).ShouldNot(HaveOccurred()) + Ω(mem.Total).Should(BeNumerically(">", 0)) + }) + }) + + Describe("Disk", func() { + It("gets the total disk space", func() { + usage := sigar.FileSystemUsage{} + err := usage.Get(os.TempDir()) + + Ω(err).ShouldNot(HaveOccurred()) + Ω(usage.Total).Should(BeNumerically(">", 0)) + }) + }) +}) diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/.gitignore b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/.gitignore new file mode 100644 index 000000000..5f6b48eae --- /dev/null +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/.gitignore @@ -0,0 +1,2 @@ +# temporary symlink for testing +testing/data/symlink diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/.travis.yml b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/.travis.yml index 3926838ac..927a999f0 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/.travis.yml +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/.travis.yml @@ -1,8 +1,9 @@ language: go sudo: false go: - - 1.3.1 - - 1.4 + - 1.3.3 + - 1.4.2 + - 1.5.1 - tip env: - GOARCH=amd64 diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/AUTHORS b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/AUTHORS index 2febb1f03..418a808fb 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/AUTHORS +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/AUTHORS @@ -11,8 +11,10 @@ Ben McCann Brendan Fosberry Brian Lalor Brian Palmer +Bryan Boreham Burke Libbey Carlos Diaz-Padron +Cesar Wong Cezar Sa Espinola Cheah Chu Yeow cheneydeng @@ -27,12 +29,14 @@ David Huie Dawn Chen Dinesh Subhraveti Ed +Erez Horev Eric Anderson Ewout Prangsma Fabio Rehm Fatih Arslan Flavia Missi Francisco Souza +Grégoire Delattre Guillermo Álvarez Fernández He Simei Ivan Mikushin @@ -43,6 +47,7 @@ Jawher Moussa Jean-Baptiste Dalido Jeff Mitchell Jeffrey Hulten +Jen Andre Johan Euphrosine Kamil Domanski Karan Misra @@ -50,6 +55,7 @@ Kim, Hirokuni Kyle Allan Liron Levin Liu Peng +Lorenz Leutgeb Lucas Clemente Lucas Weiblen Mantas Matelis @@ -66,12 +72,14 @@ Paul Morie Paul Weil Peter Edge Peter Jihoon Kim +Phil Lu Philippe Lafoucrière Rafe Colton Rob Miller Robert Williamson Salvador Gironès Sam Rijs +Samuel Karp Simon Eskildsen Simon Menke Skolos diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth.go index fccd55740..30e3af3eb 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth.go @@ -16,7 +16,8 @@ import ( "strings" ) -var AuthParseError error = errors.New("Failed to read authentication from dockercfg") +// ErrCannotParseDockercfg is the error returned by NewAuthConfigurations when the dockercfg cannot be parsed. +var ErrCannotParseDockercfg = errors.New("Failed to read authentication from dockercfg") // AuthConfiguration represents authentication options to use in the PushImage // method. It represents the authentication in the Docker index server. @@ -33,6 +34,10 @@ type AuthConfigurations struct { Configs map[string]AuthConfiguration `json:"configs"` } +// AuthConfigurations119 is used to serialize a set of AuthConfigurations +// for Docker API >= 1.19. +type AuthConfigurations119 map[string]AuthConfiguration + // dockerConfig represents a registry authentation configuration from the // .dockercfg file. type dockerConfig struct { @@ -103,7 +108,7 @@ func authConfigs(confs map[string]dockerConfig) (*AuthConfigurations, error) { } userpass := strings.Split(string(data), ":") if len(userpass) != 2 { - return nil, AuthParseError + return nil, ErrCannotParseDockercfg } c.Configs[reg] = AuthConfiguration{ Email: conf.Email, @@ -117,17 +122,15 @@ func authConfigs(confs map[string]dockerConfig) (*AuthConfigurations, error) { // AuthCheck validates the given credentials. It returns nil if successful. // -// See https://goo.gl/vPoEfJ for more details. +// See https://goo.gl/m2SleN for more details. func (c *Client) AuthCheck(conf *AuthConfiguration) error { if conf == nil { return fmt.Errorf("conf is nil") } - body, statusCode, err := c.do("POST", "/auth", doOptions{data: conf}) + resp, err := c.do("POST", "/auth", doOptions{data: conf}) if err != nil { return err } - if statusCode > 400 { - return fmt.Errorf("auth error (%d): %s", statusCode, body) - } + resp.Body.Close() return nil } diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth_test.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth_test.go index fc0ffab84..b3d4f8fc8 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth_test.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/auth_test.go @@ -41,7 +41,7 @@ func TestAuthBadConfig(t *testing.T) { auth := base64.StdEncoding.EncodeToString([]byte("userpass")) read := strings.NewReader(fmt.Sprintf(`{"docker.io":{"auth":"%s","email":"user@example.com"}}`, auth)) ac, err := NewAuthConfigurations(read) - if err != AuthParseError { + if err != ErrCannotParseDockercfg { t.Errorf("Incorrect error returned %v\n", err) } if ac != nil { diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/build_test.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/build_test.go index a4864db83..c9640f205 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/build_test.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/build_test.go @@ -35,6 +35,16 @@ func TestBuildImageMultipleContextsError(t *testing.T) { func TestBuildImageContextDirDockerignoreParsing(t *testing.T) { fakeRT := &FakeRoundTripper{message: "", status: http.StatusOK} client := newTestClient(fakeRT) + + if err := os.Symlink("doesnotexist", "testing/data/symlink"); err != nil { + t.Errorf("error creating symlink on demand: %s", err) + } + defer func() { + if err := os.Remove("testing/data/symlink"); err != nil { + t.Errorf("error removing symlink on demand: %s", err) + } + }() + var buf bytes.Buffer opts := BuildImageOptions{ Name: "testImage", diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/change.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/change.go index e7b056c3f..d133594d4 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/change.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/change.go @@ -23,7 +23,7 @@ const ( // Change represents a change in a container. // -// See http://goo.gl/QkW9sH for more details. +// See https://goo.gl/9GsTIF for more details. type Change struct { Path string Kind ChangeType diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client.go index 986bbb3d2..3ba44e864 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client.go @@ -4,7 +4,7 @@ // Package docker provides a client for the Docker remote API. // -// See http://goo.gl/G3plxW for more details on the remote API. +// See https://goo.gl/G3plxW for more details on the remote API. package docker import ( @@ -45,6 +45,8 @@ var ( ErrConnectionRefused = errors.New("cannot connect to Docker endpoint") apiVersion112, _ = NewAPIVersion("1.12") + + apiVersion119, _ = NewAPIVersion("1.19") ) // APIVersion is an internal representation of a version of the Remote API. @@ -128,6 +130,7 @@ type Client struct { SkipServerVersionCheck bool HTTPClient *http.Client TLSConfig *tls.Config + Dialer *net.Dialer endpoint string endpointURL *url.URL @@ -135,6 +138,7 @@ type Client struct { requestedAPIVersion APIVersion serverAPIVersion APIVersion expectedAPIVersion APIVersion + unixHTTPClient *http.Client } // NewClient returns a Client instance ready for communication with the given @@ -189,6 +193,7 @@ func NewVersionedClient(endpoint string, apiVersionString string) (*Client, erro } return &Client{ HTTPClient: http.DefaultClient, + Dialer: &net.Dialer{}, endpoint: endpoint, endpointURL: u, eventMonitor: new(eventMonitoringState), @@ -300,6 +305,7 @@ func NewVersionedTLSClientFromBytes(endpoint string, certPEMBlock, keyPEMBlock, return &Client{ HTTPClient: &http.Client{Transport: tr}, TLSConfig: tlsConfig, + Dialer: &net.Dialer{}, endpoint: endpoint, endpointURL: u, eventMonitor: new(eventMonitoringState), @@ -324,32 +330,40 @@ func (c *Client) checkAPIVersion() error { return nil } +// Endpoint returns the current endpoint. It's useful for getting the endpoint +// when using functions that get this data from the environment (like +// NewClientFromEnv. +func (c *Client) Endpoint() string { + return c.endpoint +} + // Ping pings the docker server // -// See http://goo.gl/stJENm for more details. +// See https://goo.gl/kQCfJj for more details. func (c *Client) Ping() error { path := "/_ping" - body, status, err := c.do("GET", path, doOptions{}) + resp, err := c.do("GET", path, doOptions{}) if err != nil { return err } - if status != http.StatusOK { - return newError(status, body) + if resp.StatusCode != http.StatusOK { + return newError(resp) } + resp.Body.Close() return nil } func (c *Client) getServerAPIVersionString() (version string, err error) { - body, status, err := c.do("GET", "/version", doOptions{}) + resp, err := c.do("GET", "/version", doOptions{}) if err != nil { return "", err } - if status != http.StatusOK { - return "", fmt.Errorf("Received unexpected status %d while trying to retrieve the server version", status) + defer resp.Body.Close() + if resp.StatusCode != http.StatusOK { + return "", fmt.Errorf("Received unexpected status %d while trying to retrieve the server version", resp.StatusCode) } var versionResponse map[string]interface{} - err = json.Unmarshal(body, &versionResponse) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&versionResponse); err != nil { return "", err } if version, ok := (versionResponse["ApiVersion"]).(string); ok { @@ -363,24 +377,35 @@ type doOptions struct { forceJSON bool } -func (c *Client) do(method, path string, doOptions doOptions) ([]byte, int, error) { +func (c *Client) do(method, path string, doOptions doOptions) (*http.Response, error) { var params io.Reader if doOptions.data != nil || doOptions.forceJSON { buf, err := json.Marshal(doOptions.data) if err != nil { - return nil, -1, err + return nil, err } params = bytes.NewBuffer(buf) } if path != "/version" && !c.SkipServerVersionCheck && c.expectedAPIVersion == nil { err := c.checkAPIVersion() if err != nil { - return nil, -1, err + return nil, err } } - req, err := http.NewRequest(method, c.getURL(path), params) + + httpClient := c.HTTPClient + protocol := c.endpointURL.Scheme + var u string + if protocol == "unix" { + httpClient = c.unixClient() + u = c.getFakeUnixURL(path) + } else { + u = c.getURL(path) + } + + req, err := http.NewRequest(method, u, params) if err != nil { - return nil, -1, err + return nil, err } req.Header.Set("User-Agent", userAgent) if doOptions.data != nil { @@ -388,40 +413,19 @@ func (c *Client) do(method, path string, doOptions doOptions) ([]byte, int, erro } else if method == "POST" { req.Header.Set("Content-Type", "plain/text") } - var resp *http.Response - protocol := c.endpointURL.Scheme - address := c.endpointURL.Path - if protocol == "unix" { - var dial net.Conn - dial, err = net.Dial(protocol, address) - if err != nil { - return nil, -1, err - } - defer dial.Close() - breader := bufio.NewReader(dial) - err = req.Write(dial) - if err != nil { - return nil, -1, err - } - resp, err = http.ReadResponse(breader, req) - } else { - resp, err = c.HTTPClient.Do(req) - } + + resp, err := httpClient.Do(req) if err != nil { if strings.Contains(err.Error(), "connection refused") { - return nil, -1, ErrConnectionRefused + return nil, ErrConnectionRefused } - return nil, -1, err - } - defer resp.Body.Close() - body, err := ioutil.ReadAll(resp.Body) - if err != nil { - return nil, -1, err + return nil, err } + if resp.StatusCode < 200 || resp.StatusCode >= 400 { - return nil, resp.StatusCode, newError(resp.StatusCode, body) + return nil, newError(resp) } - return body, resp.StatusCode, nil + return resp, nil } type streamOptions struct { @@ -462,12 +466,16 @@ func (c *Client) stream(method, path string, streamOptions streamOptions) error address := c.endpointURL.Path if streamOptions.stdout == nil { streamOptions.stdout = ioutil.Discard + } else if t, ok := streamOptions.stdout.(io.Closer); ok { + defer t.Close() } if streamOptions.stderr == nil { streamOptions.stderr = ioutil.Discard + } else if t, ok := streamOptions.stderr.(io.Closer); ok { + defer t.Close() } if protocol == "unix" { - dial, err := net.Dial(protocol, address) + dial, err := c.Dialer.Dial(protocol, address) if err != nil { return err } @@ -503,11 +511,7 @@ func (c *Client) stream(method, path string, streamOptions streamOptions) error } defer resp.Body.Close() if resp.StatusCode < 200 || resp.StatusCode >= 400 { - body, err := ioutil.ReadAll(resp.Body) - if err != nil { - return err - } - return newError(resp.StatusCode, body) + return newError(resp) } if streamOptions.useJSONDecoder || resp.Header.Get("Content-Type") == "application/json" { // if we want to get raw json stream, just copy it back to output @@ -583,6 +587,8 @@ func (c *Client) hijack(method, path string, hijackOptions hijackOptions) error return err } req.Header.Set("Content-Type", "plain/text") + req.Header.Set("Connection", "Upgrade") + req.Header.Set("Upgrade", "tcp") protocol := c.endpointURL.Scheme address := c.endpointURL.Path if protocol != "unix" { @@ -591,12 +597,12 @@ func (c *Client) hijack(method, path string, hijackOptions hijackOptions) error } var dial net.Conn if c.TLSConfig != nil && protocol != "unix" { - dial, err = tlsDial(protocol, address, c.TLSConfig) + dial, err = tlsDialWithDialer(c.Dialer, protocol, address, c.TLSConfig) if err != nil { return err } } else { - dial, err = net.Dial(protocol, address) + dial, err = c.Dialer.Dial(protocol, address) if err != nil { return err } @@ -612,13 +618,16 @@ func (c *Client) hijack(method, path string, hijackOptions hijackOptions) error defer rwc.Close() errChanOut := make(chan error, 1) errChanIn := make(chan error, 1) - exit := make(chan bool) go func() { - defer close(exit) - defer close(errChanOut) + defer func() { + if hijackOptions.in != nil { + if closer, ok := hijackOptions.in.(io.Closer); ok { + closer.Close() + } + } + }() var err error if hijackOptions.setRawTerminal { - // When TTY is ON, use regular copy _, err = io.Copy(hijackOptions.stdout, br) } else { _, err = stdcopy.StdCopy(hijackOptions.stdout, hijackOptions.stderr, br) @@ -626,17 +635,15 @@ func (c *Client) hijack(method, path string, hijackOptions hijackOptions) error errChanOut <- err }() go func() { + var err error if hijackOptions.in != nil { - _, err := io.Copy(rwc, hijackOptions.in) - errChanIn <- err - } else { - errChanIn <- nil + _, err = io.Copy(rwc, hijackOptions.in) } + errChanIn <- err rwc.(interface { CloseWrite() error }).CloseWrite() }() - <-exit errIn := <-errChanIn errOut := <-errChanOut if errIn != nil { @@ -657,6 +664,41 @@ func (c *Client) getURL(path string) string { return fmt.Sprintf("%s%s", urlStr, path) } +// getFakeUnixURL returns the URL needed to make an HTTP request over a UNIX +// domain socket to the given path. +func (c *Client) getFakeUnixURL(path string) string { + u := *c.endpointURL // Copy. + + // Override URL so that net/http will not complain. + u.Scheme = "http" + u.Host = "unix.sock" // Doesn't matter what this is - it's not used. + u.Path = "" + + urlStr := strings.TrimRight(u.String(), "/") + + if c.requestedAPIVersion != nil { + return fmt.Sprintf("%s/v%s%s", urlStr, c.requestedAPIVersion, path) + } + return fmt.Sprintf("%s%s", urlStr, path) +} + +func (c *Client) unixClient() *http.Client { + if c.unixHTTPClient != nil { + return c.unixHTTPClient + } + + socketPath := c.endpointURL.Path + c.unixHTTPClient = &http.Client{ + Transport: &http.Transport{ + Dial: func(network, addr string) (net.Conn, error) { + return c.Dialer.Dial("unix", socketPath) + }, + }, + } + + return c.unixHTTPClient +} + type jsonMessage struct { Status string `json:"status,omitempty"` Progress string `json:"progress,omitempty"` @@ -738,8 +780,13 @@ type Error struct { Message string } -func newError(status int, body []byte) *Error { - return &Error{Status: status, Message: string(body)} +func newError(resp *http.Response) *Error { + defer resp.Body.Close() + data, err := ioutil.ReadAll(resp.Body) + if err != nil { + return &Error{Status: resp.StatusCode, Message: fmt.Sprintf("cannot read body, err: %v", err)} + } + return &Error{Status: resp.StatusCode, Message: string(data)} } func (e *Error) Error() string { diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client_test.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client_test.go index c00c3d30c..67230a4e1 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client_test.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/client_test.go @@ -5,6 +5,7 @@ package docker import ( + "bytes" "fmt" "io/ioutil" "net" @@ -161,6 +162,16 @@ func TestNewTLSClient(t *testing.T) { } } +func TestEndpoint(t *testing.T) { + client, err := NewVersionedClient("http://localhost:4243", "1.12") + if err != nil { + t.Fatal(err) + } + if endpoint := client.Endpoint(); endpoint != client.endpoint { + t.Errorf("Client.Endpoint(): want %q. Got %q", client.endpoint, endpoint) + } +} + func TestGetURL(t *testing.T) { var tests = []struct { endpoint string @@ -185,8 +196,34 @@ func TestGetURL(t *testing.T) { } } +func TestGetFakeUnixURL(t *testing.T) { + var tests = []struct { + endpoint string + path string + expected string + }{ + {"unix://var/run/docker.sock", "/", "http://unix.sock/"}, + {"unix://var/run/docker.socket", "/", "http://unix.sock/"}, + {"unix://var/run/docker.sock", "/containers/ps", "http://unix.sock/containers/ps"}, + } + for _, tt := range tests { + client, _ := NewClient(tt.endpoint) + client.endpoint = tt.endpoint + client.SkipServerVersionCheck = true + got := client.getFakeUnixURL(tt.path) + if got != tt.expected { + t.Errorf("getURL(%q): Got %s. Want %s.", tt.path, got, tt.expected) + } + } +} + func TestError(t *testing.T) { - err := newError(400, []byte("bad parameter")) + fakeBody := ioutil.NopCloser(bytes.NewBufferString("bad parameter")) + resp := &http.Response{ + StatusCode: 400, + Body: fakeBody, + } + err := newError(resp) expected := Error{Status: 400, Message: "bad parameter"} if !reflect.DeepEqual(expected, *err) { t.Errorf("Wrong error type. Want %#v. Got %#v.", expected, *err) @@ -334,7 +371,7 @@ func TestPingErrorWithUnixSocket(t *testing.T) { } defer li.Close() if err != nil { - t.Fatalf("Expected to get listner, but failed: %#v", err) + t.Fatalf("Expected to get listener, but failed: %#v", err) } fd, err := li.Accept() @@ -345,7 +382,7 @@ func TestPingErrorWithUnixSocket(t *testing.T) { buf := make([]byte, 512) nr, err := fd.Read(buf) - // Create invalid response message to occur error + // Create invalid response message to trigger error. data := buf[0:nr] for i := 0; i < 10; i++ { data[i] = 63 @@ -366,6 +403,7 @@ func TestPingErrorWithUnixSocket(t *testing.T) { u, _ := parseEndpoint(endpoint, false) client := Client{ HTTPClient: http.DefaultClient, + Dialer: &net.Dialer{}, endpoint: endpoint, endpointURL: u, SkipServerVersionCheck: true, diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container.go index 89430975b..497cfd00c 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container.go @@ -5,7 +5,6 @@ package docker import ( - "bytes" "encoding/json" "errors" "fmt" @@ -23,7 +22,7 @@ var ErrContainerAlreadyExists = errors.New("container already exists") // ListContainersOptions specify parameters to the ListContainers function. // -// See http://goo.gl/6Y4Gz7 for more details. +// See https://goo.gl/47a6tO for more details. type ListContainersOptions struct { All bool Size bool @@ -41,33 +40,33 @@ type APIPort struct { IP string `json:"IP,omitempty" yaml:"IP,omitempty"` } -// APIContainers represents a container. -// -// See http://goo.gl/QeFH7U for more details. +// APIContainers represents each container in the list returned by +// ListContainers. type APIContainers struct { - ID string `json:"Id" yaml:"Id"` - Image string `json:"Image,omitempty" yaml:"Image,omitempty"` - Command string `json:"Command,omitempty" yaml:"Command,omitempty"` - Created int64 `json:"Created,omitempty" yaml:"Created,omitempty"` - Status string `json:"Status,omitempty" yaml:"Status,omitempty"` - Ports []APIPort `json:"Ports,omitempty" yaml:"Ports,omitempty"` - SizeRw int64 `json:"SizeRw,omitempty" yaml:"SizeRw,omitempty"` - SizeRootFs int64 `json:"SizeRootFs,omitempty" yaml:"SizeRootFs,omitempty"` - Names []string `json:"Names,omitempty" yaml:"Names,omitempty"` + ID string `json:"Id" yaml:"Id"` + Image string `json:"Image,omitempty" yaml:"Image,omitempty"` + Command string `json:"Command,omitempty" yaml:"Command,omitempty"` + Created int64 `json:"Created,omitempty" yaml:"Created,omitempty"` + Status string `json:"Status,omitempty" yaml:"Status,omitempty"` + Ports []APIPort `json:"Ports,omitempty" yaml:"Ports,omitempty"` + SizeRw int64 `json:"SizeRw,omitempty" yaml:"SizeRw,omitempty"` + SizeRootFs int64 `json:"SizeRootFs,omitempty" yaml:"SizeRootFs,omitempty"` + Names []string `json:"Names,omitempty" yaml:"Names,omitempty"` + Labels map[string]string `json:"Labels,omitempty" yaml:"Labels, omitempty"` } // ListContainers returns a slice of containers matching the given criteria. // -// See http://goo.gl/6Y4Gz7 for more details. +// See https://goo.gl/47a6tO for more details. func (c *Client) ListContainers(opts ListContainersOptions) ([]APIContainers, error) { path := "/containers/json?" + queryString(opts) - body, _, err := c.do("GET", path, doOptions{}) + resp, err := c.do("GET", path, doOptions{}) if err != nil { return nil, err } + defer resp.Body.Close() var containers []APIContainers - err = json.Unmarshal(body, &containers) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&containers); err != nil { return nil, err } return containers, nil @@ -206,6 +205,7 @@ type Config struct { DNS []string `json:"Dns,omitempty" yaml:"Dns,omitempty"` // For Docker API v1.9 and below only Image string `json:"Image,omitempty" yaml:"Image,omitempty"` Volumes map[string]struct{} `json:"Volumes,omitempty" yaml:"Volumes,omitempty"` + VolumeDriver string `json:"VolumeDriver,omitempty" yaml:"VolumeDriver,omitempty"` VolumesFrom string `json:"VolumesFrom,omitempty" yaml:"VolumesFrom,omitempty"` WorkingDir string `json:"WorkingDir,omitempty" yaml:"WorkingDir,omitempty"` MacAddress string `json:"MacAddress,omitempty" yaml:"MacAddress,omitempty"` @@ -213,9 +213,21 @@ type Config struct { NetworkDisabled bool `json:"NetworkDisabled,omitempty" yaml:"NetworkDisabled,omitempty"` SecurityOpts []string `json:"SecurityOpts,omitempty" yaml:"SecurityOpts,omitempty"` OnBuild []string `json:"OnBuild,omitempty" yaml:"OnBuild,omitempty"` + Mounts []Mount `json:"Mounts,omitempty" yaml:"Mounts,omitempty"` Labels map[string]string `json:"Labels,omitempty" yaml:"Labels,omitempty"` } +// Mount represents a mount point in the container. +// +// It has been added in the version 1.20 of the Docker API, available since +// Docker 1.8. +type Mount struct { + Source string + Destination string + Mode string + RW bool +} + // LogConfig defines the log driver type and the configuration for it. type LogConfig struct { Type string `json:"Type,omitempty" yaml:"Type,omitempty"` @@ -259,13 +271,14 @@ type Container struct { NetworkSettings *NetworkSettings `json:"NetworkSettings,omitempty" yaml:"NetworkSettings,omitempty"` - SysInitPath string `json:"SysInitPath,omitempty" yaml:"SysInitPath,omitempty"` - ResolvConfPath string `json:"ResolvConfPath,omitempty" yaml:"ResolvConfPath,omitempty"` - HostnamePath string `json:"HostnamePath,omitempty" yaml:"HostnamePath,omitempty"` - HostsPath string `json:"HostsPath,omitempty" yaml:"HostsPath,omitempty"` - LogPath string `json:"LogPath,omitempty" yaml:"LogPath,omitempty"` - Name string `json:"Name,omitempty" yaml:"Name,omitempty"` - Driver string `json:"Driver,omitempty" yaml:"Driver,omitempty"` + SysInitPath string `json:"SysInitPath,omitempty" yaml:"SysInitPath,omitempty"` + ResolvConfPath string `json:"ResolvConfPath,omitempty" yaml:"ResolvConfPath,omitempty"` + HostnamePath string `json:"HostnamePath,omitempty" yaml:"HostnamePath,omitempty"` + HostsPath string `json:"HostsPath,omitempty" yaml:"HostsPath,omitempty"` + LogPath string `json:"LogPath,omitempty" yaml:"LogPath,omitempty"` + Name string `json:"Name,omitempty" yaml:"Name,omitempty"` + Driver string `json:"Driver,omitempty" yaml:"Driver,omitempty"` + Mounts []Mount `json:"Mounts,omitempty" yaml:"Mounts,omitempty"` Volumes map[string]string `json:"Volumes,omitempty" yaml:"Volumes,omitempty"` VolumesRW map[string]bool `json:"VolumesRW,omitempty" yaml:"VolumesRW,omitempty"` @@ -279,7 +292,7 @@ type Container struct { // RenameContainerOptions specify parameters to the RenameContainer function. // -// See http://goo.gl/L00hoj for more details. +// See https://goo.gl/laSOIy for more details. type RenameContainerOptions struct { // ID of container to rename ID string `qs:"-"` @@ -290,27 +303,31 @@ type RenameContainerOptions struct { // RenameContainer updates and existing containers name // -// See http://goo.gl/L00hoj for more details. +// See https://goo.gl/laSOIy for more details. func (c *Client) RenameContainer(opts RenameContainerOptions) error { - _, _, err := c.do("POST", fmt.Sprintf("/containers/"+opts.ID+"/rename?%s", queryString(opts)), doOptions{}) - return err + resp, err := c.do("POST", fmt.Sprintf("/containers/"+opts.ID+"/rename?%s", queryString(opts)), doOptions{}) + if err != nil { + return err + } + resp.Body.Close() + return nil } // InspectContainer returns information about a container by its ID. // -// See http://goo.gl/CxVuJ5 for more details. +// See https://goo.gl/RdIq0b for more details. func (c *Client) InspectContainer(id string) (*Container, error) { path := "/containers/" + id + "/json" - body, status, err := c.do("GET", path, doOptions{}) - if status == http.StatusNotFound { - return nil, &NoSuchContainer{ID: id} - } + resp, err := c.do("GET", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, &NoSuchContainer{ID: id} + } return nil, err } + defer resp.Body.Close() var container Container - err = json.Unmarshal(body, &container) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&container); err != nil { return nil, err } return &container, nil @@ -318,19 +335,19 @@ func (c *Client) InspectContainer(id string) (*Container, error) { // ContainerChanges returns changes in the filesystem of the given container. // -// See http://goo.gl/QkW9sH for more details. +// See https://goo.gl/9GsTIF for more details. func (c *Client) ContainerChanges(id string) ([]Change, error) { path := "/containers/" + id + "/changes" - body, status, err := c.do("GET", path, doOptions{}) - if status == http.StatusNotFound { - return nil, &NoSuchContainer{ID: id} - } + resp, err := c.do("GET", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, &NoSuchContainer{ID: id} + } return nil, err } + defer resp.Body.Close() var changes []Change - err = json.Unmarshal(body, &changes) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&changes); err != nil { return nil, err } return changes, nil @@ -338,7 +355,7 @@ func (c *Client) ContainerChanges(id string) ([]Change, error) { // CreateContainerOptions specify parameters to the CreateContainer function. // -// See http://goo.gl/2xxQQK for more details. +// See https://goo.gl/WxQzrr for more details. type CreateContainerOptions struct { Name string Config *Config `qs:"-"` @@ -348,10 +365,10 @@ type CreateContainerOptions struct { // CreateContainer creates a new container, returning the container instance, // or an error in case of failure. // -// See http://goo.gl/mErxNp for more details. +// See https://goo.gl/WxQzrr for more details. func (c *Client) CreateContainer(opts CreateContainerOptions) (*Container, error) { path := "/containers/create?" + queryString(opts) - body, status, err := c.do( + resp, err := c.do( "POST", path, doOptions{ @@ -365,18 +382,21 @@ func (c *Client) CreateContainer(opts CreateContainerOptions) (*Container, error }, ) - if status == http.StatusNotFound { - return nil, ErrNoSuchImage - } - if status == http.StatusConflict { - return nil, ErrContainerAlreadyExists + if e, ok := err.(*Error); ok { + if e.Status == http.StatusNotFound { + return nil, ErrNoSuchImage + } + if e.Status == http.StatusConflict { + return nil, ErrContainerAlreadyExists + } } + if err != nil { return nil, err } + defer resp.Body.Close() var container Container - err = json.Unmarshal(body, &container) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&container); err != nil { return nil, err } @@ -434,125 +454,135 @@ type Device struct { // HostConfig contains the container options related to starting a container on // a given host type HostConfig struct { - Binds []string `json:"Binds,omitempty" yaml:"Binds,omitempty"` - CapAdd []string `json:"CapAdd,omitempty" yaml:"CapAdd,omitempty"` - CapDrop []string `json:"CapDrop,omitempty" yaml:"CapDrop,omitempty"` - ContainerIDFile string `json:"ContainerIDFile,omitempty" yaml:"ContainerIDFile,omitempty"` - LxcConf []KeyValuePair `json:"LxcConf,omitempty" yaml:"LxcConf,omitempty"` - Privileged bool `json:"Privileged,omitempty" yaml:"Privileged,omitempty"` - PortBindings map[Port][]PortBinding `json:"PortBindings,omitempty" yaml:"PortBindings,omitempty"` - Links []string `json:"Links,omitempty" yaml:"Links,omitempty"` - PublishAllPorts bool `json:"PublishAllPorts,omitempty" yaml:"PublishAllPorts,omitempty"` - DNS []string `json:"Dns,omitempty" yaml:"Dns,omitempty"` // For Docker API v1.10 and above only - DNSSearch []string `json:"DnsSearch,omitempty" yaml:"DnsSearch,omitempty"` - ExtraHosts []string `json:"ExtraHosts,omitempty" yaml:"ExtraHosts,omitempty"` - VolumesFrom []string `json:"VolumesFrom,omitempty" yaml:"VolumesFrom,omitempty"` - NetworkMode string `json:"NetworkMode,omitempty" yaml:"NetworkMode,omitempty"` - IpcMode string `json:"IpcMode,omitempty" yaml:"IpcMode,omitempty"` - PidMode string `json:"PidMode,omitempty" yaml:"PidMode,omitempty"` - UTSMode string `json:"UTSMode,omitempty" yaml:"UTSMode,omitempty"` - RestartPolicy RestartPolicy `json:"RestartPolicy,omitempty" yaml:"RestartPolicy,omitempty"` - Devices []Device `json:"Devices,omitempty" yaml:"Devices,omitempty"` - LogConfig LogConfig `json:"LogConfig,omitempty" yaml:"LogConfig,omitempty"` - ReadonlyRootfs bool `json:"ReadonlyRootfs,omitempty" yaml:"ReadonlyRootfs,omitempty"` - SecurityOpt []string `json:"SecurityOpt,omitempty" yaml:"SecurityOpt,omitempty"` - CgroupParent string `json:"CgroupParent,omitempty" yaml:"CgroupParent,omitempty"` - Memory int64 `json:"Memory,omitempty" yaml:"Memory,omitempty"` - MemorySwap int64 `json:"MemorySwap,omitempty" yaml:"MemorySwap,omitempty"` - CPUShares int64 `json:"CpuShares,omitempty" yaml:"CpuShares,omitempty"` - CPUSet string `json:"Cpuset,omitempty" yaml:"Cpuset,omitempty"` - CPUQuota int64 `json:"CpuQuota,omitempty" yaml:"CpuQuota,omitempty"` - CPUPeriod int64 `json:"CpuPeriod,omitempty" yaml:"CpuPeriod,omitempty"` - Ulimits []ULimit `json:"Ulimits,omitempty" yaml:"Ulimits,omitempty"` + Binds []string `json:"Binds,omitempty" yaml:"Binds,omitempty"` + CapAdd []string `json:"CapAdd,omitempty" yaml:"CapAdd,omitempty"` + CapDrop []string `json:"CapDrop,omitempty" yaml:"CapDrop,omitempty"` + ContainerIDFile string `json:"ContainerIDFile,omitempty" yaml:"ContainerIDFile,omitempty"` + LxcConf []KeyValuePair `json:"LxcConf,omitempty" yaml:"LxcConf,omitempty"` + Privileged bool `json:"Privileged,omitempty" yaml:"Privileged,omitempty"` + PortBindings map[Port][]PortBinding `json:"PortBindings,omitempty" yaml:"PortBindings,omitempty"` + Links []string `json:"Links,omitempty" yaml:"Links,omitempty"` + PublishAllPorts bool `json:"PublishAllPorts,omitempty" yaml:"PublishAllPorts,omitempty"` + DNS []string `json:"Dns,omitempty" yaml:"Dns,omitempty"` // For Docker API v1.10 and above only + DNSSearch []string `json:"DnsSearch,omitempty" yaml:"DnsSearch,omitempty"` + ExtraHosts []string `json:"ExtraHosts,omitempty" yaml:"ExtraHosts,omitempty"` + VolumesFrom []string `json:"VolumesFrom,omitempty" yaml:"VolumesFrom,omitempty"` + NetworkMode string `json:"NetworkMode,omitempty" yaml:"NetworkMode,omitempty"` + IpcMode string `json:"IpcMode,omitempty" yaml:"IpcMode,omitempty"` + PidMode string `json:"PidMode,omitempty" yaml:"PidMode,omitempty"` + UTSMode string `json:"UTSMode,omitempty" yaml:"UTSMode,omitempty"` + RestartPolicy RestartPolicy `json:"RestartPolicy,omitempty" yaml:"RestartPolicy,omitempty"` + Devices []Device `json:"Devices,omitempty" yaml:"Devices,omitempty"` + LogConfig LogConfig `json:"LogConfig,omitempty" yaml:"LogConfig,omitempty"` + ReadonlyRootfs bool `json:"ReadonlyRootfs,omitempty" yaml:"ReadonlyRootfs,omitempty"` + SecurityOpt []string `json:"SecurityOpt,omitempty" yaml:"SecurityOpt,omitempty"` + CgroupParent string `json:"CgroupParent,omitempty" yaml:"CgroupParent,omitempty"` + Memory int64 `json:"Memory,omitempty" yaml:"Memory,omitempty"` + MemorySwap int64 `json:"MemorySwap,omitempty" yaml:"MemorySwap,omitempty"` + MemorySwappiness int64 `json:"MemorySwappiness,omitempty" yaml:"MemorySwappiness,omitempty"` + OOMKillDisable bool `json:"OomKillDisable,omitempty" yaml:"OomKillDisable"` + CPUShares int64 `json:"CpuShares,omitempty" yaml:"CpuShares,omitempty"` + CPUSet string `json:"Cpuset,omitempty" yaml:"Cpuset,omitempty"` + CPUSetCPUs string `json:"CpusetCpus,omitempty" yaml:"CpusetCpus,omitempty"` + CPUSetMEMs string `json:"CpusetMems,omitempty" yaml:"CpusetMems,omitempty"` + CPUQuota int64 `json:"CpuQuota,omitempty" yaml:"CpuQuota,omitempty"` + CPUPeriod int64 `json:"CpuPeriod,omitempty" yaml:"CpuPeriod,omitempty"` + BlkioWeight int64 `json:"BlkioWeight,omitempty" yaml:"BlkioWeight"` + Ulimits []ULimit `json:"Ulimits,omitempty" yaml:"Ulimits,omitempty"` } // StartContainer starts a container, returning an error in case of failure. // -// See http://goo.gl/iM5GYs for more details. +// See https://goo.gl/MrBAJv for more details. func (c *Client) StartContainer(id string, hostConfig *HostConfig) error { path := "/containers/" + id + "/start" - _, status, err := c.do("POST", path, doOptions{data: hostConfig, forceJSON: true}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: id, Err: err} - } - if status == http.StatusNotModified { - return &ContainerAlreadyRunning{ID: id} - } + resp, err := c.do("POST", path, doOptions{data: hostConfig, forceJSON: true}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: id, Err: err} + } return err } + if resp.StatusCode == http.StatusNotModified { + return &ContainerAlreadyRunning{ID: id} + } + resp.Body.Close() return nil } // StopContainer stops a container, killing it after the given timeout (in // seconds). // -// See http://goo.gl/EbcpXt for more details. +// See https://goo.gl/USqsFt for more details. func (c *Client) StopContainer(id string, timeout uint) error { path := fmt.Sprintf("/containers/%s/stop?t=%d", id, timeout) - _, status, err := c.do("POST", path, doOptions{}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: id} - } - if status == http.StatusNotModified { - return &ContainerNotRunning{ID: id} - } + resp, err := c.do("POST", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: id} + } return err } + if resp.StatusCode == http.StatusNotModified { + return &ContainerNotRunning{ID: id} + } + resp.Body.Close() return nil } // RestartContainer stops a container, killing it after the given timeout (in // seconds), during the stop process. // -// See http://goo.gl/VOzR2n for more details. +// See https://goo.gl/QzsDnz for more details. func (c *Client) RestartContainer(id string, timeout uint) error { path := fmt.Sprintf("/containers/%s/restart?t=%d", id, timeout) - _, status, err := c.do("POST", path, doOptions{}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: id} - } + resp, err := c.do("POST", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: id} + } return err } + resp.Body.Close() return nil } // PauseContainer pauses the given container. // -// See http://goo.gl/AM5t42 for more details. +// See https://goo.gl/OF7W9X for more details. func (c *Client) PauseContainer(id string) error { path := fmt.Sprintf("/containers/%s/pause", id) - _, status, err := c.do("POST", path, doOptions{}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: id} - } + resp, err := c.do("POST", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: id} + } return err } + resp.Body.Close() return nil } // UnpauseContainer unpauses the given container. // -// See http://goo.gl/eBrNSL for more details. +// See https://goo.gl/7dwyPA for more details. func (c *Client) UnpauseContainer(id string) error { path := fmt.Sprintf("/containers/%s/unpause", id) - _, status, err := c.do("POST", path, doOptions{}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: id} - } + resp, err := c.do("POST", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: id} + } return err } + resp.Body.Close() return nil } // TopResult represents the list of processes running in a container, as // returned by /containers//top. // -// See http://goo.gl/qu4gse for more details. +// See https://goo.gl/Rb46aY for more details. type TopResult struct { Titles []string Processes [][]string @@ -560,7 +590,7 @@ type TopResult struct { // TopContainer returns processes running inside a container // -// See http://goo.gl/qu4gse for more details. +// See https://goo.gl/Rb46aY for more details. func (c *Client) TopContainer(id string, psArgs string) (TopResult, error) { var args string var result TopResult @@ -568,15 +598,15 @@ func (c *Client) TopContainer(id string, psArgs string) (TopResult, error) { args = fmt.Sprintf("?ps_args=%s", psArgs) } path := fmt.Sprintf("/containers/%s/top%s", id, args) - body, status, err := c.do("GET", path, doOptions{}) - if status == http.StatusNotFound { - return result, &NoSuchContainer{ID: id} - } + resp, err := c.do("GET", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return result, &NoSuchContainer{ID: id} + } return result, err } - err = json.Unmarshal(body, &result) - if err != nil { + defer resp.Body.Close() + if err := json.NewDecoder(resp.Body).Decode(&result); err != nil { return result, err } return result, nil @@ -584,7 +614,7 @@ func (c *Client) TopContainer(id string, psArgs string) (TopResult, error) { // Stats represents container statistics, returned by /containers//stats. // -// See http://goo.gl/DFMiYD for more details. +// See https://goo.gl/GNmLHb for more details. type Stats struct { Read time.Time `json:"read,omitempty" yaml:"read,omitempty"` Network struct { @@ -674,7 +704,7 @@ type BlkioStatsEntry struct { // StatsOptions specify parameters to the Stats function. // -// See http://goo.gl/DFMiYD for more details. +// See https://goo.gl/GNmLHb for more details. type StatsOptions struct { ID string Stats chan<- *Stats @@ -690,9 +720,10 @@ type StatsOptions struct { // This function is blocking, similar to a streaming call for logs, and should be run // on a separate goroutine from the caller. Note that this function will block until // the given container is removed, not just exited. When finished, this function -// will close the given channel. Alternatively, function can be stopped by signaling on the Done channel +// will close the given channel. Alternatively, function can be stopped by +// signaling on the Done channel. // -// See http://goo.gl/DFMiYD for more details. +// See https://goo.gl/GNmLHb for more details. func (c *Client) Stats(opts StatsOptions) (retErr error) { errC := make(chan error, 1) readCloser, writeCloser := io.Pipe() @@ -750,7 +781,7 @@ func (c *Client) Stats(opts StatsOptions) (retErr error) { decoder := json.NewDecoder(readCloser) stats := new(Stats) - for err := decoder.Decode(&stats); err != io.EOF; err = decoder.Decode(stats) { + for err := decoder.Decode(stats); err != io.EOF; err = decoder.Decode(stats) { if err != nil { return err } @@ -763,7 +794,7 @@ func (c *Client) Stats(opts StatsOptions) (retErr error) { // KillContainerOptions represents the set of options that can be used in a // call to KillContainer. // -// See http://goo.gl/TFkECx for more details. +// See https://goo.gl/hkS9i8 for more details. type KillContainerOptions struct { // The ID of the container. ID string `qs:"-"` @@ -773,24 +804,26 @@ type KillContainerOptions struct { Signal Signal } -// KillContainer kills a container, returning an error in case of failure. +// KillContainer sends a signal to a container, returning an error in case of +// failure. // -// See http://goo.gl/TFkECx for more details. +// See https://goo.gl/hkS9i8 for more details. func (c *Client) KillContainer(opts KillContainerOptions) error { path := "/containers/" + opts.ID + "/kill" + "?" + queryString(opts) - _, status, err := c.do("POST", path, doOptions{}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: opts.ID} - } + resp, err := c.do("POST", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: opts.ID} + } return err } + resp.Body.Close() return nil } // RemoveContainerOptions encapsulates options to remove a container. // -// See http://goo.gl/ZB83ji for more details. +// See https://goo.gl/RQyX62 for more details. type RemoveContainerOptions struct { // The ID of the container. ID string `qs:"-"` @@ -806,64 +839,107 @@ type RemoveContainerOptions struct { // RemoveContainer removes a container, returning an error in case of failure. // -// See http://goo.gl/ZB83ji for more details. +// See https://goo.gl/RQyX62 for more details. func (c *Client) RemoveContainer(opts RemoveContainerOptions) error { path := "/containers/" + opts.ID + "?" + queryString(opts) - _, status, err := c.do("DELETE", path, doOptions{}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: opts.ID} - } + resp, err := c.do("DELETE", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: opts.ID} + } return err } + resp.Body.Close() return nil } -// CopyFromContainerOptions is the set of options that can be used when copying -// files or folders from a container. +// UploadToContainerOptions is the set of options that can be used when +// uploading an archive into a container. // -// See http://goo.gl/rINMlw for more details. +// See https://goo.gl/Ss97HW for more details. +type UploadToContainerOptions struct { + InputStream io.Reader `json:"-" qs:"-"` + Path string `qs:"path"` + NoOverwriteDirNonDir bool `qs:"noOverwriteDirNonDir"` +} + +// UploadToContainer uploads a tar archive to be extracted to a path in the +// filesystem of the container. +// +// See https://goo.gl/Ss97HW for more details. +func (c *Client) UploadToContainer(id string, opts UploadToContainerOptions) error { + url := fmt.Sprintf("/containers/%s/archive?", id) + queryString(opts) + + return c.stream("PUT", url, streamOptions{ + in: opts.InputStream, + }) +} + +// DownloadFromContainerOptions is the set of options that can be used when +// downloading resources from a container. +// +// See https://goo.gl/KnZJDX for more details. +type DownloadFromContainerOptions struct { + OutputStream io.Writer `json:"-" qs:"-"` + Path string `qs:"path"` +} + +// DownloadFromContainer downloads a tar archive of files or folders in a container. +// +// See https://goo.gl/KnZJDX for more details. +func (c *Client) DownloadFromContainer(id string, opts DownloadFromContainerOptions) error { + url := fmt.Sprintf("/containers/%s/archive?", id) + queryString(opts) + + return c.stream("GET", url, streamOptions{ + setRawTerminal: true, + stdout: opts.OutputStream, + }) +} + +// CopyFromContainerOptions has been DEPRECATED, please use DownloadFromContainerOptions along with DownloadFromContainer. +// +// See https://goo.gl/R2jevW for more details. type CopyFromContainerOptions struct { OutputStream io.Writer `json:"-"` Container string `json:"-"` Resource string } -// CopyFromContainer copy files or folders from a container, using a given -// resource. +// CopyFromContainer has been DEPRECATED, please use DownloadFromContainerOptions along with DownloadFromContainer. // -// See http://goo.gl/rINMlw for more details. +// See https://goo.gl/R2jevW for more details. func (c *Client) CopyFromContainer(opts CopyFromContainerOptions) error { if opts.Container == "" { return &NoSuchContainer{ID: opts.Container} } url := fmt.Sprintf("/containers/%s/copy", opts.Container) - body, status, err := c.do("POST", url, doOptions{data: opts}) - if status == http.StatusNotFound { - return &NoSuchContainer{ID: opts.Container} - } + resp, err := c.do("POST", url, doOptions{data: opts}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchContainer{ID: opts.Container} + } return err } - _, err = io.Copy(opts.OutputStream, bytes.NewBuffer(body)) + defer resp.Body.Close() + _, err = io.Copy(opts.OutputStream, resp.Body) return err } // WaitContainer blocks until the given container stops, return the exit code // of the container status. // -// See http://goo.gl/J88DHU for more details. +// See https://goo.gl/Gc1rge for more details. func (c *Client) WaitContainer(id string) (int, error) { - body, status, err := c.do("POST", "/containers/"+id+"/wait", doOptions{}) - if status == http.StatusNotFound { - return 0, &NoSuchContainer{ID: id} - } + resp, err := c.do("POST", "/containers/"+id+"/wait", doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return 0, &NoSuchContainer{ID: id} + } return 0, err } + defer resp.Body.Close() var r struct{ StatusCode int } - err = json.Unmarshal(body, &r) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&r); err != nil { return 0, err } return r.StatusCode, nil @@ -871,7 +947,7 @@ func (c *Client) WaitContainer(id string) (int, error) { // CommitContainerOptions aggregates parameters to the CommitContainer method. // -// See http://goo.gl/Jn8pe8 for more details. +// See https://goo.gl/mqfoCw for more details. type CommitContainerOptions struct { Container string Repository string `qs:"repo"` @@ -883,19 +959,19 @@ type CommitContainerOptions struct { // CommitContainer creates a new image from a container's changes. // -// See http://goo.gl/Jn8pe8 for more details. +// See https://goo.gl/mqfoCw for more details. func (c *Client) CommitContainer(opts CommitContainerOptions) (*Image, error) { path := "/commit?" + queryString(opts) - body, status, err := c.do("POST", path, doOptions{data: opts.Run}) - if status == http.StatusNotFound { - return nil, &NoSuchContainer{ID: opts.Container} - } + resp, err := c.do("POST", path, doOptions{data: opts.Run}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, &NoSuchContainer{ID: opts.Container} + } return nil, err } + defer resp.Body.Close() var image Image - err = json.Unmarshal(body, &image) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&image); err != nil { return nil, err } return &image, nil @@ -904,7 +980,7 @@ func (c *Client) CommitContainer(opts CommitContainerOptions) (*Image, error) { // AttachToContainerOptions is the set of options that can be used when // attaching to a container. // -// See http://goo.gl/RRAhws for more details. +// See https://goo.gl/NKpkFk for more details. type AttachToContainerOptions struct { Container string `qs:"-"` InputStream io.Reader `qs:"-"` @@ -939,7 +1015,7 @@ type AttachToContainerOptions struct { // AttachToContainer attaches to a container, using the given options. // -// See http://goo.gl/RRAhws for more details. +// See https://goo.gl/NKpkFk for more details. func (c *Client) AttachToContainer(opts AttachToContainerOptions) error { if opts.Container == "" { return &NoSuchContainer{ID: opts.Container} @@ -957,7 +1033,7 @@ func (c *Client) AttachToContainer(opts AttachToContainerOptions) error { // LogsOptions represents the set of options used when getting logs from a // container. // -// See http://goo.gl/rLhKSU for more details. +// See https://goo.gl/yl8PGm for more details. type LogsOptions struct { Container string `qs:"-"` OutputStream io.Writer `qs:"-"` @@ -975,7 +1051,7 @@ type LogsOptions struct { // Logs gets stdout and stderr logs from the specified container. // -// See http://goo.gl/rLhKSU for more details. +// See https://goo.gl/yl8PGm for more details. func (c *Client) Logs(opts LogsOptions) error { if opts.Container == "" { return &NoSuchContainer{ID: opts.Container} @@ -992,18 +1068,24 @@ func (c *Client) Logs(opts LogsOptions) error { } // ResizeContainerTTY resizes the terminal to the given height and width. +// +// See https://goo.gl/xERhCc for more details. func (c *Client) ResizeContainerTTY(id string, height, width int) error { params := make(url.Values) params.Set("h", strconv.Itoa(height)) params.Set("w", strconv.Itoa(width)) - _, _, err := c.do("POST", "/containers/"+id+"/resize?"+params.Encode(), doOptions{}) - return err + resp, err := c.do("POST", "/containers/"+id+"/resize?"+params.Encode(), doOptions{}) + if err != nil { + return err + } + resp.Body.Close() + return nil } // ExportContainerOptions is the set of parameters to the ExportContainer // method. // -// See http://goo.gl/hnzE62 for more details. +// See https://goo.gl/dOkTyk for more details. type ExportContainerOptions struct { ID string OutputStream io.Writer @@ -1012,7 +1094,7 @@ type ExportContainerOptions struct { // ExportContainer export the contents of container id as tar archive // and prints the exported contents to stdout. // -// See http://goo.gl/hnzE62 for more details. +// See https://goo.gl/dOkTyk for more details. func (c *Client) ExportContainer(opts ExportContainerOptions) error { if opts.ID == "" { return &NoSuchContainer{ID: opts.ID} diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container_test.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container_test.go index 00966aa19..d42c82fa0 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container_test.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/container_test.go @@ -1122,10 +1122,7 @@ func TestAttachToContainerRawTerminalFalse(t *testing.T) { Stream: true, RawTerminal: false, } - err := client.AttachToContainer(opts) - if err != nil { - t.Fatal(err) - } + client.AttachToContainer(opts) expected := map[string][]string{ "stdin": {"1"}, "stdout": {"1"}, @@ -1374,6 +1371,7 @@ func TestExportContainerViaUnixSocket(t *testing.T) { u, _ := parseEndpoint(endpoint, false) client := Client{ HTTPClient: http.DefaultClient, + Dialer: &net.Dialer{}, endpoint: endpoint, endpointURL: u, SkipServerVersionCheck: true, @@ -1429,17 +1427,60 @@ func TestExportContainerNoId(t *testing.T) { } } +func TestUploadToContainer(t *testing.T) { + content := "File content" + in := stdinMock{bytes.NewBufferString(content)} + fakeRT := &FakeRoundTripper{status: http.StatusOK} + client := newTestClient(fakeRT) + opts := UploadToContainerOptions{ + Path: "abc", + InputStream: in, + } + err := client.UploadToContainer("a123456", opts) + if err != nil { + t.Errorf("UploadToContainer: caught error %#v while uploading archive to container, expected nil", err) + } + + req := fakeRT.requests[0] + + if req.Method != "PUT" { + t.Errorf("UploadToContainer{Path:abc}: Wrong HTTP method. Want PUT. Got %s", req.Method) + } + + if pathParam := req.URL.Query().Get("path"); pathParam != "abc" { + t.Errorf("ListImages({Path:abc}): Wrong parameter. Want path=abc. Got path=%s", pathParam) + } + +} + +func TestDownloadFromContainer(t *testing.T) { + filecontent := "File content" + client := newTestClient(&FakeRoundTripper{message: filecontent, status: http.StatusOK}) + + var out bytes.Buffer + opts := DownloadFromContainerOptions{ + OutputStream: &out, + } + err := client.DownloadFromContainer("a123456", opts) + if err != nil { + t.Errorf("DownloadFromContainer: caught error %#v while downloading from container, expected nil", err.Error()) + } + if out.String() != filecontent { + t.Errorf("DownloadFromContainer: wrong stdout. Want %#v. Got %#v.", filecontent, out.String()) + } +} + func TestCopyFromContainer(t *testing.T) { content := "File content" out := stdoutMock{bytes.NewBufferString(content)} client := newTestClient(&FakeRoundTripper{status: http.StatusOK}) opts := CopyFromContainerOptions{ Container: "a123456", - OutputStream: out, + OutputStream: &out, } err := client.CopyFromContainer(opts) if err != nil { - t.Errorf("CopyFromContainer: caugh error %#v while copying from container, expected nil", err.Error()) + t.Errorf("CopyFromContainer: caught error %#v while copying from container, expected nil", err.Error()) } if out.String() != content { t.Errorf("CopyFromContainer: wrong stdout. Want %#v. Got %#v.", content, out.String()) @@ -1587,7 +1628,6 @@ func TestTopContainerWithPsArgs(t *testing.T) { } func TestStatsTimeout(t *testing.T) { - l, err := net.Listen("unix", "/tmp/docker_test.sock") if err != nil { t.Fatal(err) @@ -1597,7 +1637,7 @@ func TestStatsTimeout(t *testing.T) { go func() { l.Accept() received = true - time.Sleep(time.Millisecond * 250) + time.Sleep(time.Second) }() client, _ := NewClient("unix:///tmp/docker_test.sock") client.SkipServerVersionCheck = true diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/event.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/event.go index 5a85983cc..eaffddb82 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/event.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/event.go @@ -5,7 +5,6 @@ package docker import ( - "crypto/tls" "encoding/json" "errors" "fmt" @@ -260,9 +259,9 @@ func (c *Client) eventHijack(startTime int64, eventChan chan *APIEvents, errChan var dial net.Conn var err error if c.TLSConfig == nil { - dial, err = net.Dial(protocol, address) + dial, err = c.Dialer.Dial(protocol, address) } else { - dial, err = tls.Dial(protocol, address, c.TLSConfig) + dial, err = tlsDialWithDialer(c.Dialer, protocol, address, c.TLSConfig) } if err != nil { return err diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/exec.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/exec.go index bc7c5cfcd..f3b705fa0 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/exec.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/exec.go @@ -2,8 +2,6 @@ // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. -// Docs can currently be found at https://github.com/docker/docker/blob/master/docs/sources/reference/api/docker_remote_api_v1.15.md#exec-create - package docker import ( @@ -15,9 +13,15 @@ import ( "strconv" ) +// Exec is the type representing a `docker exec` instance and containing the +// instance ID +type Exec struct { + ID string `json:"Id,omitempty" yaml:"Id,omitempty"` +} + // CreateExecOptions specify parameters to the CreateExecContainer function. // -// See http://goo.gl/8izrzI for more details +// See https://goo.gl/1KSIb7 for more details type CreateExecOptions struct { AttachStdin bool `json:"AttachStdin,omitempty" yaml:"AttachStdin,omitempty"` AttachStdout bool `json:"AttachStdout,omitempty" yaml:"AttachStdout,omitempty"` @@ -28,9 +32,31 @@ type CreateExecOptions struct { User string `json:"User,omitempty" yaml:"User,omitempty"` } +// CreateExec sets up an exec instance in a running container `id`, returning the exec +// instance, or an error in case of failure. +// +// See https://goo.gl/1KSIb7 for more details +func (c *Client) CreateExec(opts CreateExecOptions) (*Exec, error) { + path := fmt.Sprintf("/containers/%s/exec", opts.Container) + resp, err := c.do("POST", path, doOptions{data: opts}) + if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, &NoSuchContainer{ID: opts.Container} + } + return nil, err + } + defer resp.Body.Close() + var exec Exec + if err := json.NewDecoder(resp.Body).Decode(&exec); err != nil { + return nil, err + } + + return &exec, nil +} + // StartExecOptions specify parameters to the StartExecContainer function. // -// See http://goo.gl/JW8Lxl for more details +// See https://goo.gl/iQCnto for more details type StartExecOptions struct { Detach bool `json:"Detach,omitempty" yaml:"Detach,omitempty"` @@ -51,67 +77,11 @@ type StartExecOptions struct { Success chan struct{} `json:"-"` } -// Exec is the type representing a `docker exec` instance and containing the -// instance ID -type Exec struct { - ID string `json:"Id,omitempty" yaml:"Id,omitempty"` -} - -// ExecProcessConfig is a type describing the command associated to a Exec -// instance. It's used in the ExecInspect type. -// -// See http://goo.gl/ypQULN for more details -type ExecProcessConfig struct { - Privileged bool `json:"privileged,omitempty" yaml:"privileged,omitempty"` - User string `json:"user,omitempty" yaml:"user,omitempty"` - Tty bool `json:"tty,omitempty" yaml:"tty,omitempty"` - EntryPoint string `json:"entrypoint,omitempty" yaml:"entrypoint,omitempty"` - Arguments []string `json:"arguments,omitempty" yaml:"arguments,omitempty"` -} - -// ExecInspect is a type with details about a exec instance, including the -// exit code if the command has finished running. It's returned by a api -// call to /exec/(id)/json -// -// See http://goo.gl/ypQULN for more details -type ExecInspect struct { - ID string `json:"ID,omitempty" yaml:"ID,omitempty"` - Running bool `json:"Running,omitempty" yaml:"Running,omitempty"` - ExitCode int `json:"ExitCode,omitempty" yaml:"ExitCode,omitempty"` - OpenStdin bool `json:"OpenStdin,omitempty" yaml:"OpenStdin,omitempty"` - OpenStderr bool `json:"OpenStderr,omitempty" yaml:"OpenStderr,omitempty"` - OpenStdout bool `json:"OpenStdout,omitempty" yaml:"OpenStdout,omitempty"` - ProcessConfig ExecProcessConfig `json:"ProcessConfig,omitempty" yaml:"ProcessConfig,omitempty"` - Container Container `json:"Container,omitempty" yaml:"Container,omitempty"` -} - -// CreateExec sets up an exec instance in a running container `id`, returning the exec -// instance, or an error in case of failure. -// -// See http://goo.gl/8izrzI for more details -func (c *Client) CreateExec(opts CreateExecOptions) (*Exec, error) { - path := fmt.Sprintf("/containers/%s/exec", opts.Container) - body, status, err := c.do("POST", path, doOptions{data: opts}) - if status == http.StatusNotFound { - return nil, &NoSuchContainer{ID: opts.Container} - } - if err != nil { - return nil, err - } - var exec Exec - err = json.Unmarshal(body, &exec) - if err != nil { - return nil, err - } - - return &exec, nil -} - // StartExec starts a previously set up exec instance id. If opts.Detach is // true, it returns after starting the exec command. Otherwise, it sets up an // interactive session with the exec command. // -// See http://goo.gl/JW8Lxl for more details +// See https://goo.gl/iQCnto for more details func (c *Client) StartExec(id string, opts StartExecOptions) error { if id == "" { return &NoSuchExec{ID: id} @@ -120,13 +90,14 @@ func (c *Client) StartExec(id string, opts StartExecOptions) error { path := fmt.Sprintf("/exec/%s/start", id) if opts.Detach { - _, status, err := c.do("POST", path, doOptions{data: opts}) - if status == http.StatusNotFound { - return &NoSuchExec{ID: id} - } + resp, err := c.do("POST", path, doOptions{data: opts}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return &NoSuchExec{ID: id} + } return err } + defer resp.Body.Close() return nil } @@ -144,32 +115,62 @@ func (c *Client) StartExec(id string, opts StartExecOptions) error { // is valid only if Tty was specified as part of creating and starting the exec // command. // -// See http://goo.gl/YDSx1f for more details +// See https://goo.gl/e1JpsA for more details func (c *Client) ResizeExecTTY(id string, height, width int) error { params := make(url.Values) params.Set("h", strconv.Itoa(height)) params.Set("w", strconv.Itoa(width)) path := fmt.Sprintf("/exec/%s/resize?%s", id, params.Encode()) - _, _, err := c.do("POST", path, doOptions{}) - return err + resp, err := c.do("POST", path, doOptions{}) + if err != nil { + return err + } + resp.Body.Close() + return nil +} + +// ExecProcessConfig is a type describing the command associated to a Exec +// instance. It's used in the ExecInspect type. +type ExecProcessConfig struct { + Privileged bool `json:"privileged,omitempty" yaml:"privileged,omitempty"` + User string `json:"user,omitempty" yaml:"user,omitempty"` + Tty bool `json:"tty,omitempty" yaml:"tty,omitempty"` + EntryPoint string `json:"entrypoint,omitempty" yaml:"entrypoint,omitempty"` + Arguments []string `json:"arguments,omitempty" yaml:"arguments,omitempty"` +} + +// ExecInspect is a type with details about a exec instance, including the +// exit code if the command has finished running. It's returned by a api +// call to /exec/(id)/json +// +// See https://goo.gl/gPtX9R for more details +type ExecInspect struct { + ID string `json:"ID,omitempty" yaml:"ID,omitempty"` + Running bool `json:"Running,omitempty" yaml:"Running,omitempty"` + ExitCode int `json:"ExitCode,omitempty" yaml:"ExitCode,omitempty"` + OpenStdin bool `json:"OpenStdin,omitempty" yaml:"OpenStdin,omitempty"` + OpenStderr bool `json:"OpenStderr,omitempty" yaml:"OpenStderr,omitempty"` + OpenStdout bool `json:"OpenStdout,omitempty" yaml:"OpenStdout,omitempty"` + ProcessConfig ExecProcessConfig `json:"ProcessConfig,omitempty" yaml:"ProcessConfig,omitempty"` + Container Container `json:"Container,omitempty" yaml:"Container,omitempty"` } // InspectExec returns low-level information about the exec command id. // -// See http://goo.gl/ypQULN for more details +// See https://goo.gl/gPtX9R for more details func (c *Client) InspectExec(id string) (*ExecInspect, error) { path := fmt.Sprintf("/exec/%s/json", id) - body, status, err := c.do("GET", path, doOptions{}) - if status == http.StatusNotFound { - return nil, &NoSuchExec{ID: id} - } + resp, err := c.do("GET", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, &NoSuchExec{ID: id} + } return nil, err } + defer resp.Body.Close() var exec ExecInspect - err = json.Unmarshal(body, &exec) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&exec); err != nil { return nil, err } return &exec, nil diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image.go index 4bceb0d3d..c54264616 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image.go @@ -11,7 +11,6 @@ import ( "errors" "fmt" "io" - "io/ioutil" "net/http" "net/url" "os" @@ -46,16 +45,6 @@ type Image struct { VirtualSize int64 `json:"VirtualSize,omitempty" yaml:"VirtualSize,omitempty"` } -// ImageHistory represent a layer in an image's history returned by the -// ImageHistory call. -type ImageHistory struct { - ID string `json:"Id" yaml:"Id"` - Tags []string `json:"Tags,omitempty" yaml:"Tags,omitempty"` - Created int64 `json:"Created,omitempty" yaml:"Created,omitempty"` - CreatedBy string `json:"CreatedBy,omitempty" yaml:"CreatedBy,omitempty"` - Size int64 `json:"Size,omitempty" yaml:"Size,omitempty"` -} - // ImagePre012 serves the same purpose as the Image type except that it is for // earlier versions of the Docker API (pre-012 to be specific) type ImagePre012 struct { @@ -72,15 +61,6 @@ type ImagePre012 struct { Size int64 `json:"size,omitempty"` } -// ListImagesOptions specify parameters to the ListImages function. -// -// See http://goo.gl/HRVN1Z for more details. -type ListImagesOptions struct { - All bool - Filters map[string][]string - Digests bool -} - var ( // ErrNoSuchImage is the error returned when the image does not exist. ErrNoSuchImage = errors.New("no such image") @@ -102,37 +82,56 @@ var ( ErrMustSpecifyNames = errors.New("must specify at least one name to export") ) +// ListImagesOptions specify parameters to the ListImages function. +// +// See https://goo.gl/xBe1u3 for more details. +type ListImagesOptions struct { + All bool + Filters map[string][]string + Digests bool +} + // ListImages returns the list of available images in the server. // -// See http://goo.gl/HRVN1Z for more details. +// See https://goo.gl/xBe1u3 for more details. func (c *Client) ListImages(opts ListImagesOptions) ([]APIImages, error) { path := "/images/json?" + queryString(opts) - body, _, err := c.do("GET", path, doOptions{}) + resp, err := c.do("GET", path, doOptions{}) if err != nil { return nil, err } + defer resp.Body.Close() var images []APIImages - err = json.Unmarshal(body, &images) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&images); err != nil { return nil, err } return images, nil } +// ImageHistory represent a layer in an image's history returned by the +// ImageHistory call. +type ImageHistory struct { + ID string `json:"Id" yaml:"Id"` + Tags []string `json:"Tags,omitempty" yaml:"Tags,omitempty"` + Created int64 `json:"Created,omitempty" yaml:"Created,omitempty"` + CreatedBy string `json:"CreatedBy,omitempty" yaml:"CreatedBy,omitempty"` + Size int64 `json:"Size,omitempty" yaml:"Size,omitempty"` +} + // ImageHistory returns the history of the image by its name or ID. // -// See http://goo.gl/2oJmNs for more details. +// See https://goo.gl/8bnTId for more details. func (c *Client) ImageHistory(name string) ([]ImageHistory, error) { - body, status, err := c.do("GET", "/images/"+name+"/history", doOptions{}) - if status == http.StatusNotFound { - return nil, ErrNoSuchImage - } + resp, err := c.do("GET", "/images/"+name+"/history", doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, ErrNoSuchImage + } return nil, err } + defer resp.Body.Close() var history []ImageHistory - err = json.Unmarshal(body, &history) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&history); err != nil { return nil, err } return history, nil @@ -140,19 +139,23 @@ func (c *Client) ImageHistory(name string) ([]ImageHistory, error) { // RemoveImage removes an image by its name or ID. // -// See http://goo.gl/znj0wM for more details. +// See https://goo.gl/V3ZWnK for more details. func (c *Client) RemoveImage(name string) error { - _, status, err := c.do("DELETE", "/images/"+name, doOptions{}) - if status == http.StatusNotFound { - return ErrNoSuchImage + resp, err := c.do("DELETE", "/images/"+name, doOptions{}) + if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return ErrNoSuchImage + } + return err } - return err + resp.Body.Close() + return nil } // RemoveImageOptions present the set of options available for removing an image // from a registry. // -// See http://goo.gl/6V48bF for more details. +// See https://goo.gl/V3ZWnK for more details. type RemoveImageOptions struct { Force bool `qs:"force"` NoPrune bool `qs:"noprune"` @@ -161,40 +164,43 @@ type RemoveImageOptions struct { // RemoveImageExtended removes an image by its name or ID. // Extra params can be passed, see RemoveImageOptions // -// See http://goo.gl/znj0wM for more details. +// See https://goo.gl/V3ZWnK for more details. func (c *Client) RemoveImageExtended(name string, opts RemoveImageOptions) error { uri := fmt.Sprintf("/images/%s?%s", name, queryString(&opts)) - _, status, err := c.do("DELETE", uri, doOptions{}) - if status == http.StatusNotFound { - return ErrNoSuchImage + resp, err := c.do("DELETE", uri, doOptions{}) + if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return ErrNoSuchImage + } + return err } - return err + resp.Body.Close() + return nil } // InspectImage returns an image by its name or ID. // -// See http://goo.gl/Q112NY for more details. +// See https://goo.gl/jHPcg6 for more details. func (c *Client) InspectImage(name string) (*Image, error) { - body, status, err := c.do("GET", "/images/"+name+"/json", doOptions{}) - if status == http.StatusNotFound { - return nil, ErrNoSuchImage - } + resp, err := c.do("GET", "/images/"+name+"/json", doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, ErrNoSuchImage + } return nil, err } + defer resp.Body.Close() var image Image // if the caller elected to skip checking the server's version, assume it's the latest if c.SkipServerVersionCheck || c.expectedAPIVersion.GreaterThanOrEqualTo(apiVersion112) { - err = json.Unmarshal(body, &image) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&image); err != nil { return nil, err } } else { var imagePre012 ImagePre012 - err = json.Unmarshal(body, &imagePre012) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&imagePre012); err != nil { return nil, err } @@ -216,7 +222,7 @@ func (c *Client) InspectImage(name string) (*Image, error) { // PushImageOptions represents options to use in the PushImage method. // -// See http://goo.gl/pN8A3P for more details. +// See https://goo.gl/zPtZaT for more details. type PushImageOptions struct { // Name of the image Name string @@ -236,7 +242,7 @@ type PushImageOptions struct { // An empty instance of AuthConfiguration may be used for unauthenticated // pushes. // -// See http://goo.gl/pN8A3P for more details. +// See https://goo.gl/zPtZaT for more details. func (c *Client) PushImage(opts PushImageOptions, auth AuthConfiguration) error { if opts.Name == "" { return ErrNoSuchImage @@ -259,7 +265,7 @@ func (c *Client) PushImage(opts PushImageOptions, auth AuthConfiguration) error // PullImageOptions present the set of options available for pulling an image // from a registry. // -// See http://goo.gl/ACyYNS for more details. +// See https://goo.gl/iJkZjD for more details. type PullImageOptions struct { Repository string `qs:"fromImage"` Registry string @@ -268,9 +274,10 @@ type PullImageOptions struct { RawJSONStream bool `qs:"-"` } -// PullImage pulls an image from a remote registry, logging progress to opts.OutputStream. +// PullImage pulls an image from a remote registry, logging progress to +// opts.OutputStream. // -// See http://goo.gl/ACyYNS for more details. +// See https://goo.gl/iJkZjD for more details. func (c *Client) PullImage(opts PullImageOptions, auth AuthConfiguration) error { if opts.Repository == "" { return ErrNoSuchImage @@ -296,14 +303,14 @@ func (c *Client) createImage(qs string, headers map[string]string, in io.Reader, // LoadImageOptions represents the options for LoadImage Docker API Call // -// See http://goo.gl/Y8NNCq for more details. +// See https://goo.gl/JyClMX for more details. type LoadImageOptions struct { InputStream io.Reader } // LoadImage imports a tarball docker image // -// See http://goo.gl/Y8NNCq for more details. +// See https://goo.gl/JyClMX for more details. func (c *Client) LoadImage(opts LoadImageOptions) error { return c.stream("POST", "/images/load", streamOptions{ setRawTerminal: true, @@ -311,17 +318,17 @@ func (c *Client) LoadImage(opts LoadImageOptions) error { }) } -// ExportImageOptions represent the options for ExportImage Docker API call +// ExportImageOptions represent the options for ExportImage Docker API call. // -// See http://goo.gl/mi6kvk for more details. +// See https://goo.gl/le7vK8 for more details. type ExportImageOptions struct { Name string OutputStream io.Writer } -// ExportImage exports an image (as a tar file) into the stream +// ExportImage exports an image (as a tar file) into the stream. // -// See http://goo.gl/mi6kvk for more details. +// See https://goo.gl/le7vK8 for more details. func (c *Client) ExportImage(opts ExportImageOptions) error { return c.stream("GET", fmt.Sprintf("/images/%s/get", opts.Name), streamOptions{ setRawTerminal: true, @@ -331,7 +338,7 @@ func (c *Client) ExportImage(opts ExportImageOptions) error { // ExportImagesOptions represent the options for ExportImages Docker API call // -// See http://goo.gl/YeZzQK for more details. +// See https://goo.gl/huC7HA for more details. type ExportImagesOptions struct { Names []string OutputStream io.Writer `qs:"-"` @@ -339,7 +346,7 @@ type ExportImagesOptions struct { // ExportImages exports one or more images (as a tar file) into the stream // -// See http://goo.gl/YeZzQK for more details. +// See https://goo.gl/huC7HA for more details. func (c *Client) ExportImages(opts ExportImagesOptions) error { if opts.Names == nil || len(opts.Names) == 0 { return ErrMustSpecifyNames @@ -353,7 +360,7 @@ func (c *Client) ExportImages(opts ExportImagesOptions) error { // ImportImageOptions present the set of informations available for importing // an image from a source file or the stdin. // -// See http://goo.gl/PhBKnS for more details. +// See https://goo.gl/iJkZjD for more details. type ImportImageOptions struct { Repository string `qs:"repo"` Source string `qs:"fromSrc"` @@ -366,7 +373,7 @@ type ImportImageOptions struct { // ImportImage imports an image from a url, a file or stdin // -// See http://goo.gl/PhBKnS for more details. +// See https://goo.gl/iJkZjD for more details. func (c *Client) ImportImage(opts ImportImageOptions) error { if opts.Repository == "" { return ErrNoSuchImage @@ -379,8 +386,7 @@ func (c *Client) ImportImage(opts ImportImageOptions) error { if err != nil { return err } - b, err := ioutil.ReadAll(f) - opts.InputStream = bytes.NewBuffer(b) + opts.InputStream = f opts.Source = "-" } return c.createImage(queryString(&opts), nil, opts.InputStream, opts.OutputStream, opts.RawJSONStream) @@ -415,12 +421,12 @@ type BuildImageOptions struct { // BuildImage builds an image from a tarball's url or a Dockerfile in the input // stream. // -// See http://goo.gl/7nuGXa for more details. +// See https://goo.gl/xySxCe for more details. func (c *Client) BuildImage(opts BuildImageOptions) error { if opts.OutputStream == nil { return ErrMissingOutputStream } - headers, err := headersWithAuth(opts.Auth, opts.AuthConfigs) + headers, err := headersWithAuth(opts.Auth, c.versionedAuthConfigs(opts.AuthConfigs)) if err != nil { return err } @@ -452,9 +458,19 @@ func (c *Client) BuildImage(opts BuildImageOptions) error { }) } +func (c *Client) versionedAuthConfigs(authConfigs AuthConfigurations) interface{} { + if c.serverAPIVersion == nil { + c.checkAPIVersion() + } + if c.serverAPIVersion != nil && c.serverAPIVersion.GreaterThanOrEqualTo(apiVersion119) { + return AuthConfigurations119(authConfigs.Configs) + } + return authConfigs +} + // TagImageOptions present the set of options to tag an image. // -// See http://goo.gl/5g6qFy for more details. +// See https://goo.gl/98ZzkU for more details. type TagImageOptions struct { Repo string Tag string @@ -463,15 +479,16 @@ type TagImageOptions struct { // TagImage adds a tag to the image identified by the given name. // -// See http://goo.gl/5g6qFy for more details. +// See https://goo.gl/98ZzkU for more details. func (c *Client) TagImage(name string, opts TagImageOptions) error { if name == "" { return ErrNoSuchImage } - _, status, err := c.do("POST", fmt.Sprintf("/images/"+name+"/tag?%s", + resp, err := c.do("POST", fmt.Sprintf("/images/"+name+"/tag?%s", queryString(&opts)), doOptions{}) + defer resp.Body.Close() - if status == http.StatusNotFound { + if resp.StatusCode == http.StatusNotFound { return ErrNoSuchImage } @@ -497,7 +514,7 @@ func headersWithAuth(auths ...interface{}) (map[string]string, error) { return nil, err } headers["X-Registry-Auth"] = base64.URLEncoding.EncodeToString(buf.Bytes()) - case AuthConfigurations: + case AuthConfigurations, AuthConfigurations119: var buf bytes.Buffer if err := json.NewEncoder(&buf).Encode(auth); err != nil { return nil, err @@ -509,9 +526,9 @@ func headersWithAuth(auths ...interface{}) (map[string]string, error) { return headers, nil } -// APIImageSearch reflect the result of a search on the dockerHub +// APIImageSearch reflect the result of a search on the Docker Hub. // -// See http://goo.gl/xI5lLZ for more details. +// See https://goo.gl/AYjyrF for more details. type APIImageSearch struct { Description string `json:"description,omitempty" yaml:"description,omitempty"` IsOfficial bool `json:"is_official,omitempty" yaml:"is_official,omitempty"` @@ -522,15 +539,15 @@ type APIImageSearch struct { // SearchImages search the docker hub with a specific given term. // -// See http://goo.gl/xI5lLZ for more details. +// See https://goo.gl/AYjyrF for more details. func (c *Client) SearchImages(term string) ([]APIImageSearch, error) { - body, _, err := c.do("GET", "/images/search?term="+term, doOptions{}) + resp, err := c.do("GET", "/images/search?term="+term, doOptions{}) + defer resp.Body.Close() if err != nil { return nil, err } var searchResult []APIImageSearch - err = json.Unmarshal(body, &searchResult) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&searchResult); err != nil { return nil, err } return searchResult, nil diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image_test.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image_test.go index d6bce64cf..11d45e72a 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image_test.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/image_test.go @@ -9,6 +9,7 @@ import ( "encoding/base64" "encoding/json" "io/ioutil" + "net" "net/http" "net/url" "os" @@ -21,11 +22,14 @@ import ( func newTestClient(rt *FakeRoundTripper) Client { endpoint := "http://localhost:4243" u, _ := parseEndpoint("http://localhost:4243", false) + testAPIVersion, _ := NewAPIVersion("1.17") client := Client{ HTTPClient: &http.Client{Transport: rt}, + Dialer: &net.Dialer{}, endpoint: endpoint, endpointURL: u, SkipServerVersionCheck: true, + serverAPIVersion: testAPIVersion, } return client } diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/misc.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/misc.go index 42d1c7e48..34c96531a 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/misc.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/misc.go @@ -4,21 +4,19 @@ package docker -import ( - "bytes" - "strings" -) +import "strings" // Version returns version information about the docker server. // -// See http://goo.gl/BOZrF5 for more details. +// See https://goo.gl/ND9R8L for more details. func (c *Client) Version() (*Env, error) { - body, _, err := c.do("GET", "/version", doOptions{}) + resp, err := c.do("GET", "/version", doOptions{}) if err != nil { return nil, err } + defer resp.Body.Close() var env Env - if err := env.Decode(bytes.NewReader(body)); err != nil { + if err := env.Decode(resp.Body); err != nil { return nil, err } return &env, nil @@ -26,15 +24,15 @@ func (c *Client) Version() (*Env, error) { // Info returns system-wide information about the Docker server. // -// See http://goo.gl/wmqZsW for more details. +// See https://goo.gl/ElTHi2 for more details. func (c *Client) Info() (*Env, error) { - body, _, err := c.do("GET", "/info", doOptions{}) + resp, err := c.do("GET", "/info", doOptions{}) if err != nil { return nil, err } + defer resp.Body.Close() var info Env - err = info.Decode(bytes.NewReader(body)) - if err != nil { + if err := info.Decode(resp.Body); err != nil { return nil, err } return &info, nil diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/network.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/network.go index 0d3e2d43f..e21e7dfea 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/network.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/network.go @@ -38,12 +38,13 @@ type Endpoint struct { // // See https://goo.gl/4hCNtZ for more details. func (c *Client) ListNetworks() ([]Network, error) { - body, _, err := c.do("GET", "/networks", doOptions{}) + resp, err := c.do("GET", "/networks", doOptions{}) if err != nil { return nil, err } + defer resp.Body.Close() var networks []Network - if err := json.Unmarshal(body, &networks); err != nil { + if err := json.NewDecoder(resp.Body).Decode(&networks); err != nil { return nil, err } return networks, nil @@ -54,15 +55,16 @@ func (c *Client) ListNetworks() ([]Network, error) { // See https://goo.gl/4hCNtZ for more details. func (c *Client) NetworkInfo(id string) (*Network, error) { path := "/networks/" + id - body, status, err := c.do("GET", path, doOptions{}) - if status == http.StatusNotFound { - return nil, &NoSuchNetwork{ID: id} - } + resp, err := c.do("GET", path, doOptions{}) if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, &NoSuchNetwork{ID: id} + } return nil, err } + defer resp.Body.Close() var network Network - if err := json.Unmarshal(body, &network); err != nil { + if err := json.NewDecoder(resp.Body).Decode(&network); err != nil { return nil, err } return &network, nil @@ -83,35 +85,34 @@ type CreateNetworkOptions struct { // // See http://goo.gl/mErxNp for more details. func (c *Client) CreateNetwork(opts CreateNetworkOptions) (*Network, error) { - body, status, err := c.do( + resp, err := c.do( "POST", "/networks", doOptions{ data: opts, }, ) - - if status == http.StatusConflict { - return nil, ErrNetworkAlreadyExists - } if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusConflict { + return nil, ErrNetworkAlreadyExists + } return nil, err } + defer resp.Body.Close() type createNetworkResponse struct { ID string } var ( network Network - resp createNetworkResponse + cnr createNetworkResponse ) - err = json.Unmarshal(body, &resp) - if err != nil { + if err := json.NewDecoder(resp.Body).Decode(&cnr); err != nil { return nil, err } network.Name = opts.Name - network.ID = resp.ID + network.ID = cnr.ID network.Type = opts.NetworkType return &network, nil diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/data/symlink b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/data/symlink deleted file mode 120000 index 3ddf86a35..000000000 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/data/symlink +++ /dev/null @@ -1 +0,0 @@ -doesnotexist \ No newline at end of file diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server.go index 41d872771..46f6b6721 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server.go @@ -12,6 +12,7 @@ import ( "encoding/json" "errors" "fmt" + "io/ioutil" mathrand "math/rand" "net" "net/http" @@ -532,7 +533,7 @@ func (s *DockerServer) startContainer(w http.ResponseWriter, r *http.Request) { } container.HostConfig = &hostConfig if container.State.Running { - http.Error(w, "Container already running", http.StatusBadRequest) + http.Error(w, "", http.StatusNotModified) return } container.State.Running = true @@ -610,14 +611,34 @@ func (s *DockerServer) attachContainer(w http.ResponseWriter, r *http.Request) { http.Error(w, err.Error(), http.StatusInternalServerError) return } + wg := sync.WaitGroup{} + if r.URL.Query().Get("stdin") == "1" { + wg.Add(1) + go func() { + ioutil.ReadAll(conn) + wg.Done() + }() + } outStream := stdcopy.NewStdWriter(conn, stdcopy.Stdout) if container.State.Running { - fmt.Fprintf(outStream, "Container %q is running\n", container.ID) + fmt.Fprintf(outStream, "Container is running\n") } else { - fmt.Fprintf(outStream, "Container %q is not running\n", container.ID) + fmt.Fprintf(outStream, "Container is not running\n") } fmt.Fprintln(outStream, "What happened?") fmt.Fprintln(outStream, "Something happened") + wg.Wait() + if r.URL.Query().Get("stream") == "1" { + for { + time.Sleep(1e6) + s.cMut.RLock() + if !container.State.Running { + s.cMut.RUnlock() + break + } + s.cMut.RUnlock() + } + } conn.Close() } @@ -936,7 +957,7 @@ func (s *DockerServer) createExecContainer(w http.ResponseWriter, r *http.Reques func (s *DockerServer) startExecContainer(w http.ResponseWriter, r *http.Request) { id := mux.Vars(r)["id"] - if exec, err := s.getExec(id); err == nil { + if exec, err := s.getExec(id, false); err == nil { s.execMut.Lock() exec.Running = true s.execMut.Unlock() @@ -958,7 +979,7 @@ func (s *DockerServer) startExecContainer(w http.ResponseWriter, r *http.Request func (s *DockerServer) resizeExecContainer(w http.ResponseWriter, r *http.Request) { id := mux.Vars(r)["id"] - if _, err := s.getExec(id); err == nil { + if _, err := s.getExec(id, false); err == nil { w.WriteHeader(http.StatusOK) return } @@ -967,7 +988,7 @@ func (s *DockerServer) resizeExecContainer(w http.ResponseWriter, r *http.Reques func (s *DockerServer) inspectExecContainer(w http.ResponseWriter, r *http.Request) { id := mux.Vars(r)["id"] - if exec, err := s.getExec(id); err == nil { + if exec, err := s.getExec(id, true); err == nil { w.WriteHeader(http.StatusOK) w.Header().Set("Content-Type", "application/json") json.NewEncoder(w).Encode(exec) @@ -976,11 +997,15 @@ func (s *DockerServer) inspectExecContainer(w http.ResponseWriter, r *http.Reque w.WriteHeader(http.StatusNotFound) } -func (s *DockerServer) getExec(id string) (*docker.ExecInspect, error) { +func (s *DockerServer) getExec(id string, copy bool) (*docker.ExecInspect, error) { s.execMut.RLock() defer s.execMut.RUnlock() for _, exec := range s.execs { if exec.ID == id { + if copy { + cp := *exec + exec = &cp + } return exec, nil } } diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server_test.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server_test.go index 36789abb3..bb24ce560 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server_test.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/testing/server_test.go @@ -5,9 +5,11 @@ package testing import ( + "bufio" "bytes" "encoding/json" "fmt" + "io/ioutil" "math/rand" "net" "net/http" @@ -624,8 +626,8 @@ func TestStartContainerAlreadyRunning(t *testing.T) { path := fmt.Sprintf("/containers/%s/start", server.containers[0].ID) request, _ := http.NewRequest("POST", path, bytes.NewBuffer([]byte("null"))) server.ServeHTTP(recorder, request) - if recorder.Code != http.StatusBadRequest { - t.Errorf("StartContainer: wrong status code. Want %d. Got %d.", http.StatusBadRequest, recorder.Code) + if recorder.Code != http.StatusNotModified { + t.Errorf("StartContainer: wrong status code. Want %d. Got %d.", http.StatusNotModified, recorder.Code) } } @@ -845,22 +847,41 @@ func TestWaitContainerNotFound(t *testing.T) { } } +type HijackableResponseRecorder struct { + httptest.ResponseRecorder + readCh chan []byte +} + +func (r *HijackableResponseRecorder) Hijack() (net.Conn, *bufio.ReadWriter, error) { + myConn, otherConn := net.Pipe() + r.readCh = make(chan []byte) + go func() { + data, _ := ioutil.ReadAll(myConn) + r.readCh <- data + }() + return otherConn, nil, nil +} + +func (r *HijackableResponseRecorder) HijackBuffer() string { + return string(<-r.readCh) +} + func TestAttachContainer(t *testing.T) { server := DockerServer{} addContainers(&server, 1) server.containers[0].State.Running = true server.buildMuxer() - recorder := httptest.NewRecorder() + recorder := &HijackableResponseRecorder{} path := fmt.Sprintf("/containers/%s/attach?logs=1", server.containers[0].ID) request, _ := http.NewRequest("POST", path, nil) server.ServeHTTP(recorder, request) lines := []string{ - fmt.Sprintf("\x01\x00\x00\x00\x03\x00\x00\x00Container %q is running", server.containers[0].ID), - "What happened?", - "Something happened", + "\x01\x00\x00\x00\x00\x00\x00\x15Container is running", + "\x01\x00\x00\x00\x00\x00\x00\x0fWhat happened?", + "\x01\x00\x00\x00\x00\x00\x00\x13Something happened", } expected := strings.Join(lines, "\n") + "\n" - if body := recorder.Body.String(); body == expected { + if body := recorder.HijackBuffer(); body != expected { t.Errorf("AttachContainer: wrong body. Want %q. Got %q.", expected, body) } } @@ -868,7 +889,7 @@ func TestAttachContainer(t *testing.T) { func TestAttachContainerNotFound(t *testing.T) { server := DockerServer{} server.buildMuxer() - recorder := httptest.NewRecorder() + recorder := &HijackableResponseRecorder{} path := "/containers/abc123/attach?logs=1" request, _ := http.NewRequest("POST", path, nil) server.ServeHTTP(recorder, request) @@ -877,6 +898,44 @@ func TestAttachContainerNotFound(t *testing.T) { } } +func TestAttachContainerWithStreamBlocks(t *testing.T) { + server := DockerServer{} + addContainers(&server, 1) + server.containers[0].State.Running = true + server.buildMuxer() + path := fmt.Sprintf("/containers/%s/attach?logs=1&stdout=1&stream=1", server.containers[0].ID) + request, _ := http.NewRequest("POST", path, nil) + done := make(chan string) + go func() { + recorder := &HijackableResponseRecorder{} + server.ServeHTTP(recorder, request) + done <- recorder.HijackBuffer() + }() + select { + case <-done: + t.Fatalf("attach stream returned before container is stopped") + case <-time.After(500 * time.Millisecond): + } + server.cMut.Lock() + server.containers[0].State.Running = false + server.cMut.Unlock() + var body string + select { + case body = <-done: + case <-time.After(5 * time.Second): + t.Fatalf("timed out waiting for attach to finish") + } + lines := []string{ + "\x01\x00\x00\x00\x00\x00\x00\x15Container is running", + "\x01\x00\x00\x00\x00\x00\x00\x0fWhat happened?", + "\x01\x00\x00\x00\x00\x00\x00\x13Something happened", + } + expected := strings.Join(lines, "\n") + "\n" + if body != expected { + t.Errorf("AttachContainer: wrong body. Want %q. Got %q.", expected, body) + } +} + func TestRemoveContainer(t *testing.T) { server := DockerServer{} addContainers(&server, 1) @@ -1690,7 +1749,7 @@ func addNetworks(server *DockerServer, n int) { ID: fmt.Sprintf("%x", rand.Int()%10000), Type: "bridge", Endpoints: []*docker.Endpoint{ - &docker.Endpoint{ + { Name: "blah", ID: fmt.Sprintf("%x", rand.Int()%10000), Network: netid, diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/tls.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/tls.go index 11d571761..55f43174b 100644 --- a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/tls.go +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/tls.go @@ -94,7 +94,3 @@ func tlsDialWithDialer(dialer *net.Dialer, network, addr string, config *tls.Con // wrapper which holds both the TLS and raw connections. return &tlsClientCon{conn, rawConn}, nil } - -func tlsDial(network, addr string, config *tls.Config) (net.Conn, error) { - return tlsDialWithDialer(new(net.Dialer), network, addr, config) -} diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/volume.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/volume.go new file mode 100644 index 000000000..a989a6eee --- /dev/null +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/volume.go @@ -0,0 +1,127 @@ +// Copyright 2015 go-dockerclient authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package docker + +import ( + "encoding/json" + "errors" + "net/http" +) + +var ( + // ErrNoSuchVolume is the error returned when the volume does not exist. + ErrNoSuchVolume = errors.New("no such volume") + + // ErrVolumeInUse is the error returned when the volume requested to be removed is still in use. + ErrVolumeInUse = errors.New("volume in use and cannot be removed") +) + +// Volume represents a volume. +// +// See https://goo.gl/FZA4BK for more details. +type Volume struct { + Name string `json:"Name" yaml:"Name"` + Driver string `json:"Driver,omitempty" yaml:"Driver,omitempty"` + Mountpoint string `json:"Mountpoint,omitempty" yaml:"Mountpoint,omitempty"` +} + +// ListVolumesOptions specify parameters to the ListVolumes function. +// +// See https://goo.gl/FZA4BK for more details. +type ListVolumesOptions struct { + Filters map[string][]string +} + +// ListVolumes returns a list of available volumes in the server. +// +// See https://goo.gl/FZA4BK for more details. +func (c *Client) ListVolumes(opts ListVolumesOptions) ([]Volume, error) { + resp, err := c.do("GET", "/volumes?"+queryString(opts), doOptions{}) + if err != nil { + return nil, err + } + defer resp.Body.Close() + m := make(map[string]interface{}) + if err := json.NewDecoder(resp.Body).Decode(&m); err != nil { + return nil, err + } + var volumes []Volume + volumesJSON, ok := m["Volumes"] + if !ok { + return volumes, nil + } + data, err := json.Marshal(volumesJSON) + if err != nil { + return nil, err + } + if err := json.Unmarshal(data, &volumes); err != nil { + return nil, err + } + return volumes, nil +} + +// CreateVolumeOptions specify parameters to the CreateVolume function. +// +// See https://goo.gl/pBUbZ9 for more details. +type CreateVolumeOptions struct { + Name string + Driver string + DriverOpts map[string]string +} + +// CreateVolume creates a volume on the server. +// +// See https://goo.gl/pBUbZ9 for more details. +func (c *Client) CreateVolume(opts CreateVolumeOptions) (*Volume, error) { + resp, err := c.do("POST", "/volumes", doOptions{data: opts}) + if err != nil { + return nil, err + } + defer resp.Body.Close() + var volume Volume + if err := json.NewDecoder(resp.Body).Decode(&volume); err != nil { + return nil, err + } + return &volume, nil +} + +// InspectVolume returns a volume by its name. +// +// See https://goo.gl/0g9A6i for more details. +func (c *Client) InspectVolume(name string) (*Volume, error) { + resp, err := c.do("GET", "/volumes/"+name, doOptions{}) + if err != nil { + if e, ok := err.(*Error); ok && e.Status == http.StatusNotFound { + return nil, ErrNoSuchVolume + } + return nil, err + } + defer resp.Body.Close() + var volume Volume + if err := json.NewDecoder(resp.Body).Decode(&volume); err != nil { + return nil, err + } + return &volume, nil +} + +// RemoveVolume removes a volume by its name. +// +// See https://goo.gl/79GNQz for more details. +func (c *Client) RemoveVolume(name string) error { + resp, err := c.do("DELETE", "/volumes/"+name, doOptions{}) + if err != nil { + if e, ok := err.(*Error); ok { + if e.Status == http.StatusNotFound { + return ErrNoSuchVolume + } + if e.Status == http.StatusConflict { + return ErrVolumeInUse + } + } + return nil + } + defer resp.Body.Close() + return nil +} diff --git a/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/volume_test.go b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/volume_test.go new file mode 100644 index 000000000..9707c09cd --- /dev/null +++ b/Godeps/_workspace/src/github.com/fsouza/go-dockerclient/volume_test.go @@ -0,0 +1,142 @@ +// Copyright 2015 go-dockerclient authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package docker + +import ( + "encoding/json" + "net/http" + "net/url" + "reflect" + "testing" +) + +func TestListVolumes(t *testing.T) { + volumesData := `[ + { + "Name": "tardis", + "Driver": "local", + "Mountpoint": "/var/lib/docker/volumes/tardis" + }, + { + "Name": "foo", + "Driver": "bar", + "Mountpoint": "/var/lib/docker/volumes/bar" + } +]` + body := `{ "Volumes": ` + volumesData + ` }` + var expected []Volume + if err := json.Unmarshal([]byte(volumesData), &expected); err != nil { + t.Fatal(err) + } + client := newTestClient(&FakeRoundTripper{message: body, status: http.StatusOK}) + volumes, err := client.ListVolumes(ListVolumesOptions{}) + if err != nil { + t.Error(err) + } + if !reflect.DeepEqual(volumes, expected) { + t.Errorf("ListVolumes: Wrong return value. Want %#v. Got %#v.", expected, volumes) + } +} + +func TestCreateVolume(t *testing.T) { + body := `{ + "Name": "tardis", + "Driver": "local", + "Mountpoint": "/var/lib/docker/volumes/tardis" + }` + var expected Volume + if err := json.Unmarshal([]byte(body), &expected); err != nil { + t.Fatal(err) + } + fakeRT := &FakeRoundTripper{message: body, status: http.StatusOK} + client := newTestClient(fakeRT) + volume, err := client.CreateVolume( + CreateVolumeOptions{ + Name: "tardis", + Driver: "local", + DriverOpts: map[string]string{ + "foo": "bar", + }, + }, + ) + if err != nil { + t.Fatal(err) + } + if !reflect.DeepEqual(volume, &expected) { + t.Errorf("CreateVolume: Wrong return value. Want %#v. Got %#v.", expected, volume) + } + req := fakeRT.requests[0] + expectedMethod := "POST" + if req.Method != expectedMethod { + t.Errorf("CreateVolume(): Wrong HTTP method. Want %s. Got %s.", expectedMethod, req.Method) + } + u, _ := url.Parse(client.getURL("/volumes")) + if req.URL.Path != u.Path { + t.Errorf("CreateVolume(): Wrong request path. Want %q. Got %q.", u.Path, req.URL.Path) + } +} + +func TestInspectVolume(t *testing.T) { + body := `{ + "Name": "tardis", + "Driver": "local", + "Mountpoint": "/var/lib/docker/volumes/tardis" + }` + var expected Volume + if err := json.Unmarshal([]byte(body), &expected); err != nil { + t.Fatal(err) + } + fakeRT := &FakeRoundTripper{message: body, status: http.StatusOK} + client := newTestClient(fakeRT) + name := "tardis" + volume, err := client.InspectVolume(name) + if err != nil { + t.Fatal(err) + } + if !reflect.DeepEqual(volume, &expected) { + t.Errorf("InspectVolume: Wrong return value. Want %#v. Got %#v.", expected, volume) + } + req := fakeRT.requests[0] + expectedMethod := "GET" + if req.Method != expectedMethod { + t.Errorf("InspectVolume(%q): Wrong HTTP method. Want %s. Got %s.", name, expectedMethod, req.Method) + } + u, _ := url.Parse(client.getURL("/volumes/" + name)) + if req.URL.Path != u.Path { + t.Errorf("CreateVolume(%q): Wrong request path. Want %q. Got %q.", name, u.Path, req.URL.Path) + } +} + +func TestRemoveVolume(t *testing.T) { + name := "test" + fakeRT := &FakeRoundTripper{message: "", status: http.StatusNoContent} + client := newTestClient(fakeRT) + if err := client.RemoveVolume(name); err != nil { + t.Fatal(err) + } + req := fakeRT.requests[0] + expectedMethod := "DELETE" + if req.Method != expectedMethod { + t.Errorf("RemoveVolume(%q): Wrong HTTP method. Want %s. Got %s.", name, expectedMethod, req.Method) + } + u, _ := url.Parse(client.getURL("/volumes/" + name)) + if req.URL.Path != u.Path { + t.Errorf("RemoveVolume(%q): Wrong request path. Want %q. Got %q.", name, u.Path, req.URL.Path) + } +} + +func TestRemoveVolumeNotFound(t *testing.T) { + client := newTestClient(&FakeRoundTripper{message: "no such volume", status: http.StatusNotFound}) + if err := client.RemoveVolume("test:"); err != ErrNoSuchVolume { + t.Errorf("RemoveVolume: wrong error. Want %#v. Got %#v.", ErrNoSuchVolume, err) + } +} + +func TestRemoveVolumeInUse(t *testing.T) { + client := newTestClient(&FakeRoundTripper{message: "volume in use and cannot be removed", status: http.StatusConflict}) + if err := client.RemoveVolume("test:"); err != ErrVolumeInUse { + t.Errorf("RemoveVolume: wrong error. Want %#v. Got %#v.", ErrVolumeInUse, err) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/.gitignore b/Godeps/_workspace/src/github.com/influxdb/influxdb/.gitignore new file mode 100644 index 000000000..147b88122 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/.gitignore @@ -0,0 +1,67 @@ +*~ +src/ + +config.json +/bin/ + +/pkg/ + +TAGS + +# vim temp files +*.swp + +*.test +/query/a.out* +.DS_Store + +# ignore generated files. +cmd/influxd/version.go + +# executables + +influxd +**/influxd +!**/influxd/ + +influx +**/influx +!**/influx/ + +influxdb +**/influxdb +!**/influxdb/ + +/benchmark-tool +/main +/benchmark-storage +godef +gosym +gocode +inspect-raft + +# dependencies +out_rpm/ +packages/ + +# autconf +autom4te.cache/ +config.log +config.status +Makefile + +# log file +influxdb.log +benchmark.log + +# config file +config.toml + +# test data files +integration/migration_data/ + +# goide project files +.idea + +# goconvey config files +*.goconvey diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/CHANGELOG.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/CHANGELOG.md new file mode 100644 index 000000000..89eb85a48 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/CHANGELOG.md @@ -0,0 +1,1607 @@ +## v0.9.3 [unreleased] + +### Release Notes + +There are breaking changes in this release. + - To store data points as integers you must now append i to the number if using the line protocol. + - If you have a UDP input configured, you should check the UDP section of [the new sample configuration file](https://github.com/influxdb/influxdb/blob/master/etc/config.sample.toml) to learn how to modify existing configuration files, as 0.9.3 now expects multiple UDP inputs. + - Configuration files must now have an entry for `wal-dir` in the `[data]` section. Check [new sample configuration file](https://github.com/influxdb/influxdb/blob/master/etc/config.sample.toml) for more details. + - The implicit `GROUP BY *` that was added to every `SELECT *` has been removed. Instead any tags in the data are now part of the columns in the returned query. + +Please see the *Features* section below for full details. + +### Features +- [#3376](https://github.com/influxdb/influxdb/pull/3376): Support for remote shard query mapping +- [#3372](https://github.com/influxdb/influxdb/pull/3372): Support joining nodes to existing cluster +- [#3426](https://github.com/influxdb/influxdb/pull/3426): Additional logging for continuous queries. Thanks @jhorwit2 +- [#3478](https://github.com/influxdb/influxdb/pull/3478): Support incremental cluster joins +- [#3519](https://github.com/influxdb/influxdb/pull/3519): **--BREAKING CHANGE--** Update line protocol to require trailing i for field values that are integers +- [#3529](https://github.com/influxdb/influxdb/pull/3529): Add TLS support for OpenTSDB plugin. Thanks @nathanielc +- [#3421](https://github.com/influxdb/influxdb/issues/3421): Should update metastore and cluster if IP or hostname changes +- [#3502](https://github.com/influxdb/influxdb/pull/3502): Importer for 0.8.9 data via the CLI +- [#3564](https://github.com/influxdb/influxdb/pull/3564): Fix alias, maintain column sort order +- [#3585](https://github.com/influxdb/influxdb/pull/3585): Additional test coverage for non-existent fields +- [#3246](https://github.com/influxdb/influxdb/issues/3246): Allow overriding of configuration parameters using environment variables +- [#3599](https://github.com/influxdb/influxdb/pull/3599): **--BREAKING CHANGE--** Support multiple UDP inputs. Thanks @tpitale +- [#3636](https://github.com/influxdb/influxdb/pull/3639): Cap auto-created retention policy replica count at 3 +- [#3641](https://github.com/influxdb/influxdb/pull/3641): Logging enhancements and single-node rename +- [#3635](https://github.com/influxdb/influxdb/pull/3635): Add build branch to version output. +- [#3115](https://github.com/influxdb/influxdb/pull/3115): Various init.d script improvements. Thanks @KoeSystems. +- [#3628](https://github.com/influxdb/influxdb/pull/3628): Wildcard expansion of tags and fields for raw queries +- [#3721](https://github.com/influxdb/influxdb/pull/3721): interpret number literals compared against time as nanoseconds from epoch +- [#3514](https://github.com/influxdb/influxdb/issues/3514): Implement WAL outside BoltDB with compaction +- [#3544](https://github.com/influxdb/influxdb/pull/3544): Implement compression on top of BoltDB +- [#3795](https://github.com/influxdb/influxdb/pull/3795): Throttle import +- [#3584](https://github.com/influxdb/influxdb/pull/3584): Import/export documenation + +### Bugfixes +- [#3405](https://github.com/influxdb/influxdb/pull/3405): Prevent database panic when fields are missing. Thanks @jhorwit2 +- [#3411](https://github.com/influxdb/influxdb/issues/3411): 500 timeout on write +- [#3420](https://github.com/influxdb/influxdb/pull/3420): Catch opentsdb malformed tags. Thanks @nathanielc. +- [#3404](https://github.com/influxdb/influxdb/pull/3404): Added support for escaped single quotes in query string. Thanks @jhorwit2 +- [#3414](https://github.com/influxdb/influxdb/issues/3414): Shard mappers perform query re-writing +- [#3525](https://github.com/influxdb/influxdb/pull/3525): check if fields are valid during parse time. +- [#3511](https://github.com/influxdb/influxdb/issues/3511): Sending a large number of tag causes panic +- [#3288](https://github.com/influxdb/influxdb/issues/3288): Run go fuzz on the line-protocol input +- [#3545](https://github.com/influxdb/influxdb/issues/3545): Fix parsing string fields with newlines +- [#3579](https://github.com/influxdb/influxdb/issues/3579): Revert breaking change to `client.NewClient` function +- [#3580](https://github.com/influxdb/influxdb/issues/3580): Do not allow wildcards with fields in select statements +- [#3530](https://github.com/influxdb/influxdb/pull/3530): Aliasing a column no longer works +- [#3436](https://github.com/influxdb/influxdb/issues/3436): Fix panic in hinted handoff queue processor +- [#3401](https://github.com/influxdb/influxdb/issues/3401): Derivative on non-numeric fields panics db +- [#3583](https://github.com/influxdb/influxdb/issues/3583): Inserting value in scientific notation with a trailing i causes panic +- [#3611](https://github.com/influxdb/influxdb/pull/3611): Fix query arithmetic with integers +- [#3326](https://github.com/influxdb/influxdb/issues/3326): simple regex query fails with cryptic error +- [#3618](https://github.com/influxdb/influxdb/pull/3618): Fix collectd stats panic on i386. Thanks @richterger +- [#3625](https://github.com/influxdb/influxdb/pull/3625): Don't panic when aggregate and raw queries are in a single statement +- [#3629](https://github.com/influxdb/influxdb/pull/3629): Use sensible batching defaults for Graphite. +- [#3638](https://github.com/influxdb/influxdb/pull/3638): Cluster config fixes and removal of meta.peers config field +- [#3640](https://github.com/influxdb/influxdb/pull/3640): Shutdown Graphite service when signal received. +- [#3632](https://github.com/influxdb/influxdb/issues/3632): Make single-node host renames more seamless +- [#3656](https://github.com/influxdb/influxdb/issues/3656): Silence snapshotter logger for testing +- [#3651](https://github.com/influxdb/influxdb/pull/3651): Fully remove series when dropped. +- [#3517](https://github.com/influxdb/influxdb/pull/3517): Batch CQ writes to avoid timeouts. Thanks @dim. +- [#3522](https://github.com/influxdb/influxdb/pull/3522): Consume CQ results on request timeouts. Thanks @dim. +- [#3646](https://github.com/influxdb/influxdb/pull/3646): Fix nil FieldCodec panic. +- [#3672](https://github.com/influxdb/influxdb/pull/3672): Reduce in-memory index by 20%-30% +- [#3673](https://github.com/influxdb/influxdb/pull/3673): Improve query performance by removing unnecessary tagset sorting. +- [#3676](https://github.com/influxdb/influxdb/pull/3676): Improve query performance by memomizing mapper output keys. +- [#3686](https://github.com/influxdb/influxdb/pull/3686): Ensure 'p' parameter is not logged, even on OPTIONS requests. +- [#3687](https://github.com/influxdb/influxdb/issues/3687): Fix panic: runtime error: makeslice: len out of range in hinted handoff +- [#3697](https://github.com/influxdb/influxdb/issues/3697): Correctly merge non-chunked results for same series. Fix issue #3242. +- [#3708](https://github.com/influxdb/influxdb/issues/3708): Fix double escaping measurement name during cluster replication +- [#3704](https://github.com/influxdb/influxdb/issues/3704): cluster replication issue for measurement name containing backslash +- [#3681](https://github.com/influxdb/influxdb/issues/3681): Quoted measurement names fail +- [#3681](https://github.com/influxdb/influxdb/issues/3682): Fix inserting string value with backslashes +- [#3735](https://github.com/influxdb/influxdb/issues/3735): Append to small bz1 blocks +- [#3736](https://github.com/influxdb/influxdb/pull/3736): Update shard group duration with retention policy changes. Thanks for the report @papylhomme +- [#3539](https://github.com/influxdb/influxdb/issues/3539): parser incorrectly accepts NaN as numerical value, but not always +- [#3790](https://github.com/influxdb/influxdb/pull/3790): Fix line protocol parsing equals in measurements and NaN values +- [#3778](https://github.com/influxdb/influxdb/pull/3778): Don't panic if SELECT on time. +- [#3824](https://github.com/influxdb/influxdb/issues/3824): tsdb.Point.MarshalBinary needs to support all number types +- [#3828](https://github.com/influxdb/influxdb/pull/3828): Support all number types when decoding a point +- [#3853](https://github.com/influxdb/influxdb/pull/3853): Use 4KB default block size for bz1 +- [#3607](https://github.com/influxdb/influxdb/issues/3607): Fix unable to query influxdb due to deadlock in metastore. Thanks @ccutrer! + +## v0.9.2 [2015-07-24] + +### Features +- [#3177](https://github.com/influxdb/influxdb/pull/3177): Client supports making HTTPS requests. Thanks @jipperinbham +- [#3299](https://github.com/influxdb/influxdb/pull/3299): Refactor query engine for distributed query support. +- [#3334](https://github.com/influxdb/influxdb/pull/3334): Clean shutdown of influxd. Thanks @mcastilho + +### Bugfixes + +- [#3180](https://github.com/influxdb/influxdb/pull/3180): Log GOMAXPROCS, version, and commit on startup. +- [#3218](https://github.com/influxdb/influxdb/pull/3218): Allow write timeouts to be configurable. +- [#3184](https://github.com/influxdb/influxdb/pull/3184): Support basic auth in admin interface. Thanks @jipperinbham! +- [#3236](https://github.com/influxdb/influxdb/pull/3236): Fix display issues in admin interface. +- [#3232](https://github.com/influxdb/influxdb/pull/3232): Set logging prefix for metastore. +- [#3230](https://github.com/influxdb/influxdb/issues/3230): panic: unable to parse bool value +- [#3245](https://github.com/influxdb/influxdb/issues/3245): Error using graphite plugin with multiple filters +- [#3223](https://github.com/influxdb/influxdb/issues/323): default graphite template cannot have extra tags +- [#3255](https://github.com/influxdb/influxdb/pull/3255): Flush WAL on start-up as soon as possible. +- [#3289](https://github.com/influxdb/influxdb/issues/3289): InfluxDB crashes on floats without decimal +- [#3298](https://github.com/influxdb/influxdb/pull/3298): Corrected WAL & flush parameters in default config. Thanks @jhorwit2 +- [#3152](https://github.com/influxdb/influxdb/issues/3159): High CPU Usage with unsorted writes +- [#3307](https://github.com/influxdb/influxdb/pull/3307): Fix regression parsing boolean values True/False +- [#3304](https://github.com/influxdb/influxdb/pull/3304): Fixed httpd logger to log user from query params. Thanks @jhorwit2 +- [#3332](https://github.com/influxdb/influxdb/pull/3332): Add SLIMIT and SOFFSET to string version of AST. +- [#3335](https://github.com/influxdb/influxdb/pull/3335): Don't drop all data on DROP DATABASE. Thanks to @PierreF for the report +- [#2761](https://github.com/influxdb/influxdb/issues/2761): Make SHOW RETENTION POLICIES consistent with other queries. +- [#3356](https://github.com/influxdb/influxdb/pull/3356): Disregard semicolons after database name in use command. Thanks @timraymond. +- [#3351](https://github.com/influxdb/influxdb/pull/3351): Handle malformed regex comparisons during parsing. Thanks @rnubel +- [#3244](https://github.com/influxdb/influxdb/pull/3244): Wire up admin privilege grant and revoke. +- [#3259](https://github.com/influxdb/influxdb/issues/3259): Respect privileges for queries. +- [#3256](https://github.com/influxdb/influxdb/pull/3256): Remove unnecessary timeout in WaitForLeader(). Thanks @cannium. +- [#3380](https://github.com/influxdb/influxdb/issue/3380): Parser fix, only allow ORDER BY ASC and ORDER BY time ASC. +- [#3319](https://github.com/influxdb/influxdb/issues/3319): restarting process irrevocably BREAKS measurements with spaces +- [#3453](https://github.com/influxdb/influxdb/issues/3453): Remove outdated `dump` command from CLI. +- [#3463](https://github.com/influxdb/influxdb/issues/3463): Fix aggregate queries and time precision on where clauses. + +## v0.9.1 [2015-07-02] + +### Features + +- [2650](https://github.com/influxdb/influxdb/pull/2650): Add SHOW GRANTS FOR USER statement. Thanks @n1tr0g. +- [3125](https://github.com/influxdb/influxdb/pull/3125): Graphite Input Protocol Parsing +- [2746](https://github.com/influxdb/influxdb/pull/2746): New Admin UI/interface +- [3036](https://github.com/influxdb/influxdb/pull/3036): Write Ahead Log (WAL) +- [3014](https://github.com/influxdb/influxdb/issues/3014): Implement Raft snapshots + +### Bugfixes + +- [3013](https://github.com/influxdb/influxdb/issues/3013): Panic error with inserting values with commas +- [#2956](https://github.com/influxdb/influxdb/issues/2956): Type mismatch in derivative +- [#2908](https://github.com/influxdb/influxdb/issues/2908): Field mismatch error messages need to be updated +- [#2931](https://github.com/influxdb/influxdb/pull/2931): Services and reporting should wait until cluster has leader. +- [#2943](https://github.com/influxdb/influxdb/issues/2943): Ensure default retention policies are fully replicated +- [#2948](https://github.com/influxdb/influxdb/issues/2948): Field mismatch error message to include measurement name +- [#2919](https://github.com/influxdb/influxdb/issues/2919): Unable to insert negative floats +- [#2935](https://github.com/influxdb/influxdb/issues/2935): Hook CPU and memory profiling back up. +- [#2960](https://github.com/influxdb/influxdb/issues/2960): Cluster Write Errors. +- [#2928](https://github.com/influxdb/influxdb/pull/2928): Start work to set InfluxDB version in HTTP response headers. Thanks @neonstalwart. +- [#2969](https://github.com/influxdb/influxdb/pull/2969): Actually set HTTP version in responses. +- [#2993](https://github.com/influxdb/influxdb/pull/2993): Don't log each UDP batch. +- [#2994](https://github.com/influxdb/influxdb/pull/2994): Don't panic during wilcard expansion if no default database specified. +- [#3002](https://github.com/influxdb/influxdb/pull/3002): Remove measurement from shard's index on DROP MEASUREMENT. +- [#3021](https://github.com/influxdb/influxdb/pull/3021): Correct set HTTP write trace logging. Thanks @vladlopes. +- [#3027](https://github.com/influxdb/influxdb/pull/3027): Enforce minimum retention policy duration of 1 hour. +- [#3030](https://github.com/influxdb/influxdb/pull/3030): Fix excessive logging of shard creation. +- [#3038](https://github.com/influxdb/influxdb/pull/3038): Don't check deleted shards for precreation. Thanks @vladlopes. +- [#3033](https://github.com/influxdb/influxdb/pull/3033): Add support for marshaling `uint64` in client. +- [#3090](https://github.com/influxdb/influxdb/pull/3090): Remove database from TSDB index on DROP DATABASE. +- [#2944](https://github.com/influxdb/influxdb/issues/2944): Don't require "WHERE time" when creating continuous queries. +- [#3075](https://github.com/influxdb/influxdb/pull/3075): GROUP BY correctly when different tags have same value. +- [#3078](https://github.com/influxdb/influxdb/pull/3078): Fix CLI panic on malformed INSERT. +- [#2102](https://github.com/influxdb/influxdb/issues/2102): Re-work Graphite input and metric processing +- [#2996](https://github.com/influxdb/influxdb/issues/2996): Graphite Input Parsing +- [#3136](https://github.com/influxdb/influxdb/pull/3136): Fix various issues with init.d script. Thanks @ miguelcnf. +- [#2996](https://github.com/influxdb/influxdb/issues/2996): Graphite Input Parsing +- [#3127](https://github.com/influxdb/influxdb/issues/3127): Trying to insert a number larger than the largest signed 64-bit number kills influxd +- [#3131](https://github.com/influxdb/influxdb/pull/3131): Copy batch tags to each point before marshalling +- [#3155](https://github.com/influxdb/influxdb/pull/3155): Instantiate UDP batcher before listening for UDP traffic, otherwise a panic may result. +- [#2678](https://github.com/influxdb/influxdb/issues/2678): Server allows tags with an empty string for the key and/or value +- [#3061](https://github.com/influxdb/influxdb/issues/3061): syntactically incorrect line protocol insert panics the database +- [#2608](https://github.com/influxdb/influxdb/issues/2608): drop measurement while writing points to that measurement has race condition that can panic +- [#3183](https://github.com/influxdb/influxdb/issues/3183): using line protocol measurement names cannot contain commas +- [#3193](https://github.com/influxdb/influxdb/pull/3193): Fix panic for SHOW STATS and in collectd +- [#3102](https://github.com/influxdb/influxdb/issues/3102): Add authentication cache +- [#3209](https://github.com/influxdb/influxdb/pull/3209): Dump Run() errors to stderr +- [#3217](https://github.com/influxdb/influxdb/pull/3217): Allow WAL partition flush delay to be configurable. + +## v0.9.0 [2015-06-11] + +### Bugfixes + +- [#2869](https://github.com/influxdb/influxdb/issues/2869): Adding field to existing measurement causes panic +- [#2849](https://github.com/influxdb/influxdb/issues/2849): RC32: Frequent write errors +- [#2700](https://github.com/influxdb/influxdb/issues/2700): Incorrect error message in database EncodeFields +- [#2897](https://github.com/influxdb/influxdb/pull/2897): Ensure target Graphite database exists +- [#2898](https://github.com/influxdb/influxdb/pull/2898): Ensure target openTSDB database exists +- [#2895](https://github.com/influxdb/influxdb/pull/2895): Use Graphite input defaults where necessary +- [#2900](https://github.com/influxdb/influxdb/pull/2900): Use openTSDB input defaults where necessary +- [#2886](https://github.com/influxdb/influxdb/issues/2886): Refactor backup & restore +- [#2804](https://github.com/influxdb/influxdb/pull/2804): BREAKING: change time literals to be single quoted in InfluxQL. Thanks @nvcook42! +- [#2906](https://github.com/influxdb/influxdb/pull/2906): Restrict replication factor to the cluster size +- [#2905](https://github.com/influxdb/influxdb/pull/2905): Restrict clusters to 3 peers +- [#2904](https://github.com/influxdb/influxdb/pull/2904): Re-enable server reporting. +- [#2917](https://github.com/influxdb/influxdb/pull/2917): Fix int64 field values. +- [#2920](https://github.com/influxdb/influxdb/issues/2920): Ensure collectd database exists + +## v0.9.0-rc33 [2015-06-09] + +### Bugfixes + +- [#2816](https://github.com/influxdb/influxdb/pull/2816): Enable UDP service. Thanks @renan- +- [#2824](https://github.com/influxdb/influxdb/pull/2824): Add missing call to WaitGroup.Done in execConn. Thanks @liyichao +- [#2823](https://github.com/influxdb/influxdb/pull/2823): Convert OpenTSDB to a service. +- [#2838](https://github.com/influxdb/influxdb/pull/2838): Set auto-created retention policy period to infinite. +- [#2829](https://github.com/influxdb/influxdb/pull/2829): Re-enable Graphite support as a new Service-style component. +- [#2814](https://github.com/influxdb/influxdb/issues/2814): Convert collectd to a service. +- [#2852](https://github.com/influxdb/influxdb/pull/2852): Don't panic when altering retention policies. Thanks for the report @huhongbo +- [#2857](https://github.com/influxdb/influxdb/issues/2857): Fix parsing commas in string field values. +- [#2833](https://github.com/influxdb/influxdb/pull/2833): Make the default config valid. +- [#2859](https://github.com/influxdb/influxdb/pull/2859): Fix panic on aggregate functions. +- [#2878](https://github.com/influxdb/influxdb/pull/2878): Re-enable shard precreation. +- [2865](https://github.com/influxdb/influxdb/pull/2865) -- Return an empty set of results if database does not exist in shard metadata. + +### Features +- [2858](https://github.com/influxdb/influxdb/pull/2858): Support setting openTSDB write consistency. + +## v0.9.0-rc32 [2015-06-07] + +### Release Notes + +This released introduced an updated write path and clustering design. The data format has also changed, so you'll need to wipe out your data to upgrade from RC31. There should be no other data changes before v0.9.0 is released. + +### Features +- [#1997](https://github.com/influxdb/influxdb/pull/1997): Update SELECT * to return tag values. +- [#2599](https://github.com/influxdb/influxdb/issues/2599): Add "epoch" URL param and return JSON time values as epoch instead of date strings. +- [#2682](https://github.com/influxdb/influxdb/issues/2682): Adding pr checklist to CONTRIBUTING.md +- [#2683](https://github.com/influxdb/influxdb/issues/2683): Add batching support to Graphite inputs. +- [#2687](https://github.com/influxdb/influxdb/issues/2687): Add batching support to Collectd inputs. +- [#2696](https://github.com/influxdb/influxdb/pull/2696): Add line protocol. This is now the preferred way to write data. +- [#2751](https://github.com/influxdb/influxdb/pull/2751): Add UDP input. UDP only supports the line protocol now. +- [#2684](https://github.com/influxdb/influxdb/pull/2684): Include client timeout configuration. Thanks @vladlopes! + +### Bugfixes +- [#2776](https://github.com/influxdb/influxdb/issues/2776): Re-implement retention policy enforcement. +- [#2635](https://github.com/influxdb/influxdb/issues/2635): Fix querying against boolean field in WHERE clause. +- [#2644](https://github.com/influxdb/influxdb/issues/2644): Make SHOW queries work with FROM //. +- [#2501](https://github.com/influxdb/influxdb/issues/2501): Name the FlagSet for the shell and add a version flag. Thanks @neonstalwart +- [#2647](https://github.com/influxdb/influxdb/issues/2647): Fixes typos in sample config file - thanks @claws! + +## v0.9.0-rc31 [2015-05-21] + +### Features +- [#1822](https://github.com/influxdb/influxdb/issues/1822): Wire up DERIVATIVE aggregate +- [#1477](https://github.com/influxdb/influxdb/issues/1477): Wire up non_negative_derivative function +- [#2557](https://github.com/influxdb/influxdb/issues/2557): Fix false positive error with `GROUP BY time` +- [#1891](https://github.com/influxdb/influxdb/issues/1891): Wire up COUNT DISTINCT aggregate +- [#1989](https://github.com/influxdb/influxdb/issues/1989): Implement `SELECT tagName FROM m` + +### Bugfixes +- [#2545](https://github.com/influxdb/influxdb/pull/2545): Use "value" as the field name for graphite input. Thanks @cannium. +- [#2558](https://github.com/influxdb/influxdb/pull/2558): Fix client response check - thanks @vladlopes! +- [#2566](https://github.com/influxdb/influxdb/pull/2566): Wait until each data write has been commited by the Raft cluster. +- [#2602](https://github.com/influxdb/influxdb/pull/2602): CLI execute command exits without cleaning up liner package. +- [#2610](https://github.com/influxdb/influxdb/pull/2610): Fix shard group creation +- [#2596](https://github.com/influxdb/influxdb/pull/2596): RC30: `panic: runtime error: index out of range` when insert data points. +- [#2592](https://github.com/influxdb/influxdb/pull/2592): Should return an error if user attempts to group by a field. +- [#2499](https://github.com/influxdb/influxdb/pull/2499): Issuing a select query with tag as a values causes panic. +- [#2612](https://github.com/influxdb/influxdb/pull/2612): Query planner should validate distinct is passed a field. +- [#2531](https://github.com/influxdb/influxdb/issues/2531): Fix select with 3 or more terms in where clause. +- [#2564](https://github.com/influxdb/influxdb/issues/2564): Change "name" to "measurement" in JSON for writes. + +## PRs +- [#2569](https://github.com/influxdb/influxdb/pull/2569): Add derivative functions +- [#2598](https://github.com/influxdb/influxdb/pull/2598): Implement tag support in SELECT statements +- [#2624](https://github.com/influxdb/influxdb/pull/2624): Remove references to SeriesID in `DROP SERIES` handlers. + +## v0.9.0-rc30 [2015-05-12] + +### Release Notes + +This release has a breaking API change for writes -- the field previously called `timestamp` has been renamed to `time`. + +### Features +- [#2254](https://github.com/influxdb/influxdb/pull/2254): Add Support for OpenTSDB HTTP interface. Thanks @tcolgate +- [#2525](https://github.com/influxdb/influxdb/pull/2525): Serve broker diagnostics over HTTP +- [#2186](https://github.com/influxdb/influxdb/pull/2186): The default status code for queries is now `200 OK` +- [#2298](https://github.com/influxdb/influxdb/pull/2298): Successful writes now return a status code of `204 No Content` - thanks @neonstalwart! +- [#2549](https://github.com/influxdb/influxdb/pull/2549): Raft election timeout to 5 seconds, so system is more forgiving of CPU loads. +- [#2568](https://github.com/influxdb/influxdb/pull/2568): Wire up SELECT DISTINCT. + +### Bugfixes +- [#2535](https://github.com/influxdb/influxdb/pull/2535): Return exit status 0 if influxd already running. Thanks @haim0n. +- [#2521](https://github.com/influxdb/influxdb/pull/2521): Don't truncate topic data until fully replicated. +- [#2509](https://github.com/influxdb/influxdb/pull/2509): Parse config file correctly during restore. Thanks @neonstalwart +- [#2536](https://github.com/influxdb/influxdb/issues/2532): Set leader ID on restart of single-node cluster. +- [#2448](https://github.com/influxdb/influxdb/pull/2448): Fix inconsistent data type - thanks @cannium! +- [#2108](https://github.com/influxdb/influxdb/issues/2108): Change `timestamp` to `time` - thanks @neonstalwart! +- [#2539](https://github.com/influxdb/influxdb/issues/2539): Add additional vote request logging. +- [#2541](https://github.com/influxdb/influxdb/issues/2541): Update messaging client connection index with every message. +- [#2542](https://github.com/influxdb/influxdb/issues/2542): Throw parser error for invalid aggregate without where time. +- [#2548](https://github.com/influxdb/influxdb/issues/2548): Return an error when numeric aggregate applied to non-numeric data. +- [#2487](https://github.com/influxdb/influxdb/issues/2487): Aggregate query with exact timestamp causes panic. Thanks @neonstalwart! +- [#2552](https://github.com/influxdb/influxdb/issues/2552): Run CQ that is actually passed into go-routine. +- [#2553](https://github.com/influxdb/influxdb/issues/2553): Fix race condition during CQ execution. +- [#2557](https://github.com/influxdb/influxdb/issues/2557): RC30 WHERE time filter Regression. + +## v0.9.0-rc29 [2015-05-05] + +### Features +- [#2410](https://github.com/influxdb/influxdb/pull/2410): If needed, brokers respond with data nodes for peer shard replication. +- [#2469](https://github.com/influxdb/influxdb/pull/2469): Reduce default max topic size from 1GB to 50MB. +- [#1824](https://github.com/influxdb/influxdb/pull/1824): Wire up MEDIAN aggregate. Thanks @neonstalwart! + +### Bugfixes +- [#2446](https://github.com/influxdb/influxdb/pull/2446): Correctly count number of queries executed. Thanks @neonstalwart +- [#2452](https://github.com/influxdb/influxdb/issues/2452): Fix panic with shard stats on multiple clusters +- [#2453](https://github.com/influxdb/influxdb/pull/2453): Do not require snapshot on Log.WriteEntriesTo(). +- [#2460](https://github.com/influxdb/influxdb/issues/2460): Collectd input should use "value" for fields values. Fixes 2412. Thanks @josh-padnick +- [#2465](https://github.com/influxdb/influxdb/pull/2465): HTTP response logging paniced with chunked requests. Thanks @Jackkoz +- [#2475](https://github.com/influxdb/influxdb/pull/2475): RLock server when checking if shards groups are required during write. +- [#2471](https://github.com/influxdb/influxdb/issues/2471): Function calls normalized to be lower case. Fixes percentile not working when called uppercase. Thanks @neonstalwart +- [#2281](https://github.com/influxdb/influxdb/issues/2281): Fix Bad Escape error when parsing regex + +## v0.9.0-rc28 [2015-04-27] + +### Features +- [#2410](https://github.com/influxdb/influxdb/pull/2410) Allow configuration of Raft timers +- [#2354](https://github.com/influxdb/influxdb/pull/2354) Wire up STDDEV. Thanks @neonstalwart! + +### Bugfixes +- [#2374](https://github.com/influxdb/influxdb/issues/2374): Two different panics during SELECT percentile +- [#2404](https://github.com/influxdb/influxdb/pull/2404): Mean and percentile function fixes +- [#2408](https://github.com/influxdb/influxdb/pull/2408): Fix snapshot 500 error +- [#1896](https://github.com/influxdb/influxdb/issues/1896): Excessive heartbeater logging of "connection refused" on cluster node stop +- [#2418](https://github.com/influxdb/influxdb/pull/2418): Fix raft node getting stuck in candidate state +- [#2415](https://github.com/influxdb/influxdb/pull/2415): Raft leader ID now set on election after failover. Thanks @xiaost +- [#2426](https://github.com/influxdb/influxdb/pull/2426): Fix race condition around listener address in openTSDB server. +- [#2426](https://github.com/influxdb/influxdb/pull/2426): Fix race condition around listener address in Graphite server. +- [#2429](https://github.com/influxdb/influxdb/pull/2429): Ensure no field value is null. +- [#2431](https://github.com/influxdb/influxdb/pull/2431): Always append shard path in diags. Thanks @marcosnils +- [#2441](https://github.com/influxdb/influxdb/pull/2441): Correctly release server RLock during "drop series". +- [#2445](https://github.com/influxdb/influxdb/pull/2445): Read locks and data race fixes + +## v0.9.0-rc27 [04-23-2015] + +### Features +- [#2398](https://github.com/influxdb/influxdb/pull/2398) Track more stats and report errors for shards. + +### Bugfixes +- [#2370](https://github.com/influxdb/influxdb/pull/2370): Fix data race in openTSDB endpoint. +- [#2371](https://github.com/influxdb/influxdb/pull/2371): Don't set client to nil when closing broker Fixes #2352 +- [#2372](https://github.com/influxdb/influxdb/pull/2372): Fix data race in graphite endpoint. +- [#2373](https://github.com/influxdb/influxdb/pull/2373): Actually allow HTTP logging to be controlled. +- [#2376](https://github.com/influxdb/influxdb/pull/2376): Encode all types of integers. Thanks @jtakkala. +- [#2376](https://github.com/influxdb/influxdb/pull/2376): Add shard path to existing diags value. Fix issue #2369. +- [#2386](https://github.com/influxdb/influxdb/pull/2386): Fix shard datanodes stats getting appended too many times +- [#2393](https://github.com/influxdb/influxdb/pull/2393): Fix default hostname for connecting to cluster. +- [#2390](https://github.com/influxdb/influxdb/pull/2390): Handle large sums when calculating means - thanks @neonstalwart! +- [#2391](https://github.com/influxdb/influxdb/pull/2391): Unable to write points through Go client when authentication enabled +- [#2400](https://github.com/influxdb/influxdb/pull/2400): Always send auth headers for client requests if present + +## v0.9.0-rc26 [04-21-2015] + +### Features +- [#2301](https://github.com/influxdb/influxdb/pull/2301): Distributed query load balancing and failover +- [#2336](https://github.com/influxdb/influxdb/pull/2336): Handle distributed queries when shards != data nodes +- [#2353](https://github.com/influxdb/influxdb/pull/2353): Distributed Query/Clustering Fixes + +### Bugfixes +- [#2297](https://github.com/influxdb/influxdb/pull/2297): create /var/run during startup. Thanks @neonstalwart. +- [#2312](https://github.com/influxdb/influxdb/pull/2312): Re-use httpclient for continuous queries +- [#2318](https://github.com/influxdb/influxdb/pull/2318): Remove pointless use of 'done' channel for collectd. +- [#2242](https://github.com/influxdb/influxdb/pull/2242): Distributed Query should balance requests +- [#2243](https://github.com/influxdb/influxdb/pull/2243): Use Limit Reader instead of fixed 1MB/1GB slice for DQ +- [#2190](https://github.com/influxdb/influxdb/pull/2190): Implement failover to other data nodes for distributed queries +- [#2324](https://github.com/influxdb/influxdb/issues/2324): Race in Broker.Close()/Broker.RunContinousQueryProcessing() +- [#2325](https://github.com/influxdb/influxdb/pull/2325): Cluster open fixes +- [#2326](https://github.com/influxdb/influxdb/pull/2326): Fix parse error in CREATE CONTINUOUS QUERY +- [#2300](https://github.com/influxdb/influxdb/pull/2300): Refactor integration tests. Properly close Graphite/OpenTSDB listeners. +- [#2338](https://github.com/influxdb/influxdb/pull/2338): Fix panic if tag key isn't double quoted when it should have been +- [#2340](https://github.com/influxdb/influxdb/pull/2340): Fix SHOW DIAGNOSTICS panic if any shard was non-local. +- [#2351](https://github.com/influxdb/influxdb/pull/2351): Fix data race by rlocking shard during diagnostics. +- [#2348](https://github.com/influxdb/influxdb/pull/2348): Data node fail to join cluster in 0.9.0rc25 +- [#2343](https://github.com/influxdb/influxdb/pull/2343): Node falls behind Metastore updates +- [#2334](https://github.com/influxdb/influxdb/pull/2334): Test Partial replication is very problematic +- [#2272](https://github.com/influxdb/influxdb/pull/2272): clustering: influxdb 0.9.0-rc23 panics when doing a GET with merge_metrics in a +- [#2350](https://github.com/influxdb/influxdb/pull/2350): Issue fix for :influxd -hostname localhost. +- [#2367](https://github.com/influxdb/influxdb/pull/2367): PR for issue #2350 - Always use localhost, not host name. + +## v0.9.0-rc25 [2015-04-15] + +### Bugfixes +- [#2282](https://github.com/influxdb/influxdb/pull/2282): Use "value" as field name for OpenTSDB input. +- [#2283](https://github.com/influxdb/influxdb/pull/2283): Fix bug when restarting an entire existing cluster. +- [#2293](https://github.com/influxdb/influxdb/pull/2293): Open cluster listener before starting broker. +- [#2287](https://github.com/influxdb/influxdb/pull/2287): Fix data race during SHOW RETENTION POLICIES. +- [#2288](https://github.com/influxdb/influxdb/pull/2288): Fix expression parsing bug. +- [#2294](https://github.com/influxdb/influxdb/pull/2294): Fix async response flushing (invalid chunked response error). + +## Features +- [#2276](https://github.com/influxdb/influxdb/pull/2276): Broker topic truncation. +- [#2292](https://github.com/influxdb/influxdb/pull/2292): Wire up drop CQ statement - thanks @neonstalwart! +- [#2290](https://github.com/influxdb/influxdb/pull/2290): Allow hostname argument to override default config - thanks @neonstalwart! +- [#2295](https://github.com/influxdb/influxdb/pull/2295): Use nil as default return value for MapCount - thanks @neonstalwart! +- [#2246](https://github.com/influxdb/influxdb/pull/2246): Allow HTTP logging to be controlled. + +## v0.9.0-rc24 [2015-04-13] + +### Bugfixes +- [#2255](https://github.com/influxdb/influxdb/pull/2255): Fix panic when changing default retention policy. +- [#2257](https://github.com/influxdb/influxdb/pull/2257): Add "snapshotting" pseudo state & log entry cache. +- [#2261](https://github.com/influxdb/influxdb/pull/2261): Support int64 value types. +- [#2191](https://github.com/influxdb/influxdb/pull/2191): Case-insensitive check for "fill" +- [#2274](https://github.com/influxdb/influxdb/pull/2274): Snapshot and HTTP API endpoints +- [#2265](https://github.com/influxdb/influxdb/pull/2265): Fix auth for CLI. + +## v0.9.0-rc23 [2015-04-11] + +### Features +- [#2202](https://github.com/influxdb/influxdb/pull/2202): Initial implementation of Distributed Queries +- [#2202](https://github.com/influxdb/influxdb/pull/2202): 64-bit Series IDs. INCOMPATIBLE WITH PREVIOUS DATASTORES. + +### Bugfixes +- [#2225](https://github.com/influxdb/influxdb/pull/2225): Make keywords completely case insensitive +- [#2228](https://github.com/influxdb/influxdb/pull/2228): Accept keyword default unquoted in ALTER RETENTION POLICY statement +- [#2236](https://github.com/influxdb/influxdb/pull/2236): Immediate term changes, fix stale write issue, net/http/pprof +- [#2213](https://github.com/influxdb/influxdb/pull/2213): Seed random number generator for election timeout. Thanks @cannium. + +## v0.9.0-rc22 [2015-04-09] + +### Features +- [#2214](https://github.com/influxdb/influxdb/pull/2214): Added the option to influx CLI to execute single command and exit. Thanks @n1tr0g + +### Bugfixes +- [#2223](https://github.com/influxdb/influxdb/pull/2223): Always notify term change on RequestVote + +## v0.9.0-rc21 [2015-04-09] + +### Features +- [#870](https://github.com/influxdb/influxdb/pull/870): Add support for OpenTSDB telnet input protocol. Thanks @tcolgate +- [#2180](https://github.com/influxdb/influxdb/pull/2180): Allow http write handler to decode gzipped body +- [#2175](https://github.com/influxdb/influxdb/pull/2175): Separate broker and data nodes +- [#2158](https://github.com/influxdb/influxdb/pull/2158): Allow user password to be changed. Thanks @n1tr0g +- [#2201](https://github.com/influxdb/influxdb/pull/2201): Bring back config join URLs +- [#2121](https://github.com/influxdb/influxdb/pull/2121): Parser refactor + +### Bugfixes +- [#2181](https://github.com/influxdb/influxdb/pull/2181): Fix panic on "SHOW DIAGNOSTICS". +- [#2170](https://github.com/influxdb/influxdb/pull/2170): Make sure queries on missing tags return 200 status. +- [#2197](https://github.com/influxdb/influxdb/pull/2197): Lock server during Open(). +- [#2200](https://github.com/influxdb/influxdb/pull/2200): Re-enable Continuous Queries. +- [#2203](https://github.com/influxdb/influxdb/pull/2203): Fix race condition on continuous queries. +- [#2217](https://github.com/influxdb/influxdb/pull/2217): Only revert to follower if new term is greater. +- [#2219](https://github.com/influxdb/influxdb/pull/2219): Persist term change to disk when candidate. Thanks @cannium + +## v0.9.0-rc20 [2015-04-04] + +### Features +- [#2128](https://github.com/influxdb/influxdb/pull/2128): Data node discovery from brokers +- [#2142](https://github.com/influxdb/influxdb/pull/2142): Support chunked queries +- [#2154](https://github.com/influxdb/influxdb/pull/2154): Node redirection +- [#2168](https://github.com/influxdb/influxdb/pull/2168): Return raft term from vote, add term logging + +### Bugfixes +- [#2147](https://github.com/influxdb/influxdb/pull/2147): Set Go Max procs in a better location +- [#2137](https://github.com/influxdb/influxdb/pull/2137): Refactor `results` to `response`. Breaking Go Client change. +- [#2151](https://github.com/influxdb/influxdb/pull/2151): Ignore replay commands on the metastore. +- [#2152](https://github.com/influxdb/influxdb/issues/2152): Influxd process with stats enabled crashing with 'Unsuported protocol scheme for ""' +- [#2156](https://github.com/influxdb/influxdb/pull/2156): Propagate error when resolving UDP address in Graphite UDP server. +- [#2163](https://github.com/influxdb/influxdb/pull/2163): Fix up paths for default data and run storage. +- [#2164](https://github.com/influxdb/influxdb/pull/2164): Append STDOUT/STDERR in initscript. +- [#2165](https://github.com/influxdb/influxdb/pull/2165): Better name for config section for stats and diags. +- [#2165](https://github.com/influxdb/influxdb/pull/2165): Monitoring database and retention policy are not configurable. +- [#2167](https://github.com/influxdb/influxdb/pull/2167): Add broker log recovery. +- [#2166](https://github.com/influxdb/influxdb/pull/2166): Don't panic if presented with a field of unknown type. +- [#2149](https://github.com/influxdb/influxdb/pull/2149): Fix unit tests for win32 when directory doesn't exist. +- [#2150](https://github.com/influxdb/influxdb/pull/2150): Fix unit tests for win32 when a connection is refused. + +## v0.9.0-rc19 [2015-04-01] + +### Features +- [#2143](https://github.com/influxdb/influxdb/pull/2143): Add raft term logging. + +### Bugfixes +- [#2145](https://github.com/influxdb/influxdb/pull/2145): Encode toml durations correctly which fixes default configuration generation `influxd config`. + +## v0.9.0-rc18 [2015-03-31] + +### Bugfixes +- [#2100](https://github.com/influxdb/influxdb/pull/2100): Use channel to synchronize collectd shutdown. +- [#2100](https://github.com/influxdb/influxdb/pull/2100): Synchronize access to shard index. +- [#2131](https://github.com/influxdb/influxdb/pull/2131): Optimize marshalTags(). +- [#2130](https://github.com/influxdb/influxdb/pull/2130): Make fewer calls to marshalTags(). +- [#2105](https://github.com/influxdb/influxdb/pull/2105): Support != for tag values. Fix issue #2097, thanks to @smonkewitz for bug report. +- [#2105](https://github.com/influxdb/influxdb/pull/2105): Support !~ tags values. +- [#2138](https://github.com/influxdb/influxdb/pull/2136): Use map for marshaledTags cache. + +## v0.9.0-rc17 [2015-03-29] + +### Features +- [#2076](https://github.com/influxdb/influxdb/pull/2076): Separate stdout and stderr output in init.d script +- [#2091](https://github.com/influxdb/influxdb/pull/2091): Support disabling snapshot endpoint. +- [#2081](https://github.com/influxdb/influxdb/pull/2081): Support writing diagnostic data into the internal database. +- [#2095](https://github.com/influxdb/influxdb/pull/2095): Improved InfluxDB client docs. Thanks @derailed + +### Bugfixes +- [#2093](https://github.com/influxdb/influxdb/pull/2093): Point precision not marshalled correctly. Thanks @derailed +- [#2084](https://github.com/influxdb/influxdb/pull/2084): Allowing leading underscores in identifiers. +- [#2080](https://github.com/influxdb/influxdb/pull/2080): Graphite logs in seconds, not milliseconds. +- [#2101](https://github.com/influxdb/influxdb/pull/2101): SHOW DATABASES should name returned series "databases". +- [#2104](https://github.com/influxdb/influxdb/pull/2104): Include NEQ when calculating field filters. +- [#2112](https://github.com/influxdb/influxdb/pull/2112): Set GOMAXPROCS on startup. This may have been causing extra leader elections, which would cause a number of other bugs or instability. +- [#2111](https://github.com/influxdb/influxdb/pull/2111) and [#2025](https://github.com/influxdb/influxdb/issues/2025): Raft stability fixes. Non-contiguous log error and others. +- [#2114](https://github.com/influxdb/influxdb/pull/2114): Correctly start influxd on platforms without start-stop-daemon. + +## v0.9.0-rc16 [2015-03-24] + +### Features +- [#2058](https://github.com/influxdb/influxdb/pull/2058): Track number of queries executed in stats. +- [#2059](https://github.com/influxdb/influxdb/pull/2059): Retention policies sorted by name on return to client. +- [#2061](https://github.com/influxdb/influxdb/pull/2061): Implement SHOW DIAGNOSTICS. +- [#2064](https://github.com/influxdb/influxdb/pull/2064): Allow init.d script to return influxd version. +- [#2053](https://github.com/influxdb/influxdb/pull/2053): Implment backup and restore. +- [#1631](https://github.com/influxdb/influxdb/pull/1631): Wire up DROP CONTINUOUS QUERY. + +### Bugfixes +- [#2037](https://github.com/influxdb/influxdb/pull/2037): Don't check 'configExists' at Run() level. +- [#2039](https://github.com/influxdb/influxdb/pull/2039): Don't panic if getting current user fails. +- [#2034](https://github.com/influxdb/influxdb/pull/2034): GROUP BY should require an aggregate. +- [#2040](https://github.com/influxdb/influxdb/pull/2040): Add missing top-level help for config command. +- [#2057](https://github.com/influxdb/influxdb/pull/2057): Move racy "in order" test to integration test suite. +- [#2060](https://github.com/influxdb/influxdb/pull/2060): Reload server shard map on restart. +- [#2068](https://github.com/influxdb/influxdb/pull/2068): Fix misspelled JSON field. +- [#2067](https://github.com/influxdb/influxdb/pull/2067): Fixed issue where some queries didn't properly pull back data (introduced in RC15). Fixing intervals for GROUP BY. + +## v0.9.0-rc15 [2015-03-19] + +### Features +- [#2000](https://github.com/influxdb/influxdb/pull/2000): Log broker path when broker fails to start. Thanks @gst. +- [#2007](https://github.com/influxdb/influxdb/pull/2007): Track shard-level stats. + +### Bugfixes +- [#2001](https://github.com/influxdb/influxdb/pull/2001): Ensure measurement not found returns status code 200. +- [#1985](https://github.com/influxdb/influxdb/pull/1985): Set content-type JSON header before actually writing header. Thanks @dstrek. +- [#2003](https://github.com/influxdb/influxdb/pull/2003): Set timestamp when writing monitoring stats. +- [#2004](https://github.com/influxdb/influxdb/pull/2004): Limit group by to MaxGroupByPoints (currently 100,000). +- [#2016](https://github.com/influxdb/influxdb/pull/2016): Fixing bucket alignment for group by. Thanks @jnutzmann +- [#2021](https://github.com/influxdb/influxdb/pull/2021): Remove unnecessary formatting from log message. Thanks @simonkern + + +## v0.9.0-rc14 [2015-03-18] + +### Bugfixes +- [#1999](https://github.com/influxdb/influxdb/pull/1999): Return status code 200 for measurement not found errors on show series. + +## v0.9.0-rc13 [2015-03-17] + +### Features +- [#1974](https://github.com/influxdb/influxdb/pull/1974): Add time taken for request to the http server logs. + +### Bugfixes +- [#1971](https://github.com/influxdb/influxdb/pull/1971): Fix leader id initialization. +- [#1975](https://github.com/influxdb/influxdb/pull/1975): Require `q` parameter for query endpoint. +- [#1969](https://github.com/influxdb/influxdb/pull/1969): Print loaded config. +- [#1987](https://github.com/influxdb/influxdb/pull/1987): Fix config print startup statement for when no config is provided. +- [#1990](https://github.com/influxdb/influxdb/pull/1990): Drop measurement was taking too long due to transactions. + +## v0.9.0-rc12 [2015-03-15] + +### Bugfixes +- [#1942](https://github.com/influxdb/influxdb/pull/1942): Sort wildcard names. +- [#1957](https://github.com/influxdb/influxdb/pull/1957): Graphite numbers are always float64. +- [#1955](https://github.com/influxdb/influxdb/pull/1955): Prohibit creation of databases with no name. Thanks @dullgiulio +- [#1952](https://github.com/influxdb/influxdb/pull/1952): Handle delete statement with an error. Thanks again to @dullgiulio + +### Features +- [#1935](https://github.com/influxdb/influxdb/pull/1935): Implement stateless broker for Raft. +- [#1936](https://github.com/influxdb/influxdb/pull/1936): Implement "SHOW STATS" and self-monitoring + +### Features +- [#1909](https://github.com/influxdb/influxdb/pull/1909): Implement a dump command. + +## v0.9.0-rc11 [2015-03-13] + +### Bugfixes +- [#1917](https://github.com/influxdb/influxdb/pull/1902): Creating Infinite Retention Policy Failed. +- [#1758](https://github.com/influxdb/influxdb/pull/1758): Add Graphite Integration Test. +- [#1929](https://github.com/influxdb/influxdb/pull/1929): Default Retention Policy incorrectly auto created. +- [#1930](https://github.com/influxdb/influxdb/pull/1930): Auto create database for graphite if not specified. +- [#1908](https://github.com/influxdb/influxdb/pull/1908): Cosmetic CLI output fixes. +- [#1931](https://github.com/influxdb/influxdb/pull/1931): Add default column to SHOW RETENTION POLICIES. +- [#1937](https://github.com/influxdb/influxdb/pull/1937): OFFSET should be allowed to be 0. + +### Features +- [#1902](https://github.com/influxdb/influxdb/pull/1902): Enforce retention policies to have a minimum duration. +- [#1906](https://github.com/influxdb/influxdb/pull/1906): Add show servers to query language. +- [#1925](https://github.com/influxdb/influxdb/pull/1925): Add `fill(none)`, `fill(previous)`, and `fill()` to queries. + +## v0.9.0-rc10 [2015-03-09] + +### Bugfixes +- [#1867](https://github.com/influxdb/influxdb/pull/1867): Fix race accessing topic replicas map +- [#1864](https://github.com/influxdb/influxdb/pull/1864): fix race in startStateLoop +- [#1753](https://github.com/influxdb/influxdb/pull/1874): Do Not Panic on Missing Dirs +- [#1877](https://github.com/influxdb/influxdb/pull/1877): Broker clients track broker leader +- [#1862](https://github.com/influxdb/influxdb/pull/1862): Fix memory leak in `httpd.serveWait`. Thanks @mountkin +- [#1883](https://github.com/influxdb/influxdb/pull/1883): RLock server during retention policy enforcement. Thanks @grisha +- [#1868](https://github.com/influxdb/influxdb/pull/1868): Use `BatchPoints` for `client.Write` method. Thanks @vladlopes, @georgmu, @d2g, @evanphx, @akolosov. +- [#1881](https://github.com/influxdb/influxdb/pull/1881): Update documentation for `client` package. Misc library tweaks. +- Fix queries with multiple where clauses on tags, times and fields. Fix queries that have where clauses on fields not in the select + +### Features +- [#1875](https://github.com/influxdb/influxdb/pull/1875): Support trace logging of Raft. +- [#1895](https://github.com/influxdb/influxdb/pull/1895): Auto-create a retention policy when a database is created. +- [#1897](https://github.com/influxdb/influxdb/pull/1897): Pre-create shard groups. +- [#1900](https://github.com/influxdb/influxdb/pull/1900): Change `LIMIT` to `SLIMIT` and implement `LIMIT` and `OFFSET` + +## v0.9.0-rc9 [2015-03-06] + +### Bugfixes +- [#1872](https://github.com/influxdb/influxdb/pull/1872): Fix "stale term" errors with raft + +## v0.9.0-rc8 [2015-03-05] + +### Bugfixes +- [#1836](https://github.com/influxdb/influxdb/pull/1836): Store each parsed shell command in history file. +- [#1789](https://github.com/influxdb/influxdb/pull/1789): add --config-files option to fpm command. Thanks @kylezh +- [#1859](https://github.com/influxdb/influxdb/pull/1859): Queries with a `GROUP BY *` clause were returning a 500 if done against a measurement that didn't exist + +### Features +- [#1755](https://github.com/influxdb/influxdb/pull/1848): Support JSON data ingest over UDP +- [#1857](https://github.com/influxdb/influxdb/pull/1857): Support retention policies with infinite duration +- [#1858](https://github.com/influxdb/influxdb/pull/1858): Enable detailed tracing of write path + +## v0.9.0-rc7 [2015-03-02] + +### Features +- [#1813](https://github.com/influxdb/influxdb/pull/1813): Queries for missing measurements or fields now return a 200 with an error message in the series JSON. +- [#1826](https://github.com/influxdb/influxdb/pull/1826), [#1827](https://github.com/influxdb/influxdb/pull/1827): Fixed queries with `WHERE` clauses against fields. + +### Bugfixes + +- [#1744](https://github.com/influxdb/influxdb/pull/1744): Allow retention policies to be modified without specifying replication factor. Thanks @kylezh +- [#1809](https://github.com/influxdb/influxdb/pull/1809): Packaging post-install script unconditionally removes init.d symlink. Thanks @sineos + +## v0.9.0-rc6 [2015-02-27] + +### Bugfixes + +- [#1780](https://github.com/influxdb/influxdb/pull/1780): Malformed identifiers get through the parser +- [#1775](https://github.com/influxdb/influxdb/pull/1775): Panic "index out of range" on some queries +- [#1744](https://github.com/influxdb/influxdb/pull/1744): Select shard groups which completely encompass time range. Thanks @kylezh. + +## v0.9.0-rc5 [2015-02-27] + +### Bugfixes + +- [#1752](https://github.com/influxdb/influxdb/pull/1752): remove debug log output from collectd. +- [#1720](https://github.com/influxdb/influxdb/pull/1720): Parse Series IDs as unsigned 32-bits. +- [#1767](https://github.com/influxdb/influxdb/pull/1767): Drop Series was failing across shards. Issue #1761. +- [#1773](https://github.com/influxdb/influxdb/pull/1773): Fix bug when merging series together that have unequal number of points in a group by interval +- [#1771](https://github.com/influxdb/influxdb/pull/1771): Make `SHOW SERIES` return IDs and support `LIMIT` and `OFFSET` + +### Features + +- [#1698](https://github.com/influxdb/influxdb/pull/1698): Wire up DROP MEASUREMENT + +## v0.9.0-rc4 [2015-02-24] + +### Bugfixes + +- Fix authentication issue with continuous queries +- Print version in the log on startup + +## v0.9.0-rc3 [2015-02-23] + +### Features + +- [#1659](https://github.com/influxdb/influxdb/pull/1659): WHERE against regexes: `WHERE =~ '.*asdf' +- [#1580](https://github.com/influxdb/influxdb/pull/1580): Add support for fields with bool, int, or string data types +- [#1687](https://github.com/influxdb/influxdb/pull/1687): Change `Rows` to `Series` in results output. BREAKING API CHANGE +- [#1629](https://github.com/influxdb/influxdb/pull/1629): Add support for `DROP SERIES` queries +- [#1632](https://github.com/influxdb/influxdb/pull/1632): Add support for `GROUP BY *` to return all series within a measurement +- [#1689](https://github.com/influxdb/influxdb/pull/1689): Change `SHOW TAG VALUES WITH KEY="foo"` to use the key name in the result. BREAKING API CHANGE +- [#1699](https://github.com/influxdb/influxdb/pull/1699): Add CPU and memory profiling options to daemon +- [#1672](https://github.com/influxdb/influxdb/pull/1672): Add index tracking to metastore. Makes downed node recovery actually work +- [#1591](https://github.com/influxdb/influxdb/pull/1591): Add `spread` aggregate function +- [#1576](https://github.com/influxdb/influxdb/pull/1576): Add `first` and `last` aggregate functions +- [#1573](https://github.com/influxdb/influxdb/pull/1573): Add `stddev` aggregate function +- [#1565](https://github.com/influxdb/influxdb/pull/1565): Add the admin interface back into the server and update for new API +- [#1562](https://github.com/influxdb/influxdb/pull/1562): Enforce retention policies +- [#1700](https://github.com/influxdb/influxdb/pull/1700): Change `Values` to `Fields` on writes. BREAKING API CHANGE +- [#1706](https://github.com/influxdb/influxdb/pull/1706): Add support for `LIMIT` and `OFFSET`, which work on the number of series returned in a query. To limit the number of data points use a `WHERE time` clause + +### Bugfixes + +- [#1636](https://github.com/influxdb/influxdb/issues/1636): Don't store number of fields in raw data. THIS IS A BREAKING DATA CHANGE. YOU MUST START WITH A FRESH DATABASE +- [#1701](https://github.com/influxdb/influxdb/pull/1701), [#1667](https://github.com/influxdb/influxdb/pull/1667), [#1663](https://github.com/influxdb/influxdb/pull/1663), [#1615](https://github.com/influxdb/influxdb/pull/1615): Raft fixes +- [#1644](https://github.com/influxdb/influxdb/pull/1644): Add batching support for significantly improved write performance +- [#1704](https://github.com/influxdb/influxdb/pull/1704): Fix queries that pull back raw data (i.e. ones without aggregate functions) +- [#1718](https://github.com/influxdb/influxdb/pull/1718): Return an error on write if any of the points are don't have at least one field +- [#1806](https://github.com/influxdb/influxdb/pull/1806): Fix regex parsing. Change regex syntax to use / delimiters. + + +## v0.9.0-rc1,2 [no public release] + +### Features + +- Support for tags added +- New queries for showing measurement names, tag keys, and tag values +- Renamed shard spaces to retention policies +- Deprecated matching against regex in favor of explicit writing and querying on retention policies +- Pure Go InfluxQL parser +- Switch to BoltDB as underlying datastore +- BoltDB backed metastore to store schema information +- Updated HTTP API to only have two endpoints `/query` and `/write` +- Added all administrative functions to the query language +- Change cluster architecture to have brokers and data nodes +- Switch to streaming Raft implementation +- In memory inverted index of the tag data +- Pure Go implementation! + +## v0.8.6 [2014-11-15] + +### Features + +- [Issue #973](https://github.com/influxdb/influxdb/issues/973). Support + joining using a regex or list of time series +- [Issue #1068](https://github.com/influxdb/influxdb/issues/1068). Print + the processor chain when the query is started + +### Bugfixes + +- [Issue #584](https://github.com/influxdb/influxdb/issues/584). Don't + panic if the process died while initializing +- [Issue #663](https://github.com/influxdb/influxdb/issues/663). Make + sure all sub servies are closed when are stopping InfluxDB +- [Issue #671](https://github.com/influxdb/influxdb/issues/671). Fix + the Makefile package target for Mac OSX +- [Issue #800](https://github.com/influxdb/influxdb/issues/800). Use + su instead of sudo in the init script. This fixes the startup problem + on RHEL 6. +- [Issue #925](https://github.com/influxdb/influxdb/issues/925). Don't + generate invalid query strings for single point queries +- [Issue #943](https://github.com/influxdb/influxdb/issues/943). Don't + take two snapshots at the same time +- [Issue #947](https://github.com/influxdb/influxdb/issues/947). Exit + nicely if the daemon doesn't have permission to write to the log. +- [Issue #959](https://github.com/influxdb/influxdb/issues/959). Stop using + closed connections in the protobuf client. +- [Issue #978](https://github.com/influxdb/influxdb/issues/978). Check + for valgrind and mercurial in the configure script +- [Issue #996](https://github.com/influxdb/influxdb/issues/996). Fill should + fill the time range even if no points exists in the given time range +- [Issue #1008](https://github.com/influxdb/influxdb/issues/1008). Return + an appropriate exit status code depending on whether the process exits + due to an error or exits gracefully. +- [Issue #1024](https://github.com/influxdb/influxdb/issues/1024). Hitting + open files limit causes influxdb to create shards in loop. +- [Issue #1069](https://github.com/influxdb/influxdb/issues/1069). Fix + deprecated interface endpoint in Admin UI. +- [Issue #1076](https://github.com/influxdb/influxdb/issues/1076). Fix + the timestamps of data points written by the collectd plugin. (Thanks, + @renchap for reporting this bug) +- [Issue #1078](https://github.com/influxdb/influxdb/issues/1078). Make sure + we don't resurrect shard directories for shards that have already expired +- [Issue #1085](https://github.com/influxdb/influxdb/issues/1085). Set + the connection string of the local raft node +- [Issue #1092](https://github.com/influxdb/influxdb/issues/1093). Set + the connection string of the local node in the raft snapshot. +- [Issue #1100](https://github.com/influxdb/influxdb/issues/1100). Removing + a non-existent shard space causes the cluster to panic. +- [Issue #1113](https://github.com/influxdb/influxdb/issues/1113). A nil + engine.ProcessorChain causes a panic. + +## v0.8.5 [2014-10-27] + +### Features + +- [Issue #1055](https://github.com/influxdb/influxdb/issues/1055). Allow + graphite and collectd input plugins to have separate binding address + +### Bugfixes + +- [Issue #1058](https://github.com/influxdb/influxdb/issues/1058). Use + the query language instead of the continuous query endpoints that + were removed in 0.8.4 +- [Issue #1022](https://github.com/influxdb/influxdb/issues/1022). Return + an +Inf or NaN instead of panicing when we encounter a divide by zero +- [Issue #821](https://github.com/influxdb/influxdb/issues/821). Don't + scan through points when we hit the limit +- [Issue #1051](https://github.com/influxdb/influxdb/issues/1051). Fix + timestamps when the collectd is used and low resolution timestamps + is set. + +## v0.8.4 [2014-10-24] + +### Bugfixes + +- Remove the continuous query api endpoints since the query language + has all the features needed to list and delete continuous queries. +- [Issue #778](https://github.com/influxdb/influxdb/issues/778). Selecting + from a non-existent series should give a better error message indicating + that the series doesn't exist +- [Issue #988](https://github.com/influxdb/influxdb/issues/988). Check + the arguments of `top()` and `bottom()` +- [Issue #1021](https://github.com/influxdb/influxdb/issues/1021). Make + redirecting to standard output and standard error optional instead of + going to `/dev/null`. This can now be configured by setting `$STDOUT` + in `/etc/default/influxdb` +- [Issue #985](https://github.com/influxdb/influxdb/issues/985). Make + sure we drop a shard only when there's no one using it. Otherwise, the + shard can be closed when another goroutine is writing to it which will + cause random errors and possibly corruption of the database. + +### Features + +- [Issue #1047](https://github.com/influxdb/influxdb/issues/1047). Allow + merge() to take a list of series (as opposed to a regex in #72) + +## v0.8.4-rc.1 [2014-10-21] + +### Bugfixes + +- [Issue #1040](https://github.com/influxdb/influxdb/issues/1040). Revert + to older raft snapshot if the latest one is corrupted +- [Issue #1004](https://github.com/influxdb/influxdb/issues/1004). Querying + for data outside of existing shards returns an empty response instead of + throwing a `Couldn't lookup columns` error +- [Issue #1020](https://github.com/influxdb/influxdb/issues/1020). Change + init script exit codes to conform to the lsb standards. (Thanks, @spuder) +- [Issue #1011](https://github.com/influxdb/influxdb/issues/1011). Fix + the tarball for homebrew so that rocksdb is included and the directory + structure is clean +- [Issue #1007](https://github.com/influxdb/influxdb/issues/1007). Fix + the content type when an error occurs and the client requests + compression. +- [Issue #916](https://github.com/influxdb/influxdb/issues/916). Set + the ulimit in the init script with a way to override the limit +- [Issue #742](https://github.com/influxdb/influxdb/issues/742). Fix + rocksdb for Mac OSX +- [Issue #387](https://github.com/influxdb/influxdb/issues/387). Aggregations + with group by time(1w), time(1m) and time(1y) (for week, month and + year respectively) will cause the start time and end time of the bucket + to fall on the logical boundaries of the week, month or year. +- [Issue #334](https://github.com/influxdb/influxdb/issues/334). Derivative + for queries with group by time() and fill(), will take the difference + between the first value in the bucket and the first value of the next + bucket. +- [Issue #972](https://github.com/influxdb/influxdb/issues/972). Don't + assign duplicate server ids + +### Features + +- [Issue #722](https://github.com/influxdb/influxdb/issues/722). Add + an install target to the Makefile +- [Issue #1032](https://github.com/influxdb/influxdb/issues/1032). Include + the admin ui static assets in the binary +- [Issue #1019](https://github.com/influxdb/influxdb/issues/1019). Upgrade + to rocksdb 3.5.1 +- [Issue #992](https://github.com/influxdb/influxdb/issues/992). Add + an input plugin for collectd. (Thanks, @kimor79) +- [Issue #72](https://github.com/influxdb/influxdb/issues/72). Support merge + for multiple series using regex syntax + +## v0.8.3 [2014-09-24] + +### Bugfixes + +- [Issue #885](https://github.com/influxdb/influxdb/issues/885). Multiple + queries separated by semicolons work as expected. Queries are process + sequentially +- [Issue #652](https://github.com/influxdb/influxdb/issues/652). Return an + error if an invalid column is used in the where clause +- [Issue #794](https://github.com/influxdb/influxdb/issues/794). Fix case + insensitive regex matching +- [Issue #853](https://github.com/influxdb/influxdb/issues/853). Move + cluster config from raft to API. +- [Issue #714](https://github.com/influxdb/influxdb/issues/714). Don't + panic on invalid boolean operators. +- [Issue #843](https://github.com/influxdb/influxdb/issues/843). Prevent blank database names +- [Issue #780](https://github.com/influxdb/influxdb/issues/780). Fix + fill() for all aggregators +- [Issue #923](https://github.com/influxdb/influxdb/issues/923). Enclose + table names in double quotes in the result of GetQueryString() +- [Issue #923](https://github.com/influxdb/influxdb/issues/923). Enclose + table names in double quotes in the result of GetQueryString() +- [Issue #967](https://github.com/influxdb/influxdb/issues/967). Return an + error if the storage engine can't be created +- [Issue #954](https://github.com/influxdb/influxdb/issues/954). Don't automatically + create shards which was causing too many shards to be created when used with + grafana +- [Issue #939](https://github.com/influxdb/influxdb/issues/939). Aggregation should + ignore null values and invalid values, e.g. strings with mean(). +- [Issue #964](https://github.com/influxdb/influxdb/issues/964). Parse + big int in queries properly. + +## v0.8.2 [2014-09-05] + +### Bugfixes + +- [Issue #886](https://github.com/influxdb/influxdb/issues/886). Update shard space to not set defaults + +- [Issue #867](https://github.com/influxdb/influxdb/issues/867). Add option to return shard space mappings in list series + +### Bugfixes + +- [Issue #652](https://github.com/influxdb/influxdb/issues/652). Return + a meaningful error if an invalid column is used in where clause + after joining multiple series + +## v0.8.2 [2014-09-08] + +### Features + +- Added API endpoint to update shard space definitions + +### Bugfixes + +- [Issue #886](https://github.com/influxdb/influxdb/issues/886). Shard space regexes reset after restart of InfluxDB + +## v0.8.1 [2014-09-03] + +- [Issue #896](https://github.com/influxdb/influxdb/issues/896). Allow logging to syslog. Thanks @malthe + +### Bugfixes + +- [Issue #868](https://github.com/influxdb/influxdb/issues/868). Don't panic when upgrading a snapshot from 0.7.x +- [Issue #887](https://github.com/influxdb/influxdb/issues/887). The first continuous query shouldn't trigger backfill if it had backfill disabled +- [Issue #674](https://github.com/influxdb/influxdb/issues/674). Graceful exit when config file is invalid. (Thanks, @DavidBord) +- [Issue #857](https://github.com/influxdb/influxdb/issues/857). More informative list servers api. (Thanks, @oliveagle) + +## v0.8.0 [2014-08-22] + +### Features + +- [Issue #850](https://github.com/influxdb/influxdb/issues/850). Makes the server listing more informative + +### Bugfixes + +- [Issue #779](https://github.com/influxdb/influxdb/issues/779). Deleting expired shards isn't thread safe. +- [Issue #860](https://github.com/influxdb/influxdb/issues/860). Load database config should validate shard spaces. +- [Issue #862](https://github.com/influxdb/influxdb/issues/862). Data migrator should have option to set delay time. + +## v0.8.0-rc.5 [2014-08-15] + +### Features + +- [Issue #376](https://github.com/influxdb/influxdb/issues/376). List series should support regex filtering +- [Issue #745](https://github.com/influxdb/influxdb/issues/745). Add continuous queries to the database config +- [Issue #746](https://github.com/influxdb/influxdb/issues/746). Add data migration tool for 0.8.0 + +### Bugfixes + +- [Issue #426](https://github.com/influxdb/influxdb/issues/426). Fill should fill the entire time range that is requested +- [Issue #740](https://github.com/influxdb/influxdb/issues/740). Don't emit non existent fields when joining series with different fields +- [Issue #744](https://github.com/influxdb/influxdb/issues/744). Admin site should have all assets locally +- [Issue #767](https://github.com/influxdb/influxdb/issues/768). Remove shards whenever they expire +- [Issue #781](https://github.com/influxdb/influxdb/issues/781). Don't emit non existent fields when joining series with different fields +- [Issue #791](https://github.com/influxdb/influxdb/issues/791). Move database config loader to be an API endpoint +- [Issue #809](https://github.com/influxdb/influxdb/issues/809). Migration path from 0.7 -> 0.8 +- [Issue #811](https://github.com/influxdb/influxdb/issues/811). Gogoprotobuf removed `ErrWrongType`, which is depended on by Raft +- [Issue #820](https://github.com/influxdb/influxdb/issues/820). Query non-local shard with time range to avoid getting back points not in time range +- [Issue #827](https://github.com/influxdb/influxdb/issues/827). Don't leak file descriptors in the WAL +- [Issue #830](https://github.com/influxdb/influxdb/issues/830). List series should return series in lexicographic sorted order +- [Issue #831](https://github.com/influxdb/influxdb/issues/831). Move create shard space to be db specific + +## v0.8.0-rc.4 [2014-07-29] + +### Bugfixes + +- [Issue #774](https://github.com/influxdb/influxdb/issues/774). Don't try to parse "inf" shard retention policy +- [Issue #769](https://github.com/influxdb/influxdb/issues/769). Use retention duration when determining expired shards. (Thanks, @shugo) +- [Issue #736](https://github.com/influxdb/influxdb/issues/736). Only db admins should be able to drop a series +- [Issue #713](https://github.com/influxdb/influxdb/issues/713). Null should be a valid fill value +- [Issue #644](https://github.com/influxdb/influxdb/issues/644). Graphite api should write data in batches to the coordinator +- [Issue #740](https://github.com/influxdb/influxdb/issues/740). Panic when distinct fields are selected from an inner join +- [Issue #781](https://github.com/influxdb/influxdb/issues/781). Panic when distinct fields are added after an inner join + +## v0.8.0-rc.3 [2014-07-21] + +### Bugfixes + +- [Issue #752](https://github.com/influxdb/influxdb/issues/752). `./configure` should use goroot to find gofmt +- [Issue #758](https://github.com/influxdb/influxdb/issues/758). Clarify the reason behind graphite input plugin not starting. (Thanks, @otoolep) +- [Issue #759](https://github.com/influxdb/influxdb/issues/759). Don't revert the regex in the shard space. (Thanks, @shugo) +- [Issue #760](https://github.com/influxdb/influxdb/issues/760). Removing a server should remove it from the shard server ids. (Thanks, @shugo) +- [Issue #772](https://github.com/influxdb/influxdb/issues/772). Add sentinel values to all db. This caused the last key in the db to not be fetched properly. + + +## v0.8.0-rc.2 [2014-07-15] + +- This release is to fix a build error in rc1 which caused rocksdb to not be available +- Bump up the `max-open-files` option to 1000 on all storage engines +- Lower the `write-buffer-size` to 1000 + +## v0.8.0-rc.1 [2014-07-15] + +### Features + +- [Issue #643](https://github.com/influxdb/influxdb/issues/643). Support pretty print json. (Thanks, @otoolep) +- [Issue #641](https://github.com/influxdb/influxdb/issues/641). Support multiple storage engines +- [Issue #665](https://github.com/influxdb/influxdb/issues/665). Make build tmp directory configurable in the make file. (Thanks, @dgnorton) +- [Issue #667](https://github.com/influxdb/influxdb/issues/667). Enable compression on all GET requests and when writing data +- [Issue #648](https://github.com/influxdb/influxdb/issues/648). Return permissions when listing db users. (Thanks, @nicolai86) +- [Issue #682](https://github.com/influxdb/influxdb/issues/682). Allow continuous queries to run without backfill (Thanks, @dhammika) +- [Issue #689](https://github.com/influxdb/influxdb/issues/689). **REQUIRES DATA MIGRATION** Move metadata into raft +- [Issue #255](https://github.com/influxdb/influxdb/issues/255). Support millisecond precision using `ms` suffix +- [Issue #95](https://github.com/influxdb/influxdb/issues/95). Drop database should not be synchronous +- [Issue #571](https://github.com/influxdb/influxdb/issues/571). Add support for arbitrary number of shard spaces and retention policies +- Default storage engine changed to RocksDB + +### Bugfixes + +- [Issue #651](https://github.com/influxdb/influxdb/issues/651). Change permissions of symlink which fix some installation issues. (Thanks, @Dieterbe) +- [Issue #670](https://github.com/influxdb/influxdb/issues/670). Don't warn on missing influxdb user on fresh installs +- [Issue #676](https://github.com/influxdb/influxdb/issues/676). Allow storing high precision integer values without losing any information +- [Issue #695](https://github.com/influxdb/influxdb/issues/695). Prevent having duplicate field names in the write payload. (Thanks, @seunglee150) +- [Issue #731](https://github.com/influxdb/influxdb/issues/731). Don't enable the udp plugin if the `enabled` option is set to false +- [Issue #733](https://github.com/influxdb/influxdb/issues/733). Print an `INFO` message when the input plugin is disabled +- [Issue #707](https://github.com/influxdb/influxdb/issues/707). Graphite input plugin should work payload delimited by any whitespace character +- [Issue #734](https://github.com/influxdb/influxdb/issues/734). Don't buffer non replicated writes +- [Issue #465](https://github.com/influxdb/influxdb/issues/465). Recreating a currently deleting db or series doesn't bring back the old data anymore +- [Issue #358](https://github.com/influxdb/influxdb/issues/358). **BREAKING** List series should return as a single series +- [Issue #499](https://github.com/influxdb/influxdb/issues/499). **BREAKING** Querying non-existent database or series will return an error +- [Issue #570](https://github.com/influxdb/influxdb/issues/570). InfluxDB crashes during delete/drop of database +- [Issue #592](https://github.com/influxdb/influxdb/issues/592). Drop series is inefficient + +## v0.7.3 [2014-06-13] + +### Bugfixes + +- [Issue #637](https://github.com/influxdb/influxdb/issues/637). Truncate log files if the last request wasn't written properly +- [Issue #646](https://github.com/influxdb/influxdb/issues/646). CRITICAL: Duplicate shard ids for new shards if old shards are deleted. + +## v0.7.2 [2014-05-30] + +### Features + +- [Issue #521](https://github.com/influxdb/influxdb/issues/521). MODE works on all datatypes (Thanks, @richthegeek) + +### Bugfixes + +- [Issue #418](https://github.com/influxdb/influxdb/pull/418). Requests or responses larger than MAX_REQUEST_SIZE break things. +- [Issue #606](https://github.com/influxdb/influxdb/issues/606). InfluxDB will fail to start with invalid permission if log.txt didn't exist +- [Issue #602](https://github.com/influxdb/influxdb/issues/602). Merge will fail to work across shards + +### Features + +## v0.7.1 [2014-05-29] + +### Bugfixes + +- [Issue #579](https://github.com/influxdb/influxdb/issues/579). Reject writes to nonexistent databases +- [Issue #597](https://github.com/influxdb/influxdb/issues/597). Force compaction after deleting data + +### Features + +- [Issue #476](https://github.com/influxdb/influxdb/issues/476). Support ARM architecture +- [Issue #578](https://github.com/influxdb/influxdb/issues/578). Support aliasing for expressions in parenthesis +- [Issue #544](https://github.com/influxdb/influxdb/pull/544). Support forcing node removal from a cluster +- [Issue #591](https://github.com/influxdb/influxdb/pull/591). Support multiple udp input plugins (Thanks, @tpitale) +- [Issue #600](https://github.com/influxdb/influxdb/pull/600). Report version, os, arch, and raftName once per day. + +## v0.7.0 [2014-05-23] + +### Bugfixes + +- [Issue #557](https://github.com/influxdb/influxdb/issues/557). Group by time(1y) doesn't work while time(365d) works +- [Issue #547](https://github.com/influxdb/influxdb/issues/547). Add difference function (Thanks, @mboelstra) +- [Issue #550](https://github.com/influxdb/influxdb/issues/550). Fix tests on 32-bit ARM +- [Issue #524](https://github.com/influxdb/influxdb/issues/524). Arithmetic operators and where conditions don't play nice together +- [Issue #561](https://github.com/influxdb/influxdb/issues/561). Fix missing query in parsing errors +- [Issue #563](https://github.com/influxdb/influxdb/issues/563). Add sample config for graphite over udp +- [Issue #537](https://github.com/influxdb/influxdb/issues/537). Incorrect query syntax causes internal error +- [Issue #565](https://github.com/influxdb/influxdb/issues/565). Empty series names shouldn't cause a panic +- [Issue #575](https://github.com/influxdb/influxdb/issues/575). Single point select doesn't interpret timestamps correctly +- [Issue #576](https://github.com/influxdb/influxdb/issues/576). We shouldn't set timestamps and sequence numbers when listing cq +- [Issue #560](https://github.com/influxdb/influxdb/issues/560). Use /dev/urandom instead of /dev/random +- [Issue #502](https://github.com/influxdb/influxdb/issues/502). Fix a + race condition in assigning id to db+series+field (Thanks @ohurvitz + for reporting this bug and providing a script to repro) + +### Features + +- [Issue #567](https://github.com/influxdb/influxdb/issues/567). Allow selecting from multiple series names by separating them with commas (Thanks, @peekeri) + +### Deprecated + +- [Issue #460](https://github.com/influxdb/influxdb/issues/460). Don't start automatically after installing +- [Issue #529](https://github.com/influxdb/influxdb/issues/529). Don't run influxdb as root +- [Issue #443](https://github.com/influxdb/influxdb/issues/443). Use `name` instead of `username` when returning cluster admins + +## v0.6.5 [2014-05-19] + +### Features + +- [Issue #551](https://github.com/influxdb/influxdb/issues/551). Add TOP and BOTTOM aggregate functions (Thanks, @chobie) + +### Bugfixes + +- [Issue #555](https://github.com/influxdb/influxdb/issues/555). Fix a regression introduced in the raft snapshot format + +## v0.6.4 [2014-05-16] + +### Features + +- Make the write batch size configurable (also applies to deletes) +- Optimize writing to multiple series +- [Issue #546](https://github.com/influxdb/influxdb/issues/546). Add UDP support for Graphite API (Thanks, @peekeri) + +### Bugfixes + +- Fix a bug in shard logic that caused short term shards to be clobbered with long term shards +- [Issue #489](https://github.com/influxdb/influxdb/issues/489). Remove replication factor from CreateDatabase command + +## v0.6.3 [2014-05-13] + +### Features + +- [Issue #505](https://github.com/influxdb/influxdb/issues/505). Return a version header with http the response (Thanks, @majst01) +- [Issue #520](https://github.com/influxdb/influxdb/issues/520). Print the version to the log file + +### Bugfixes + +- [Issue #516](https://github.com/influxdb/influxdb/issues/516). Close WAL log/index files when they aren't being used +- [Issue #532](https://github.com/influxdb/influxdb/issues/532). Don't log graphite connection EOF as an error +- [Issue #535](https://github.com/influxdb/influxdb/issues/535). WAL Replay hangs if response isn't received +- [Issue #538](https://github.com/influxdb/influxdb/issues/538). Don't panic if the same series existed twice in the request with different columns +- [Issue #536](https://github.com/influxdb/influxdb/issues/536). Joining the cluster after shards are creating shouldn't cause new nodes to panic +- [Issue #539](https://github.com/influxdb/influxdb/issues/539). count(distinct()) with fill shouldn't panic on empty groups +- [Issue #534](https://github.com/influxdb/influxdb/issues/534). Create a new series when interpolating + +## v0.6.2 [2014-05-09] + +### Bugfixes + +- [Issue #511](https://github.com/influxdb/influxdb/issues/511). Don't automatically create the database when a db user is created +- [Issue #512](https://github.com/influxdb/influxdb/issues/512). Group by should respect null values +- [Issue #518](https://github.com/influxdb/influxdb/issues/518). Filter Infinities and NaNs from the returned json +- [Issue #522](https://github.com/influxdb/influxdb/issues/522). Committing requests while replaying caused the WAL to skip some log files +- [Issue #369](https://github.com/influxdb/influxdb/issues/369). Fix some edge cases with WAL recovery + +## v0.6.1 [2014-05-06] + +### Bugfixes + +- [Issue #500](https://github.com/influxdb/influxdb/issues/500). Support `y` suffix in time durations +- [Issue #501](https://github.com/influxdb/influxdb/issues/501). Writes with invalid payload should be rejected +- [Issue #507](https://github.com/influxdb/influxdb/issues/507). New cluster admin passwords don't propagate properly to other nodes in a cluster +- [Issue #508](https://github.com/influxdb/influxdb/issues/508). Don't replay WAL entries for servers with no shards +- [Issue #464](https://github.com/influxdb/influxdb/issues/464). Admin UI shouldn't draw graphs for string columns +- [Issue #480](https://github.com/influxdb/influxdb/issues/480). Large values on the y-axis get cut off + +## v0.6.0 [2014-05-02] + +### Feature + +- [Issue #477](https://github.com/influxdb/influxdb/issues/477). Add a udp json interface (Thanks, Julien Ammous) +- [Issue #491](https://github.com/influxdb/influxdb/issues/491). Make initial root password settable through env variable (Thanks, Edward Muller) + +### Bugfixes + +- [Issue #469](https://github.com/influxdb/influxdb/issues/469). Drop continuous queries when a database is dropped +- [Issue #431](https://github.com/influxdb/influxdb/issues/431). Don't log to standard output if a log file is specified in the config file +- [Issue #483](https://github.com/influxdb/influxdb/issues/483). Return 409 if a database already exist (Thanks, Edward Muller) +- [Issue #486](https://github.com/influxdb/influxdb/issues/486). Columns used in the target of continuous query shouldn't be inserted in the time series +- [Issue #490](https://github.com/influxdb/influxdb/issues/490). Database user password's cannot be changed (Thanks, Edward Muller) +- [Issue #495](https://github.com/influxdb/influxdb/issues/495). Enforce write permissions properly + +## v0.5.12 [2014-04-29] + +### Bugfixes + +- [Issue #419](https://github.com/influxdb/influxdb/issues/419),[Issue #478](https://github.com/influxdb/influxdb/issues/478). Allow hostname, raft and protobuf ports to be changed, without requiring manual intervention from the user + +## v0.5.11 [2014-04-25] + +### Features + +- [Issue #471](https://github.com/influxdb/influxdb/issues/471). Read and write permissions should be settable through the http api + +### Bugfixes + +- [Issue #323](https://github.com/influxdb/influxdb/issues/323). Continuous queries should guard against data loops +- [Issue #473](https://github.com/influxdb/influxdb/issues/473). Engine memory optimization + +## v0.5.10 [2014-04-22] + +### Features + +- [Issue #463](https://github.com/influxdb/influxdb/issues/463). Allow series names to use any character (escape by wrapping in double quotes) +- [Issue #447](https://github.com/influxdb/influxdb/issues/447). Allow @ in usernames +- [Issue #466](https://github.com/influxdb/influxdb/issues/466). Allow column names to use any character (escape by wrapping in double quotes) + +### Bugfixes + +- [Issue #458](https://github.com/influxdb/influxdb/issues/458). Continuous queries with group by time() and a column should insert sequence numbers of 1 +- [Issue #457](https://github.com/influxdb/influxdb/issues/457). Deleting series that start with capital letters should work + +## v0.5.9 [2014-04-18] + +### Bugfixes + +- [Issue #446](https://github.com/influxdb/influxdb/issues/446). Check for (de)serialization errors +- [Issue #456](https://github.com/influxdb/influxdb/issues/456). Continuous queries failed if one of the group by columns had null value +- [Issue #455](https://github.com/influxdb/influxdb/issues/455). Comparison operators should ignore null values + +## v0.5.8 [2014-04-17] + +- Renamed config.toml.sample to config.sample.toml + +### Bugfixes + +- [Issue #244](https://github.com/influxdb/influxdb/issues/244). Reconstruct the query from the ast +- [Issue #449](https://github.com/influxdb/influxdb/issues/449). Heartbeat timeouts can cause reading from connection to lock up +- [Issue #451](https://github.com/influxdb/influxdb/issues/451). Reduce the aggregation state that is kept in memory so that + aggregation queries over large periods of time don't take insance amount of memory + +## v0.5.7 [2014-04-15] + +### Features + +- Queries are now logged as INFO in the log file before they run + +### Bugfixes + +- [Issue #328](https://github.com/influxdb/influxdb/issues/328). Join queries with math expressions don't work +- [Issue #440](https://github.com/influxdb/influxdb/issues/440). Heartbeat timeouts in logs +- [Issue #442](https://github.com/influxdb/influxdb/issues/442). shouldQuerySequentially didn't work as expected + causing count(*) queries on large time series to use + lots of memory +- [Issue #437](https://github.com/influxdb/influxdb/issues/437). Queries with negative constants don't parse properly +- [Issue #432](https://github.com/influxdb/influxdb/issues/432). Deleted data using a delete query is resurrected after a server restart +- [Issue #439](https://github.com/influxdb/influxdb/issues/439). Report the right location of the error in the query +- Fix some bugs with the WAL recovery on startup + +## v0.5.6 [2014-04-08] + +### Features + +- [Issue #310](https://github.com/influxdb/influxdb/issues/310). Request should support multiple timeseries +- [Issue #416](https://github.com/influxdb/influxdb/issues/416). Improve the time it takes to drop database + +### Bugfixes + +- [Issue #413](https://github.com/influxdb/influxdb/issues/413). Don't assume that group by interval is greater than a second +- [Issue #415](https://github.com/influxdb/influxdb/issues/415). Include the database when sending an auth error back to the user +- [Issue #421](https://github.com/influxdb/influxdb/issues/421). Make read timeout a config option +- [Issue #392](https://github.com/influxdb/influxdb/issues/392). Different columns in different shards returns invalid results when a query spans those shards + +### Bugfixes + +## v0.5.5 [2014-04-04] + +- Upgrade leveldb 1.10 -> 1.15 + + This should be a backward compatible change, but is here for documentation only + +### Feature + +- Add a command line option to repair corrupted leveldb databases on startup +- [Issue #401](https://github.com/influxdb/influxdb/issues/401). No limit on the number of columns in the group by clause + +### Bugfixes + +- [Issue #398](https://github.com/influxdb/influxdb/issues/398). Support now() and NOW() in the query lang +- [Issue #403](https://github.com/influxdb/influxdb/issues/403). Filtering should work with join queries +- [Issue #404](https://github.com/influxdb/influxdb/issues/404). Filtering with invalid condition shouldn't crash the server +- [Issue #405](https://github.com/influxdb/influxdb/issues/405). Percentile shouldn't crash for small number of values +- [Issue #408](https://github.com/influxdb/influxdb/issues/408). Make InfluxDB recover from internal bugs and panics +- [Issue #390](https://github.com/influxdb/influxdb/issues/390). Multiple response.WriteHeader when querying as admin +- [Issue #407](https://github.com/influxdb/influxdb/issues/407). Start processing continuous queries only after the WAL is initialized +- Close leveldb databases properly if we couldn't create a new Shard. See leveldb\_shard\_datastore\_test:131 + +## v0.5.4 [2014-04-02] + +### Bugfixes + +- [Issue #386](https://github.com/influxdb/influxdb/issues/386). Drop series should work with series containing dots +- [Issue #389](https://github.com/influxdb/influxdb/issues/389). Filtering shouldn't stop prematurely +- [Issue #341](https://github.com/influxdb/influxdb/issues/341). Make the number of shards that are queried in parallel configurable +- [Issue #394](https://github.com/influxdb/influxdb/issues/394). Support count(distinct) and count(DISTINCT) +- [Issue #362](https://github.com/influxdb/influxdb/issues/362). Limit should be enforced after aggregation + +## v0.5.3 [2014-03-31] + +### Bugfixes + +- [Issue #378](https://github.com/influxdb/influxdb/issues/378). Indexing should return if there are no requests added since the last index +- [Issue #370](https://github.com/influxdb/influxdb/issues/370). Filtering and limit should be enforced on the shards +- [Issue #379](https://github.com/influxdb/influxdb/issues/379). Boolean columns should be usable in where clauses +- [Issue #381](https://github.com/influxdb/influxdb/issues/381). Should be able to do deletes as a cluster admin + +## v0.5.2 [2014-03-28] + +### Bugfixes + +- [Issue #342](https://github.com/influxdb/influxdb/issues/342). Data resurrected after a server restart +- [Issue #367](https://github.com/influxdb/influxdb/issues/367). Influxdb won't start if the api port is commented out +- [Issue #355](https://github.com/influxdb/influxdb/issues/355). Return an error on wrong time strings +- [Issue #331](https://github.com/influxdb/influxdb/issues/331). Allow negative time values in the where clause +- [Issue #371](https://github.com/influxdb/influxdb/issues/371). Seris index isn't deleted when the series is dropped +- [Issue #360](https://github.com/influxdb/influxdb/issues/360). Store and recover continuous queries + +## v0.5.1 [2014-03-24] + +### Bugfixes + +- Revert the version of goraft due to a bug found in the latest version + +## v0.5.0 [2014-03-24] + +### Features + +- [Issue #293](https://github.com/influxdb/influxdb/pull/293). Implement a Graphite listener + +### Bugfixes + +- [Issue #340](https://github.com/influxdb/influxdb/issues/340). Writing many requests while replaying seems to cause commits out of order + +## v0.5.0-rc.6 [2014-03-20] + +### Bugfixes + +- Increase raft election timeout to avoid unecessary relections +- Sort points before writing them to avoid an explosion in the request + number when the points are written randomly +- [Issue #335](https://github.com/influxdb/influxdb/issues/335). Fixes regexp for interpolating more than one column value in continuous queries +- [Issue #318](https://github.com/influxdb/influxdb/pull/318). Support EXPLAIN queries +- [Issue #333](https://github.com/influxdb/influxdb/pull/333). Fail + when the password is too short or too long instead of passing it to + the crypto library + +## v0.5.0-rc.5 [2014-03-11] + +### Bugfixes + +- [Issue #312](https://github.com/influxdb/influxdb/issues/312). WAL should wait for server id to be set before recovering +- [Issue #301](https://github.com/influxdb/influxdb/issues/301). Use ref counting to guard against race conditions in the shard cache +- [Issue #319](https://github.com/influxdb/influxdb/issues/319). Propagate engine creation error correctly to the user +- [Issue #316](https://github.com/influxdb/influxdb/issues/316). Make + sure we don't starve goroutines if we get an access denied error + from one of the shards +- [Issue #306](https://github.com/influxdb/influxdb/issues/306). Deleting/Dropping database takes a lot of memory +- [Issue #302](https://github.com/influxdb/influxdb/issues/302). Should be able to set negative timestamps on points +- [Issue #327](https://github.com/influxdb/influxdb/issues/327). Make delete queries not use WAL. This addresses #315, #317 and #314 +- [Issue #321](https://github.com/influxdb/influxdb/issues/321). Make sure we split points on shards properly + +## v0.5.0-rc.4 [2014-03-07] + +### Bugfixes + +- [Issue #298](https://github.com/influxdb/influxdb/issues/298). Fix limit when querying multiple shards +- [Issue #305](https://github.com/influxdb/influxdb/issues/305). Shard ids not unique after restart +- [Issue #309](https://github.com/influxdb/influxdb/issues/309). Don't relog the requests on the remote server +- Fix few bugs in the WAL and refactor the way it works (this requires purging the WAL from previous rc) + +## v0.5.0-rc.3 [2014-03-03] + +### Bugfixes +- [Issue #69](https://github.com/influxdb/influxdb/issues/69). Support column aliases +- [Issue #287](https://github.com/influxdb/influxdb/issues/287). Make the lru cache size configurable +- [Issue #38](https://github.com/influxdb/influxdb/issues/38). Fix a memory leak discussed in this story +- [Issue #286](https://github.com/influxdb/influxdb/issues/286). Make the number of open shards configurable +- Make LevelDB use the max open files configuration option. + +## v0.5.0-rc.2 [2014-02-27] + +### Bugfixes + +- [Issue #274](https://github.com/influxdb/influxdb/issues/274). Crash after restart +- [Issue #277](https://github.com/influxdb/influxdb/issues/277). Ensure duplicate shards won't be created +- [Issue #279](https://github.com/influxdb/influxdb/issues/279). Limits not working on regex queries +- [Issue #281](https://github.com/influxdb/influxdb/issues/281). `./influxdb -v` should print the sha when building from source +- [Issue #283](https://github.com/influxdb/influxdb/issues/283). Dropping shard and restart in cluster causes panic. +- [Issue #288](https://github.com/influxdb/influxdb/issues/288). Sequence numbers should be unique per server id + +## v0.5.0-rc.1 [2014-02-25] + +### Bugfixes + +- Ensure large deletes don't take too much memory +- [Issue #240](https://github.com/influxdb/influxdb/pull/240). Unable to query against columns with `.` in the name. +- [Issue #250](https://github.com/influxdb/influxdb/pull/250). different result between normal and continuous query with "group by" clause +- [Issue #216](https://github.com/influxdb/influxdb/pull/216). Results with no points should exclude columns and points + +### Features + +- [Issue #243](https://github.com/influxdb/influxdb/issues/243). Should have endpoint to GET a user's attributes. +- [Issue #269](https://github.com/influxdb/influxdb/pull/269), [Issue #65](https://github.com/influxdb/influxdb/issues/65) New clustering architecture (see docs), with the side effect that queries can be distributed between multiple shards +- [Issue #164](https://github.com/influxdb/influxdb/pull/269),[Issue #103](https://github.com/influxdb/influxdb/pull/269),[Issue #166](https://github.com/influxdb/influxdb/pull/269),[Issue #165](https://github.com/influxdb/influxdb/pull/269),[Issue #132](https://github.com/influxdb/influxdb/pull/269) Make request log a log file instead of leveldb with recovery on startup + +### Deprecated + +- [Issue #189](https://github.com/influxdb/influxdb/issues/189). `/cluster_admins` and `/db/:db/users` return usernames in a `name` key instead of `username` key. +- [Issue #216](https://github.com/influxdb/influxdb/pull/216). Results with no points should exclude columns and points + +## v0.4.4 [2014-02-05] + +### Features + +- Make the leveldb max open files configurable in the toml file + +## v0.4.3 [2014-01-31] + +### Bugfixes + +- [Issue #225](https://github.com/influxdb/influxdb/issues/225). Remove a hard limit on the points returned by the datastore +- [Issue #223](https://github.com/influxdb/influxdb/issues/223). Null values caused count(distinct()) to panic +- [Issue #224](https://github.com/influxdb/influxdb/issues/224). Null values broke replication due to protobuf limitation + +## v0.4.1 [2014-01-30] + +### Features + +- [Issue #193](https://github.com/influxdb/influxdb/issues/193). Allow logging to stdout. Thanks @schmurfy +- [Issue #190](https://github.com/influxdb/influxdb/pull/190). Add support for SSL. +- [Issue #194](https://github.com/influxdb/influxdb/pull/194). Should be able to disable Admin interface. + +### Bugfixes + +- [Issue #33](https://github.com/influxdb/influxdb/issues/33). Don't call WriteHeader more than once per request +- [Issue #195](https://github.com/influxdb/influxdb/issues/195). Allow the bind address to be configurable, Thanks @schmurfy. +- [Issue #199](https://github.com/influxdb/influxdb/issues/199). Make the test timeout configurable +- [Issue #200](https://github.com/influxdb/influxdb/issues/200). Selecting `time` or `sequence_number` silently fail +- [Issue #215](https://github.com/influxdb/influxdb/pull/215). Server fails to start up after Raft log compaction and restart. + +## v0.4.0 [2014-01-17] + +## Features + +- [Issue #86](https://github.com/influxdb/influxdb/issues/86). Support arithmetic expressions in select clause +- [Issue #92](https://github.com/influxdb/influxdb/issues/92). Change '==' to '=' and '!=' to '<>' +- [Issue #88](https://github.com/influxdb/influxdb/issues/88). Support datetime strings +- [Issue #64](https://github.com/influxdb/influxdb/issues/64). Shard writes and queries across cluster with replay for briefly downed nodes (< 24 hrs) +- [Issue #78](https://github.com/influxdb/influxdb/issues/78). Sequence numbers persist across restarts so they're not reused +- [Issue #102](https://github.com/influxdb/influxdb/issues/102). Support expressions in where condition +- [Issue #101](https://github.com/influxdb/influxdb/issues/101). Support expressions in aggregates +- [Issue #62](https://github.com/influxdb/influxdb/issues/62). Support updating and deleting column values +- [Issue #96](https://github.com/influxdb/influxdb/issues/96). Replicate deletes in a cluster +- [Issue #94](https://github.com/influxdb/influxdb/issues/94). delete queries +- [Issue #116](https://github.com/influxdb/influxdb/issues/116). Use proper logging +- [Issue #40](https://github.com/influxdb/influxdb/issues/40). Use TOML instead of JSON in the config file +- [Issue #99](https://github.com/influxdb/influxdb/issues/99). Support list series in the query language +- [Issue #149](https://github.com/influxdb/influxdb/issues/149). Cluster admins should be able to perform reads and writes. +- [Issue #108](https://github.com/influxdb/influxdb/issues/108). Querying one point using `time =` +- [Issue #114](https://github.com/influxdb/influxdb/issues/114). Servers should periodically check that they're consistent. +- [Issue #93](https://github.com/influxdb/influxdb/issues/93). Should be able to drop a time series +- [Issue #177](https://github.com/influxdb/influxdb/issues/177). Support drop series in the query language. +- [Issue #184](https://github.com/influxdb/influxdb/issues/184). Implement Raft log compaction. +- [Issue #153](https://github.com/influxdb/influxdb/issues/153). Implement continuous queries + +### Bugfixes + +- [Issue #90](https://github.com/influxdb/influxdb/issues/90). Group by multiple columns panic +- [Issue #89](https://github.com/influxdb/influxdb/issues/89). 'Group by' combined with 'where' not working +- [Issue #106](https://github.com/influxdb/influxdb/issues/106). Don't panic if we only see one point and can't calculate derivative +- [Issue #105](https://github.com/influxdb/influxdb/issues/105). Panic when using a where clause that reference columns with null values +- [Issue #61](https://github.com/influxdb/influxdb/issues/61). Remove default limits from queries +- [Issue #118](https://github.com/influxdb/influxdb/issues/118). Make column names starting with '_' legal +- [Issue #121](https://github.com/influxdb/influxdb/issues/121). Don't fall back to the cluster admin auth if the db user auth fails +- [Issue #127](https://github.com/influxdb/influxdb/issues/127). Return error on delete queries with where condition that don't have time +- [Issue #117](https://github.com/influxdb/influxdb/issues/117). Fill empty groups with default values +- [Issue #150](https://github.com/influxdb/influxdb/pull/150). Fix parser for when multiple divisions look like a regex. +- [Issue #158](https://github.com/influxdb/influxdb/issues/158). Logged deletes should be stored with the time range if missing. +- [Issue #136](https://github.com/influxdb/influxdb/issues/136). Make sure writes are replicated in order to avoid triggering replays +- [Issue #145](https://github.com/influxdb/influxdb/issues/145). Server fails to join cluster if all starting at same time. +- [Issue #176](https://github.com/influxdb/influxdb/issues/176). Drop database should take effect on all nodes +- [Issue #180](https://github.com/influxdb/influxdb/issues/180). Column names not returned when running multi-node cluster and writing more than one point. +- [Issue #182](https://github.com/influxdb/influxdb/issues/182). Queries with invalid limit clause crash the server + +### Deprecated + +- deprecate '==' and '!=' in favor of '=' and '<>', respectively +- deprecate `/dbs` (for listing databases) in favor of a more consistent `/db` endpoint +- deprecate `username` field for a more consistent `name` field in `/db/:db/users` and `/cluster_admins` +- deprecate endpoints `/db/:db/admins/:user` in favor of using `/db/:db/users/:user` which should + be used to update user flags, password, etc. +- Querying for column names that don't exist no longer throws an error. + +## v0.3.2 + +## Features + +- [Issue #82](https://github.com/influxdb/influxdb/issues/82). Add endpoint for listing available admin interfaces. +- [Issue #80](https://github.com/influxdb/influxdb/issues/80). Support durations when specifying start and end time +- [Issue #81](https://github.com/influxdb/influxdb/issues/81). Add support for IN + +## Bugfixes + +- [Issue #75](https://github.com/influxdb/influxdb/issues/75). Don't allow time series names that start with underscore +- [Issue #85](https://github.com/influxdb/influxdb/issues/85). Non-existing columns exist after they have been queried before + +## v0.3.0 + +## Features + +- [Issue #51](https://github.com/influxdb/influxdb/issues/51). Implement first and last aggregates +- [Issue #35](https://github.com/influxdb/influxdb/issues/35). Support table aliases in Join Queries +- [Issue #71](https://github.com/influxdb/influxdb/issues/71). Add WillReturnSingleSeries to the Query +- [Issue #61](https://github.com/influxdb/influxdb/issues/61). Limit should default to 10k +- [Issue #59](https://github.com/influxdb/influxdb/issues/59). Add histogram aggregate function + +## Bugfixes + +- Fix join and merges when the query is a descending order query +- [Issue #57](https://github.com/influxdb/influxdb/issues/57). Don't panic when type of time != float +- [Issue #63](https://github.com/influxdb/influxdb/issues/63). Aggregate queries should not have a sequence_number column + +## v0.2.0 + +### Features + +- [Issue #37](https://github.com/influxdb/influxdb/issues/37). Support the negation of the regex matcher !~ +- [Issue #47](https://github.com/influxdb/influxdb/issues/47). Spill out query and database detail at the time of bug report + +### Bugfixes + +- [Issue #36](https://github.com/influxdb/influxdb/issues/36). The regex operator should be =~ not ~= +- [Issue #39](https://github.com/influxdb/influxdb/issues/39). Return proper content types from the http api +- [Issue #42](https://github.com/influxdb/influxdb/issues/42). Make the api consistent with the docs +- [Issue #41](https://github.com/influxdb/influxdb/issues/41). Table/Points not deleted when database is dropped +- [Issue #45](https://github.com/influxdb/influxdb/issues/45). Aggregation shouldn't mess up the order of the points +- [Issue #44](https://github.com/influxdb/influxdb/issues/44). Fix crashes on RHEL 5.9 +- [Issue #34](https://github.com/influxdb/influxdb/issues/34). Ascending order always return null for columns that have a null value +- [Issue #55](https://github.com/influxdb/influxdb/issues/55). Limit should limit the points that match the Where clause +- [Issue #53](https://github.com/influxdb/influxdb/issues/53). Writing null values via HTTP API fails + +### Deprecated + +- Preparing to deprecate `/dbs` (for listing databases) in favor of a more consistent `/db` endpoint +- Preparing to deprecate `username` field for a more consistent `name` field in the `/db/:db/users` +- Preparing to deprecate endpoints `/db/:db/admins/:user` in favor of using `/db/:db/users/:user` which should + be used to update user flags, password, etc. + +## v0.1.0 + +### Features + +- [Issue #29](https://github.com/influxdb/influxdb/issues/29). Semicolon is now optional in queries +- [Issue #31](https://github.com/influxdb/influxdb/issues/31). Support Basic Auth as well as query params for authentication. + +### Bugfixes + +- Don't allow creating users with empty username +- [Issue #22](https://github.com/influxdb/influxdb/issues/22). Don't set goroot if it was set +- [Issue #25](https://github.com/influxdb/influxdb/issues/25). Fix queries that use the median aggregator +- [Issue #26](https://github.com/influxdb/influxdb/issues/26). Default log and db directories should be in /opt/influxdb/shared/data +- [Issue #27](https://github.com/influxdb/influxdb/issues/27). Group by should not blow up if the one of the columns in group by has null values +- [Issue #30](https://github.com/influxdb/influxdb/issues/30). Column indexes/names getting off somehow +- [Issue #32](https://github.com/influxdb/influxdb/issues/32). Fix many typos in the codebase. Thanks @pborreli + +## v0.0.9 + +#### Features + +- Add stddev(...) support +- Better docs, thanks @auxesis and @d-snp. + +#### Bugfixes + +- Set PYTHONPATH and CC appropriately on mac os x. +- [Issue #18](https://github.com/influxdb/influxdb/issues/18). Fix 386 debian and redhat packages +- [Issue #23](https://github.com/influxdb/influxdb/issues/23). Fix the init scripts on redhat + +## v0.0.8 + +#### Features + +- Add a way to reset the root password from the command line. +- Add distinct(..) and derivative(...) support +- Print test coverage if running go1.2 + +#### Bugfixes + +- Fix the default admin site path in the .deb and .rpm packages. +- Fix the configuration filename in the .tar.gz package. + +## v0.0.7 + +#### Features + +- include the admin site in the repo to make it easier for newcomers. + +## v0.0.6 + +#### Features + +- Add count(distinct(..)) support + +#### Bugfixes + +- Reuse levigo read/write options. + +## v0.0.5 + +#### Features + +- Cache passwords in memory to speed up password verification +- Add MERGE and INNER JOIN support + +#### Bugfixes + +- All columns should be returned if `select *` was used +- Read/Write benchmarks + +## v0.0.2 + +#### Features + +- Add an admin UI +- Deb and RPM packages + +#### Bugfixes + +- Fix some nil pointer dereferences +- Cleanup the aggregators implementation + +## v0.0.1 [2013-10-22] + + * Initial Release diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/CONTRIBUTING.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/CONTRIBUTING.md new file mode 100644 index 000000000..56d188693 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/CONTRIBUTING.md @@ -0,0 +1,231 @@ +Contributing to InfluxDB +======================== + +Bug reports +--------------- +Before you file an issue, please search existing issues in case it has already been filed, or perhaps even fixed. If you file an issue, please include the following. +* Full details of your operating system (or distribution) e.g. 64-bit Ubuntu 14.04. +* The version of InfluxDB you are running +* Whether you installed it using a pre-built package, or built it from source. +* A small test case, if applicable, that demonstrates the issues. + +Remember the golden rule of bug reports: **The easier you make it for us to reproduce the problem, the faster it will get fixed.** +If you have never written a bug report before, or if you want to brush up on your bug reporting skills, we recommend reading [Simon Tatham's essay "How to Report Bugs Effectively."](http://www.chiark.greenend.org.uk/~sgtatham/bugs.html) + +Test cases should be in the form of `curl` commands. For example: +``` +# create database +curl -G http://localhost:8086/query --data-urlencode "q=CREATE DATABASE mydb" + +# create retention policy +curl -G http://localhost:8086/query --data-urlencode "q=CREATE RETENTION POLICY myrp ON mydb DURATION 365d REPLICATION 1 DEFAULT" + +# write data +curl -X POST http://localhost:8086/write --data-urlencode "db=mydb" --data-binary "cpu,region=useast,host=server_1,service=redis value=61" + +# Delete a Measurement +curl -G http://localhost:8086/query --data-urlencode 'db=mydb' --data-urlencode 'q=DROP MEASUREMENT cpu' + +# Query the Measurement +# Bug: expected it to return no data, but data comes back. +curl -G http://localhost:8086/query --data-urlencode 'db=mydb' --data-urlencode 'q=SELECT * from cpu' +``` +**If you don't include a clear test case like this, your issue may not be investigated, and may even be closed**. If writing the data is too difficult, please zip up your data directory and include a link to it in your bug report. + +Please note that issues are *not the place to file general questions* such as "how do I use collectd with InfluxDB?" Questions of this nature should be sent to the [Google Group](https://groups.google.com/forum/#!forum/influxdb), not filed as issues. Issues like this will be closed. + +Feature requests +--------------- +We really like to receive feature requests, as it helps us prioritize our work. Please be clear about your requirements, as incomplete feature requests may simply be closed if we don't understand what you would like to see added to InfluxDB. + +Contributing to the source code +--------------- + +InfluxDB follows standard Go project structure. This means that all +your go development are done in `$GOPATH/src`. GOPATH can be any +directory under which InfluxDB and all its dependencies will be +cloned. For more details on recommended go project's structure, see +[How to Write Go Code](http://golang.org/doc/code.html) and +[Go: Best Practices for Production Environments](http://peter.bourgon.org/go-in-production/), or you can just follow +the steps below. + +Submitting a pull request +------------ +To submit a pull request you should fork the InfluxDB repository, and make your change on a feature branch of your fork. Then generate a pull request from your branch against *master* of the InfluxDB repository. Include in your pull request details of your change -- the why *and* the how -- as well as the testing your performed. Also, be sure to run the test suite with your change in place. Changes that cause tests to fail cannot be merged. + +There will usually be some back and forth as we finalize the change, but once that completes it may be merged. + +To assist in review for the PR, please add the following to your pull request comment: + +```md +- [ ] CHANGELOG.md updated +- [ ] Rebased/mergable +- [ ] Tests pass +- [ ] Sign [CLA](http://influxdb.com/community/cla.html) (if not already signed) +``` + +Use of third-party packages +------------ +A third-party package is defined as one that is not part of the standard Go distribution. Generally speaking we prefer to minimize our use of third-party packages, and avoid them unless absolutely necessarly. We'll often write a little bit of code rather than pull in a third-party package. Of course, we do use some third-party packages -- most importantly we use [BoltDB](https://github.com/boltdb/bolt) as the storage engine. So to maximise the chance your change will be accepted by us, use only the standard libaries, or the third-party packages we have decided to use. + +For rationale, check out the post [The Case Against Third Party Libraries](http://blog.gopheracademy.com/advent-2014/case-against-3pl/). + +Signing the CLA +--------------- + +If you are going to be contributing back to InfluxDB please take a +second to sign our CLA, which can be found +[on our website](http://influxdb.com/community/cla.html). + +Installing Go +------------- +InfluxDB requires Go 1.4 or greater. + +At InfluxDB we find gvm, a Go version manager, useful for installing Go. For instructions +on how to install it see [the gvm page on github](https://github.com/moovweb/gvm). + +After installing gvm you can install and set the default go version by +running the following: + + gvm install go1.4 + gvm use go1.4 --default + +Revision Control Systems +------ +Go has the ability to import remote packages via revision control systems with the `go get` command. To ensure that you can retrieve any remote package, be sure to install the following rcs software to your system. +Currently the project only depends on `git` and `mercurial`. + +* [Install Git](http://git-scm.com/book/en/Getting-Started-Installing-Git) +* [Install Mercurial](http://mercurial.selenic.com/wiki/Download) + +Project structure +----------------- +First you need to setup the project structure: + + export GOPATH=$HOME/gocodez + mkdir -p $GOPATH/src/github.com/influxdb + cd $GOPATH/src/github.com/influxdb + git clone git@github.com:influxdb/influxdb + +You can add the line `export GOPATH=$HOME/gocodez` to your bash/zsh +file to be set for every shell instead of having to manually run it +everytime. + +We have a pre commit hook to make sure code is formatted properly +and vetted before you commit any changes. We strongly recommend using the pre +commit hook to guard against accidentally committing unformatted +code. To use the pre-commit hook, run the following: + + cd $GOPATH/src/github.com/influxdb/influxdb + cp .hooks/pre-commit .git/hooks/ + +In case the commit is rejected because it's not formatted you can run +the following to format the code: + +``` +go fmt ./... +go vet ./... +``` + +To install go vet, run the following command: +``` +go get golang.org/x/tools/cmd/vet +``` + +NOTE: If you have not installed mercurial, the above command will fail. See [Revision Control Systems](#revision-control-systems) above. + +For more information on `go vet`, [read the GoDoc](https://godoc.org/golang.org/x/tools/cmd/vet). + +Build and Test +----- + +Make sure you have Go installed and the project structure as shown above. To then build the project, execute the following commands: + +```bash +cd $GOPATH/src/github.com/influxdb +go get -u -f -t ./... +go build ./... +``` + +To then install the binaries, run the following command. They can be found in `$GOPATH/bin`. Please note that the InfluxDB binary is named `influxd`, not `influxdb`. + +```bash +go install ./... +``` + +To set the version and commit flags during the build pass the following to the build command: + +```bash +-ldflags="-X main.version $VERSION -X main.branch $BRANCH -X main.commit $COMMIT" +``` + +where `$VERSION` is the version, `$BRANCH` is the branch, and `$COMMIT` is the git commit hash. + +To run the tests, execute the following command: + +```bash +cd $GOPATH/src/github.com/influxdb/influxdb +go test -v ./... + +# run tests that match some pattern +go test -run=TestDatabase . -v + +# run tests and show coverage +go test -coverprofile /tmp/cover . && go tool cover -html /tmp/cover +``` + +To install go cover, run the following command: +``` +go get golang.org/x/tools/cmd/cover +``` + +Generated Google Protobuf code +----------------- +Most changes to the source do not require that the generated protocol buffer code be changed. But if you need to modify the protocol buffer code, you'll first need to install the protocol buffers toolchain. + +First install the [protocol buffer compiler](https://developers.google.com/protocol-buffers/ +) 2.6.1 or later for your OS: + +Then install the go plugins: + +```bash +go get github.com/gogo/protobuf/proto +go get github.com/gogo/protobuf/protoc-gen-gogo +go get github.com/gogo/protobuf/gogoproto +``` + +Finally run, `go generate` after updating any `*.proto` file: + +```bash +go generate ./... +``` +**Trouleshooting** + +If generating the protobuf code is failing for you, check each of the following: + * Ensure the protobuf library can be found. Make sure that `LD_LIBRRARY_PATH` includes the directory in which the library `libprotoc.so` has been installed. + * Ensure the command `protoc-gen-gogo`, found in `GOPATH/bin`, is on your path. This can be done by adding `GOPATH/bin` to `PATH`. + +Profiling +----- +When troubleshooting problems with CPU or memory the Go toolchain can be helpful. You can start InfluxDB with CPU or memory profiling turned on. For example: + +```sh +# start influx with profiling +./influxd -cpuprofile influxd.prof +# run queries, writes, whatever you're testing +# open up pprof +go tool pprof influxd influxd.prof +# once inside run "web", opens up browser with the CPU graph +# can also run "web " to zoom in. Or "list " to see specific lines +``` + +Continuous Integration testing +----- +InfluxDB uses CirceCI for continuous integration testing. To see how the code is built and tested, check out [this file](https://github.com/influxdb/influxdb/blob/master/circle-test.sh). It closely follows the build and test process outlined above. You can see the exact version of Go InfluxDB uses for testing by consulting that file. + +Useful links +------------ +- [Useful techniques in Go](http://arslan.io/ten-useful-techniques-in-go) +- [Go in production](http://peter.bourgon.org/go-in-production/) +- [Principles of designing Go APIs with channels](https://inconshreveable.com/07-08-2014/principles-of-designing-go-apis-with-channels/) +- [Common mistakes in Golang](http://soryy.com/blog/2014/common-mistakes-with-go-lang/). Especially this section `Loops, Closures, and Local Variables` diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/DOCKER.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/DOCKER.md new file mode 100644 index 000000000..adb33a416 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/DOCKER.md @@ -0,0 +1,44 @@ +# Docker Setup +======================== + +This document describes how to build and run a minimal InfluxDB container under Docker. Currently, it has only been tested for local development and assumes that you have a working docker environment. + +## Building Image + +To build a docker image for InfluxDB from your current checkout, run the following: + +``` +$ ./build-docker.sh +``` + +This script uses the `golang:1.5` image to build a fully static binary of `influxd` and then adds it to a minimal `scratch` image. + +To build the image using a different version of go: + +``` +$ GO_VER=1.4.2 ./build-docker.sh +``` + +Available version can be found [here](https://hub.docker.com/_/golang/). + +## Single Node Container + +This will start an interactive, single-node, that publishes the containers port `8086` and `8088` to the hosts ports `8086` and `8088` respectively. This is identical to starting `influxd` manually. + +``` +$ docker run -it -p 8086:8086 -p 8088:8088 influxdb +``` + +## Multi-Node Cluster + +This will create a simple 3-node cluster. The data is stored within the container and will be lost when the container is removed. This is only useful for test clusters. + +The `HOST_IP` env variable should be your host IP if running under linux or the virtualbox VM IP if running under OSX. On OSX, this would be something like: `$(docker-machine ip dev)` or `$(boot2docker ip)` depending on which docker tool you are using. + +``` +$ export HOST_IP= +$ docker run -it -p 8086:8088 -p 8088:8088 influxdb -hostname $HOST_IP:8088 +$ docker run -it -p 8186:8088 -p 8188:8088 influxdb -hostname $HOST_IP:8188 -join $HOST_IP:8088 +$ docker run -it -p 8286:8088 -p 8288:8088 influxdb -hostname $HOST_IP:8288 -join $HOST_IP:8088 +``` + diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/Dockerfile b/Godeps/_workspace/src/github.com/influxdb/influxdb/Dockerfile new file mode 100644 index 000000000..d30cd300d --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/Dockerfile @@ -0,0 +1,24 @@ +FROM busybox:ubuntu-14.04 + +MAINTAINER Jason Wilder "" + +# admin, http, udp, cluster, graphite, opentsdb, collectd +EXPOSE 8083 8086 8086/udp 8088 2003 4242 25826 + +WORKDIR /app + +# copy binary into image +COPY influxd /app/ + +# Add influxd to the PATH +ENV PATH=/app:$PATH + +# Generate a default config +RUN influxd config > /etc/influxdb.toml + +# Use /data for all disk storage +RUN sed -i 's/dir = "\/.*influxdb/dir = "\/data/' /etc/influxdb.toml + +VOLUME ["/data"] + +ENTRYPOINT ["influxd", "--config", "/etc/influxdb.toml"] diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/LICENSE b/Godeps/_workspace/src/github.com/influxdb/influxdb/LICENSE new file mode 100644 index 000000000..d50222706 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/LICENSE @@ -0,0 +1,20 @@ +The MIT License (MIT) + +Copyright (c) 2013-2015 Errplane Inc. + +Permission is hereby granted, free of charge, to any person obtaining a copy of +this software and associated documentation files (the "Software"), to deal in +the Software without restriction, including without limitation the rights to +use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of +the Software, and to permit persons to whom the Software is furnished to do so, +subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS +FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR +COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER +IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN +CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/QUERIES.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/QUERIES.md new file mode 100644 index 000000000..8491aa7ec --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/QUERIES.md @@ -0,0 +1,180 @@ +The top level name is called a measurement. These names can contain any characters. Then there are field names, field values, tag keys and tag values, which can also contain any characters. However, if the measurement, field, or tag contains any character other than [A-Z,a-z,0-9,_], or if it starts with a digit, it must be double-quoted. Therefore anywhere a measurement name, field name, field value, tag name, or tag value appears it should be wrapped in double quotes. + +# Databases & retention policies + +```sql +-- create a database +CREATE DATABASE + +-- create a retention policy +CREATE RETENTION POLICY ON DURATION REPLICATION [DEFAULT] + +-- alter retention policy +ALTER RETENTION POLICY ON (DURATION | REPLICATION | DEFAULT)+ + +-- drop a database +DROP DATABASE + +-- drop a retention policy +DROP RETENTION POLICY ON +``` +where `` is either `INF` for infinite retention, or an integer followed by the desired unit of time: u,ms,s,m,h,d,w for microseconds, milliseconds, seconds, minutes, hours, days, or weeks, respectively. `` must be an integer. + +If present, `DEFAULT` sets the retention policy as the default retention policy for writes and reads. + +# Users and permissions + +```sql +-- create user +CREATE USER WITH PASSWORD '' + +-- grant privilege on a database +GRANT ON TO + +-- grant cluster admin privileges +GRANT ALL [PRIVILEGES] TO + +-- revoke privilege +REVOKE ON FROM + +-- revoke all privileges for a DB +REVOKE ALL [PRIVILEGES] ON FROM + +-- revoke all privileges including cluster admin +REVOKE ALL [PRIVILEGES] FROM + +-- combine db creation with privilege assignment (user must already exist) +CREATE DATABASE GRANT TO +CREATE DATABASE REVOKE FROM + +-- delete a user +DROP USER + + +``` +where ` := READ | WRITE | All `. + +Authentication must be enabled in the influxdb.conf file for user permissions to be in effect. + +By default, newly created users have no privileges to any databases. + +Cluster administration privileges automatically grant full read and write permissions to all databases, regardless of subsequent database-specific privilege revocation statements. + +# Select + +```sql +SELECT mean(value) from cpu WHERE host = 'serverA' AND time > now() - 4h GROUP BY time(5m) + +SELECT mean(value) from cpu WHERE time > now() - 4h GROUP BY time(5m), region +``` + +## Group By + +# Delete + +# Series + +## Destroy + +```sql +DROP MEASUREMENT +DROP MEASUREMENT cpu WHERE region = 'uswest' +``` + +## Show + +Show series queries are for pulling out individual series from measurement names and tag data. They're useful for discovery. + +```sql +-- show all databases +SHOW DATABASES + +-- show measurement names +SHOW MEASUREMENTS +SHOW MEASUREMENTS LIMIT 15 +SHOW MEASUREMENTS LIMIT 10 OFFSET 40 +SHOW MEASUREMENTS WHERE service = 'redis' +-- LIMIT and OFFSET can be applied to any of the SHOW type queries + +-- show all series across all measurements/tagsets +SHOW SERIES + +-- get a show of all series for any measurements where tag key region = tak value 'uswest' +SHOW SERIES WHERE region = 'uswest' + +SHOW SERIES FROM cpu_load WHERE region = 'uswest' LIMIT 10 + +-- returns the 100 - 109 rows in the result. In the case of SHOW SERIES, which returns +-- series split into measurements. Each series counts as a row. So you could see only a +-- single measurement returned, but 10 series within it. +SHOW SERIES FROM cpu_load WHERE region = 'uswest' LIMIT 10 OFFSET 100 + +-- show all retention policies on a database +SHOW RETENTION POLICIES ON mydb + +-- get a show of all tag keys across all measurements +SHOW TAG KEYS + +-- show all the tag keys for a given measurement +SHOW TAG KEYS FROM cpu +SHOW TAG KEYS FROM temperature, wind_speed + +-- show all the tag values. note that a single WHERE TAG KEY = '...' clause is required +SHOW TAG VALUES WITH TAG KEY = 'region' +SHOW TAG VALUES FROM cpu WHERE region = 'uswest' WITH TAG KEY = 'host' + +-- and you can do stuff against fields +SHOW FIELD KEYS FROM cpu + +-- but you can't do this +SHOW FIELD VALUES +-- we don't index field values, so this query should be invalid. + +-- show all users +SHOW USERS +``` + +Note that `FROM` and `WHERE` are optional clauses in most of the show series queries. + +And the show series output looks like this: + +```json +[ + { + "name": "cpu", + "columns": ["id", "region", "host"], + "values": [ + 1, "uswest", "servera", + 2, "uswest", "serverb" + ] + }, + { + "name": "reponse_time", + "columns": ["id", "application", "host"], + "values": [ + 3, "myRailsApp", "servera" + ] + } +] +``` + +# Continuous Queries + +Continous queries are going to be inspired by MySQL `TRIGGER` syntax: + +http://dev.mysql.com/doc/refman/5.0/en/trigger-syntax.html + +Instead of having automatically-assigned ids, named continuous queries allows for some level of duplication prevention, +particularly in the case where creation is scripted. + +## Create + + CREATE CONTINUOUS QUERY AS SELECT ... FROM ... + +## Destroy + + DROP CONTINUOUS QUERY + +## List + + SHOW CONTINUOUS QUERIES diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/README.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/README.md new file mode 100644 index 000000000..a13be4127 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/README.md @@ -0,0 +1,71 @@ +# InfluxDB [![Circle CI](https://circleci.com/gh/influxdb/influxdb/tree/master.svg?style=svg)](https://circleci.com/gh/influxdb/influxdb/tree/master) + +## An Open-Source, Distributed, Time Series Database + +> InfluxDB v0.9.0 is now out. Going forward, the 0.9.x series of releases will not make breaking API changes or breaking changes to the underlying data storage. However, 0.9.0 clustering should be considered an alpha release. + +InfluxDB is an open source **distributed time series database** with +**no external dependencies**. It's useful for recording metrics, +events, and performing analytics. + +## Features + +* Built-in [HTTP API](http://influxdb.com/docs/v0.9/concepts/reading_and_writing_data.html) so you don't have to write any server side code to get up and running. +* Data can be tagged, allowing very flexible querying. +* SQL-like query language. +* Clustering is supported out of the box, so that you can scale horizontally to handle your data. +* Simple to install and manage, and fast to get data in and out. +* It aims to answer queries in real-time. That means every data point is + indexed as it comes in and is immediately available in queries that + should return in < 100ms. + +## Getting Started +*The following directions apply only to the 0.9.0 release or building from the source on master.* + +### Building + +You don't need to build the project to use it - you can use any of our +[pre-built packages](http://influxdb.com/download/index.html) to install InfluxDB. That's +the recommended way to get it running. However, if you want to contribute to the core of InfluxDB, you'll need to build. +For those adventurous enough, you can +[follow along on our docs](http://github.com/influxdb/influxdb/blob/master/CONTRIBUTING.md). + +### Starting InfluxDB +* `service influxdb start` if you have installed InfluxDB using an official Debian or RPM package. +* `$GOPATH/bin/influxd` if you have built InfluxDB from source. + +### Creating your first database + +``` +curl -G 'http://localhost:8086/query' --data-urlencode "q=CREATE DATABASE mydb" +``` + +### Insert some data +``` +curl -XPOST 'http://localhost:8086/write?db=mydb' \ +-d 'cpu,host=server01,region=uswest load=42 1434055562000000000' + +curl -XPOST 'http://localhost:8086/write?db=mydb' \ +-d 'cpu,host=server02,region=uswest load=78 1434055562000000000' + +curl -XPOST 'http://localhost:8086/write?db=mydb' \ +-d 'cpu,host=server03,region=useast load=15.4 1434055562000000000' +``` + +### Query for the data +```JSON +curl -G http://localhost:8086/query?pretty=true --data-urlencode "db=mydb" \ +--data-urlencode "q=SELECT * FROM cpu WHERE host='server01' AND time < now - 1d" +``` + +### Analyze the data +```JSON +curl -G http://localhost:8086/query?pretty=true --data-urlencode "db=mydb" \ +--data-urlencode "q=SELECT mean(load) FROM cpu WHERE region='uswest'" +``` + +## Helpful Links + +* Understand the [design goals and motivations of the project](http://influxdb.com/docs/v0.9/introduction/overview.html). +* Follow the [getting started guide](http://influxdb.com/docs/v0.9/introduction/getting_started.html) to find out how to install InfluxDB, start writing more data, and issue more queries - in just a few minutes. +* See the [HTTP API documentation to start writing a library for your favorite language](http://influxdb.com/docs/v0.9/concepts/reading_and_writing_data.html). diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/balancer.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/balancer.go new file mode 100644 index 000000000..25abbf6f1 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/balancer.go @@ -0,0 +1,78 @@ +package influxdb + +import ( + "math/rand" + + "github.com/influxdb/influxdb/meta" +) + +// Balancer represents a load-balancing algorithm for a set of nodes +type Balancer interface { + // Next returns the next Node according to the balancing method + // or nil if there are no nodes available + Next() *meta.NodeInfo +} + +type nodeBalancer struct { + nodes []meta.NodeInfo // data nodes to balance between + p int // current node index +} + +// NewNodeBalancer create a shuffled, round-robin balancer so that +// multiple instances will return nodes in randomized order and each +// each returned node will be repeated in a cycle +func NewNodeBalancer(nodes []meta.NodeInfo) Balancer { + // make a copy of the node slice so we can randomize it + // without affecting the original instance as well as ensure + // that each Balancer returns nodes in a different order + b := &nodeBalancer{} + + b.nodes = make([]meta.NodeInfo, len(nodes)) + copy(b.nodes, nodes) + + b.shuffle() + return b +} + +// shuffle randomizes the ordering the balancers available nodes +func (b *nodeBalancer) shuffle() { + for i := range b.nodes { + j := rand.Intn(i + 1) + b.nodes[i], b.nodes[j] = b.nodes[j], b.nodes[i] + } +} + +// online returns a slice of the nodes that are online +func (b *nodeBalancer) online() []meta.NodeInfo { + return b.nodes + // now := time.Now().UTC() + // up := []meta.NodeInfo{} + // for _, n := range b.nodes { + // if n.OfflineUntil.After(now) { + // continue + // } + // up = append(up, n) + // } + // return up +} + +// Next returns the next available nodes +func (b *nodeBalancer) Next() *meta.NodeInfo { + // only use online nodes + up := b.online() + + // no nodes online + if len(up) == 0 { + return nil + } + + // rollover back to the beginning + if b.p >= len(up) { + b.p = 0 + } + + d := &up[b.p] + b.p += 1 + + return d +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/balancer_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/balancer_test.go new file mode 100644 index 000000000..ca1942c33 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/balancer_test.go @@ -0,0 +1,115 @@ +package influxdb_test + +import ( + "fmt" + "testing" + + "github.com/influxdb/influxdb" + "github.com/influxdb/influxdb/meta" +) + +func NewNodes() []meta.NodeInfo { + var nodes []meta.NodeInfo + for i := 1; i <= 2; i++ { + nodes = append(nodes, meta.NodeInfo{ + ID: uint64(i), + Host: fmt.Sprintf("localhost:999%d", i), + }) + } + return nodes +} + +func TestBalancerEmptyNodes(t *testing.T) { + b := influxdb.NewNodeBalancer([]meta.NodeInfo{}) + got := b.Next() + if got != nil { + t.Errorf("expected nil, got %v", got) + } +} + +func TestBalancerUp(t *testing.T) { + nodes := NewNodes() + b := influxdb.NewNodeBalancer(nodes) + + // First node in randomized round-robin order + first := b.Next() + if first == nil { + t.Errorf("expected datanode, got %v", first) + } + + // Second node in randomized round-robin order + second := b.Next() + if second == nil { + t.Errorf("expected datanode, got %v", second) + } + + // Should never get the same node in order twice + if first.ID == second.ID { + t.Errorf("expected first != second. got %v = %v", first.ID, second.ID) + } +} + +/* +func TestBalancerDown(t *testing.T) { + nodes := NewNodes() + b := influxdb.NewNodeBalancer(nodes) + + nodes[0].Down() + + // First node in randomized round-robin order + first := b.Next() + if first == nil { + t.Errorf("expected datanode, got %v", first) + } + + // Second node should rollover to the first up node + second := b.Next() + if second == nil { + t.Errorf("expected datanode, got %v", second) + } + + // Health node should be returned each time + if first.ID != 2 && first.ID != second.ID { + t.Errorf("expected first != second. got %v = %v", first.ID, second.ID) + } +} +*/ + +/* +func TestBalancerBackUp(t *testing.T) { + nodes := newDataNodes() + b := influxdb.NewNodeBalancer(nodes) + + nodes[0].Down() + + for i := 0; i < 3; i++ { + got := b.Next() + if got == nil { + t.Errorf("expected datanode, got %v", got) + } + + if exp := uint64(2); got.ID != exp { + t.Errorf("wrong node id: exp %v, got %v", exp, got.ID) + } + } + + nodes[0].Up() + + // First node in randomized round-robin order + first := b.Next() + if first == nil { + t.Errorf("expected datanode, got %v", first) + } + + // Second node should rollover to the first up node + second := b.Next() + if second == nil { + t.Errorf("expected datanode, got %v", second) + } + + // Should get both nodes returned + if first.ID == second.ID { + t.Errorf("expected first != second. got %v = %v", first.ID, second.ID) + } +} +*/ diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/build-docker.sh b/Godeps/_workspace/src/github.com/influxdb/influxdb/build-docker.sh new file mode 100644 index 000000000..6f303201e --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/build-docker.sh @@ -0,0 +1,7 @@ +#!/bin/sh -x -e + +GO_VER=${GO_VER:-1.5} + +docker run -it -v "$GOPATH":/gopath -v "$(pwd)":/app -e "GOPATH=/gopath" -w /app golang:$GO_VER sh -c 'CGO_ENABLED=0 go build -a --installsuffix cgo --ldflags="-s" -o influxd ./cmd/influxd' + +docker build -t influxdb . diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/circle-test.sh b/Godeps/_workspace/src/github.com/influxdb/influxdb/circle-test.sh new file mode 100644 index 000000000..a53ac6487 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/circle-test.sh @@ -0,0 +1,63 @@ +#!/bin/bash +# +# This is the InfluxDB CircleCI test script. Using this script allows total control +# the environment in which the build and test is run, and matches the official +# build process for InfluxDB. + +BUILD_DIR=$HOME/influxdb-build +GO_VERSION=go1.4.2 +PARALLELISM="-parallel 256" +TIMEOUT="-timeout 480s" + +# Executes the given statement, and exits if the command returns a non-zero code. +function exit_if_fail { + command=$@ + echo "Executing '$command'" + $command + rc=$? + if [ $rc -ne 0 ]; then + echo "'$command' returned $rc." + exit $rc + fi +} + +source $HOME/.gvm/scripts/gvm +exit_if_fail gvm use $GO_VERSION + +# Set up the build directory, and then GOPATH. +exit_if_fail mkdir $BUILD_DIR +export GOPATH=$BUILD_DIR +exit_if_fail mkdir -p $GOPATH/src/github.com/influxdb + +# Dump some test config to the log. +echo "Test configuration" +echo "========================================" +echo "\$HOME: $HOME" +echo "\$GOPATH: $GOPATH" +echo "\$CIRCLE_BRANCH: $CIRCLE_BRANCH" + +# Move the checked-out source to a better location. +exit_if_fail mv $HOME/influxdb $GOPATH/src/github.com/influxdb +exit_if_fail cd $GOPATH/src/github.com/influxdb/influxdb +exit_if_fail git branch --set-upstream-to=origin/$CIRCLE_BRANCH $CIRCLE_BRANCH + +# Install the code. +exit_if_fail cd $GOPATH/src/github.com/influxdb/influxdb +exit_if_fail go get -t -d -v ./... +exit_if_fail git checkout $CIRCLE_BRANCH # 'go get' switches to master. Who knew? Switch back. +exit_if_fail go build -v ./... + +# Run the tests. +exit_if_fail go tool vet --composites=false . +case $CIRCLE_NODE_INDEX in + 0) + go test $PARALLELISM $TIMEOUT -v ./... 2>&1 | tee $CIRCLE_ARTIFACTS/test_logs.txt + rc=${PIPESTATUS[0]} + ;; + 1) + GORACE="halt_on_error=1" go test $PARALLELISM $TIMEOUT -v -race ./... 2>&1 | tee $CIRCLE_ARTIFACTS/test_logs_race.txt + rc=${PIPESTATUS[0]} + ;; +esac + +exit $rc diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/circle.yml b/Godeps/_workspace/src/github.com/influxdb/influxdb/circle.yml new file mode 100644 index 000000000..141e26fda --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/circle.yml @@ -0,0 +1,12 @@ +machine: + pre: + - bash < <(curl -s -S -L https://raw.githubusercontent.com/moovweb/gvm/master/binscripts/gvm-installer) + - source $HOME/.gvm/scripts/gvm; gvm install go1.4.2 --binary + +dependencies: + override: + - echo "Dummy override, so no Circle dependencies execute" +test: + override: + - bash circle-test.sh: + parallel: true diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/client_pool.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/client_pool.go new file mode 100644 index 000000000..fed7e18e0 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/client_pool.go @@ -0,0 +1,57 @@ +package cluster + +import ( + "net" + "sync" + + "gopkg.in/fatih/pool.v2" +) + +type clientPool struct { + mu sync.RWMutex + pool map[uint64]pool.Pool +} + +func newClientPool() *clientPool { + return &clientPool{ + pool: make(map[uint64]pool.Pool), + } +} + +func (c *clientPool) setPool(nodeID uint64, p pool.Pool) { + c.mu.Lock() + c.pool[nodeID] = p + c.mu.Unlock() +} + +func (c *clientPool) getPool(nodeID uint64) (pool.Pool, bool) { + c.mu.RLock() + p, ok := c.pool[nodeID] + c.mu.RUnlock() + return p, ok +} + +func (c *clientPool) size() int { + c.mu.RLock() + var size int + for _, p := range c.pool { + size += p.Len() + } + c.mu.RUnlock() + return size +} + +func (c *clientPool) conn(nodeID uint64) (net.Conn, error) { + c.mu.RLock() + conn, err := c.pool[nodeID].Get() + c.mu.RUnlock() + return conn, err +} + +func (c *clientPool) close() { + c.mu.Lock() + for _, p := range c.pool { + p.Close() + } + c.mu.Unlock() +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/config.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/config.go new file mode 100644 index 000000000..3a67b32d0 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/config.go @@ -0,0 +1,35 @@ +package cluster + +import ( + "time" + + "github.com/influxdb/influxdb/toml" +) + +const ( + // DefaultWriteTimeout is the default timeout for a complete write to succeed. + DefaultWriteTimeout = 5 * time.Second + + // DefaultShardWriterTimeout is the default timeout set on shard writers. + DefaultShardWriterTimeout = 5 * time.Second + + // DefaultShardMapperTimeout is the default timeout set on shard mappers. + DefaultShardMapperTimeout = 5 * time.Second +) + +// Config represents the configuration for the clustering service. +type Config struct { + ForceRemoteShardMapping bool `toml:"force-remote-mapping"` + WriteTimeout toml.Duration `toml:"write-timeout"` + ShardWriterTimeout toml.Duration `toml:"shard-writer-timeout"` + ShardMapperTimeout toml.Duration `toml:"shard-mapper-timeout"` +} + +// NewConfig returns an instance of Config with defaults. +func NewConfig() Config { + return Config{ + WriteTimeout: toml.Duration(DefaultWriteTimeout), + ShardWriterTimeout: toml.Duration(DefaultShardWriterTimeout), + ShardMapperTimeout: toml.Duration(DefaultShardMapperTimeout), + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/config_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/config_test.go new file mode 100644 index 000000000..db5e5ddc1 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/config_test.go @@ -0,0 +1,27 @@ +package cluster_test + +import ( + "testing" + "time" + + "github.com/BurntSushi/toml" + "github.com/influxdb/influxdb/cluster" +) + +func TestConfig_Parse(t *testing.T) { + // Parse configuration. + var c cluster.Config + if _, err := toml.Decode(` +shard-writer-timeout = "10s" +write-timeout = "20s" +`, &c); err != nil { + t.Fatal(err) + } + + // Validate configuration. + if time.Duration(c.ShardWriterTimeout) != 10*time.Second { + t.Fatalf("unexpected shard-writer timeout: %s", c.ShardWriterTimeout) + } else if time.Duration(c.WriteTimeout) != 20*time.Second { + t.Fatalf("unexpected write timeout s: %s", c.WriteTimeout) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/internal/data.pb.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/internal/data.pb.go new file mode 100644 index 000000000..1d53707c8 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/internal/data.pb.go @@ -0,0 +1,286 @@ +// Code generated by protoc-gen-gogo. +// source: internal/data.proto +// DO NOT EDIT! + +/* +Package internal is a generated protocol buffer package. + +It is generated from these files: + internal/data.proto + +It has these top-level messages: + WriteShardRequest + Field + Tag + Point + WriteShardResponse + MapShardRequest + MapShardResponse +*/ +package internal + +import proto "github.com/gogo/protobuf/proto" +import math "math" + +// Reference imports to suppress errors if they are not otherwise used. +var _ = proto.Marshal +var _ = math.Inf + +type WriteShardRequest struct { + ShardID *uint64 `protobuf:"varint,1,req" json:"ShardID,omitempty"` + Points []*Point `protobuf:"bytes,2,rep" json:"Points,omitempty"` + XXX_unrecognized []byte `json:"-"` +} + +func (m *WriteShardRequest) Reset() { *m = WriteShardRequest{} } +func (m *WriteShardRequest) String() string { return proto.CompactTextString(m) } +func (*WriteShardRequest) ProtoMessage() {} + +func (m *WriteShardRequest) GetShardID() uint64 { + if m != nil && m.ShardID != nil { + return *m.ShardID + } + return 0 +} + +func (m *WriteShardRequest) GetPoints() []*Point { + if m != nil { + return m.Points + } + return nil +} + +type Field struct { + Name *string `protobuf:"bytes,1,req" json:"Name,omitempty"` + Int32 *int32 `protobuf:"varint,2,opt" json:"Int32,omitempty"` + Int64 *int64 `protobuf:"varint,3,opt" json:"Int64,omitempty"` + Float64 *float64 `protobuf:"fixed64,4,opt" json:"Float64,omitempty"` + Bool *bool `protobuf:"varint,5,opt" json:"Bool,omitempty"` + String_ *string `protobuf:"bytes,6,opt" json:"String,omitempty"` + Bytes []byte `protobuf:"bytes,7,opt" json:"Bytes,omitempty"` + XXX_unrecognized []byte `json:"-"` +} + +func (m *Field) Reset() { *m = Field{} } +func (m *Field) String() string { return proto.CompactTextString(m) } +func (*Field) ProtoMessage() {} + +func (m *Field) GetName() string { + if m != nil && m.Name != nil { + return *m.Name + } + return "" +} + +func (m *Field) GetInt32() int32 { + if m != nil && m.Int32 != nil { + return *m.Int32 + } + return 0 +} + +func (m *Field) GetInt64() int64 { + if m != nil && m.Int64 != nil { + return *m.Int64 + } + return 0 +} + +func (m *Field) GetFloat64() float64 { + if m != nil && m.Float64 != nil { + return *m.Float64 + } + return 0 +} + +func (m *Field) GetBool() bool { + if m != nil && m.Bool != nil { + return *m.Bool + } + return false +} + +func (m *Field) GetString_() string { + if m != nil && m.String_ != nil { + return *m.String_ + } + return "" +} + +func (m *Field) GetBytes() []byte { + if m != nil { + return m.Bytes + } + return nil +} + +type Tag struct { + Key *string `protobuf:"bytes,1,req" json:"Key,omitempty"` + Value *string `protobuf:"bytes,2,req" json:"Value,omitempty"` + XXX_unrecognized []byte `json:"-"` +} + +func (m *Tag) Reset() { *m = Tag{} } +func (m *Tag) String() string { return proto.CompactTextString(m) } +func (*Tag) ProtoMessage() {} + +func (m *Tag) GetKey() string { + if m != nil && m.Key != nil { + return *m.Key + } + return "" +} + +func (m *Tag) GetValue() string { + if m != nil && m.Value != nil { + return *m.Value + } + return "" +} + +type Point struct { + Name *string `protobuf:"bytes,1,req" json:"Name,omitempty"` + Time *int64 `protobuf:"varint,2,req" json:"Time,omitempty"` + Fields []*Field `protobuf:"bytes,3,rep" json:"Fields,omitempty"` + Tags []*Tag `protobuf:"bytes,4,rep" json:"Tags,omitempty"` + XXX_unrecognized []byte `json:"-"` +} + +func (m *Point) Reset() { *m = Point{} } +func (m *Point) String() string { return proto.CompactTextString(m) } +func (*Point) ProtoMessage() {} + +func (m *Point) GetName() string { + if m != nil && m.Name != nil { + return *m.Name + } + return "" +} + +func (m *Point) GetTime() int64 { + if m != nil && m.Time != nil { + return *m.Time + } + return 0 +} + +func (m *Point) GetFields() []*Field { + if m != nil { + return m.Fields + } + return nil +} + +func (m *Point) GetTags() []*Tag { + if m != nil { + return m.Tags + } + return nil +} + +type WriteShardResponse struct { + Code *int32 `protobuf:"varint,1,req" json:"Code,omitempty"` + Message *string `protobuf:"bytes,2,opt" json:"Message,omitempty"` + XXX_unrecognized []byte `json:"-"` +} + +func (m *WriteShardResponse) Reset() { *m = WriteShardResponse{} } +func (m *WriteShardResponse) String() string { return proto.CompactTextString(m) } +func (*WriteShardResponse) ProtoMessage() {} + +func (m *WriteShardResponse) GetCode() int32 { + if m != nil && m.Code != nil { + return *m.Code + } + return 0 +} + +func (m *WriteShardResponse) GetMessage() string { + if m != nil && m.Message != nil { + return *m.Message + } + return "" +} + +type MapShardRequest struct { + ShardID *uint64 `protobuf:"varint,1,req" json:"ShardID,omitempty"` + Query *string `protobuf:"bytes,2,req" json:"Query,omitempty"` + ChunkSize *int32 `protobuf:"varint,3,req" json:"ChunkSize,omitempty"` + XXX_unrecognized []byte `json:"-"` +} + +func (m *MapShardRequest) Reset() { *m = MapShardRequest{} } +func (m *MapShardRequest) String() string { return proto.CompactTextString(m) } +func (*MapShardRequest) ProtoMessage() {} + +func (m *MapShardRequest) GetShardID() uint64 { + if m != nil && m.ShardID != nil { + return *m.ShardID + } + return 0 +} + +func (m *MapShardRequest) GetQuery() string { + if m != nil && m.Query != nil { + return *m.Query + } + return "" +} + +func (m *MapShardRequest) GetChunkSize() int32 { + if m != nil && m.ChunkSize != nil { + return *m.ChunkSize + } + return 0 +} + +type MapShardResponse struct { + Code *int32 `protobuf:"varint,1,req" json:"Code,omitempty"` + Message *string `protobuf:"bytes,2,opt" json:"Message,omitempty"` + Data []byte `protobuf:"bytes,3,opt" json:"Data,omitempty"` + TagSets []string `protobuf:"bytes,4,rep" json:"TagSets,omitempty"` + Fields []string `protobuf:"bytes,5,rep" json:"Fields,omitempty"` + XXX_unrecognized []byte `json:"-"` +} + +func (m *MapShardResponse) Reset() { *m = MapShardResponse{} } +func (m *MapShardResponse) String() string { return proto.CompactTextString(m) } +func (*MapShardResponse) ProtoMessage() {} + +func (m *MapShardResponse) GetCode() int32 { + if m != nil && m.Code != nil { + return *m.Code + } + return 0 +} + +func (m *MapShardResponse) GetMessage() string { + if m != nil && m.Message != nil { + return *m.Message + } + return "" +} + +func (m *MapShardResponse) GetData() []byte { + if m != nil { + return m.Data + } + return nil +} + +func (m *MapShardResponse) GetTagSets() []string { + if m != nil { + return m.TagSets + } + return nil +} + +func (m *MapShardResponse) GetFields() []string { + if m != nil { + return m.Fields + } + return nil +} + +func init() { +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/internal/data.proto b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/internal/data.proto new file mode 100644 index 000000000..e38a9de5b --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/internal/data.proto @@ -0,0 +1,49 @@ +package internal; + +message WriteShardRequest { + required uint64 ShardID = 1; + repeated Point Points = 2; +} + +message Field { + required string Name = 1; + oneof Value { + int32 Int32 = 2; + int64 Int64 = 3; + double Float64 = 4; + bool Bool = 5; + string String = 6; + bytes Bytes = 7; + } +} + +message Tag { + required string Key = 1; + required string Value = 2; +} + +message Point { + required string Name = 1; + required int64 Time = 2; + repeated Field Fields = 3; + repeated Tag Tags = 4; +} + +message WriteShardResponse { + required int32 Code = 1; + optional string Message = 2; +} + +message MapShardRequest { + required uint64 ShardID = 1; + required string Query = 2; + required int32 ChunkSize = 3; +} + +message MapShardResponse { + required int32 Code = 1; + optional string Message = 2; + optional bytes Data = 3; + repeated string TagSets = 4; + repeated string Fields = 5; +} \ No newline at end of file diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/points_writer.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/points_writer.go new file mode 100644 index 000000000..b5a78636d --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/points_writer.go @@ -0,0 +1,314 @@ +package cluster + +import ( + "errors" + "fmt" + "log" + "os" + "strings" + "sync" + "time" + + "github.com/influxdb/influxdb" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/tsdb" +) + +// ConsistencyLevel represent a required replication criteria before a write can +// be returned as successful +type ConsistencyLevel int + +const ( + // ConsistencyLevelAny allows for hinted hand off, potentially no write happened yet + ConsistencyLevelAny ConsistencyLevel = iota + + // ConsistencyLevelOne requires at least one data node acknowledged a write + ConsistencyLevelOne + + // ConsistencyLevelOne requires a quorum of data nodes to acknowledge a write + ConsistencyLevelQuorum + + // ConsistencyLevelAll requires all data nodes to acknowledge a write + ConsistencyLevelAll +) + +var ( + // ErrTimeout is returned when a write times out. + ErrTimeout = errors.New("timeout") + + // ErrPartialWrite is returned when a write partially succeeds but does + // not meet the requested consistency level. + ErrPartialWrite = errors.New("partial write") + + // ErrWriteFailed is returned when no writes succeeded. + ErrWriteFailed = errors.New("write failed") + + // ErrInvalidConsistencyLevel is returned when parsing the string version + // of a consistency level. + ErrInvalidConsistencyLevel = errors.New("invalid consistency level") +) + +func ParseConsistencyLevel(level string) (ConsistencyLevel, error) { + switch strings.ToLower(level) { + case "any": + return ConsistencyLevelAny, nil + case "one": + return ConsistencyLevelOne, nil + case "quorum": + return ConsistencyLevelQuorum, nil + case "all": + return ConsistencyLevelAll, nil + default: + return 0, ErrInvalidConsistencyLevel + } +} + +// PointsWriter handles writes across multiple local and remote data nodes. +type PointsWriter struct { + mu sync.RWMutex + closing chan struct{} + WriteTimeout time.Duration + Logger *log.Logger + + MetaStore interface { + NodeID() uint64 + Database(name string) (di *meta.DatabaseInfo, err error) + RetentionPolicy(database, policy string) (*meta.RetentionPolicyInfo, error) + CreateShardGroupIfNotExists(database, policy string, timestamp time.Time) (*meta.ShardGroupInfo, error) + ShardOwner(shardID uint64) (string, string, *meta.ShardGroupInfo) + } + + TSDBStore interface { + CreateShard(database, retentionPolicy string, shardID uint64) error + WriteToShard(shardID uint64, points []tsdb.Point) error + } + + ShardWriter interface { + WriteShard(shardID, ownerID uint64, points []tsdb.Point) error + } + + HintedHandoff interface { + WriteShard(shardID, ownerID uint64, points []tsdb.Point) error + } +} + +// NewPointsWriter returns a new instance of PointsWriter for a node. +func NewPointsWriter() *PointsWriter { + return &PointsWriter{ + closing: make(chan struct{}), + WriteTimeout: DefaultWriteTimeout, + Logger: log.New(os.Stderr, "[write] ", log.LstdFlags), + } +} + +// ShardMapping contains a mapping of a shards to a points. +type ShardMapping struct { + Points map[uint64][]tsdb.Point // The points associated with a shard ID + Shards map[uint64]*meta.ShardInfo // The shards that have been mapped, keyed by shard ID +} + +// NewShardMapping creates an empty ShardMapping +func NewShardMapping() *ShardMapping { + return &ShardMapping{ + Points: map[uint64][]tsdb.Point{}, + Shards: map[uint64]*meta.ShardInfo{}, + } +} + +// MapPoint maps a point to shard +func (s *ShardMapping) MapPoint(shardInfo *meta.ShardInfo, p tsdb.Point) { + points, ok := s.Points[shardInfo.ID] + if !ok { + s.Points[shardInfo.ID] = []tsdb.Point{p} + } else { + s.Points[shardInfo.ID] = append(points, p) + } + s.Shards[shardInfo.ID] = shardInfo +} + +func (w *PointsWriter) Open() error { + w.mu.Lock() + defer w.mu.Unlock() + if w.closing == nil { + w.closing = make(chan struct{}) + } + return nil +} + +func (w *PointsWriter) Close() error { + w.mu.Lock() + defer w.mu.Unlock() + if w.closing != nil { + close(w.closing) + w.closing = nil + } + return nil +} + +// MapShards maps the points contained in wp to a ShardMapping. If a point +// maps to a shard group or shard that does not currently exist, it will be +// created before returning the mapping. +func (w *PointsWriter) MapShards(wp *WritePointsRequest) (*ShardMapping, error) { + + // holds the start time ranges for required shard groups + timeRanges := map[time.Time]*meta.ShardGroupInfo{} + + rp, err := w.MetaStore.RetentionPolicy(wp.Database, wp.RetentionPolicy) + if err != nil { + return nil, err + } + + for _, p := range wp.Points { + timeRanges[p.Time().Truncate(rp.ShardGroupDuration)] = nil + } + + // holds all the shard groups and shards that are required for writes + for t := range timeRanges { + sg, err := w.MetaStore.CreateShardGroupIfNotExists(wp.Database, wp.RetentionPolicy, t) + if err != nil { + return nil, err + } + timeRanges[t] = sg + } + + mapping := NewShardMapping() + for _, p := range wp.Points { + sg := timeRanges[p.Time().Truncate(rp.ShardGroupDuration)] + sh := sg.ShardFor(p.HashID()) + mapping.MapPoint(&sh, p) + } + return mapping, nil +} + +// WritePoints writes across multiple local and remote data nodes according the consistency level. +func (w *PointsWriter) WritePoints(p *WritePointsRequest) error { + if p.RetentionPolicy == "" { + db, err := w.MetaStore.Database(p.Database) + if err != nil { + return err + } else if db == nil { + return influxdb.ErrDatabaseNotFound(p.Database) + } + p.RetentionPolicy = db.DefaultRetentionPolicy + } + + shardMappings, err := w.MapShards(p) + if err != nil { + return err + } + + // Write each shard in it's own goroutine and return as soon + // as one fails. + ch := make(chan error, len(shardMappings.Points)) + for shardID, points := range shardMappings.Points { + go func(shard *meta.ShardInfo, database, retentionPolicy string, points []tsdb.Point) { + ch <- w.writeToShard(shard, p.Database, p.RetentionPolicy, p.ConsistencyLevel, points) + }(shardMappings.Shards[shardID], p.Database, p.RetentionPolicy, points) + } + + for range shardMappings.Points { + select { + case <-w.closing: + return ErrWriteFailed + case err := <-ch: + if err != nil { + return err + } + } + } + return nil +} + +// writeToShards writes points to a shard and ensures a write consistency level has been met. If the write +// partially succeeds, ErrPartialWrite is returned. +func (w *PointsWriter) writeToShard(shard *meta.ShardInfo, database, retentionPolicy string, + consistency ConsistencyLevel, points []tsdb.Point) error { + // The required number of writes to achieve the requested consistency level + required := len(shard.OwnerIDs) + switch consistency { + case ConsistencyLevelAny, ConsistencyLevelOne: + required = 1 + case ConsistencyLevelQuorum: + required = required/2 + 1 + } + + // response channel for each shard writer go routine + ch := make(chan error, len(shard.OwnerIDs)) + + for _, nodeID := range shard.OwnerIDs { + go func(shardID, nodeID uint64, points []tsdb.Point) { + if w.MetaStore.NodeID() == nodeID { + err := w.TSDBStore.WriteToShard(shardID, points) + // If we've written to shard that should exist on the current node, but the store has + // not actually created this shard, tell it to create it and retry the write + if err == tsdb.ErrShardNotFound { + err = w.TSDBStore.CreateShard(database, retentionPolicy, shardID) + if err != nil { + ch <- err + return + } + err = w.TSDBStore.WriteToShard(shardID, points) + } + ch <- err + return + } + + err := w.ShardWriter.WriteShard(shardID, nodeID, points) + if err != nil && tsdb.IsRetryable(err) { + // The remote write failed so queue it via hinted handoff + hherr := w.HintedHandoff.WriteShard(shardID, nodeID, points) + + // If the write consistency level is ANY, then a successful hinted handoff can + // be considered a successful write so send nil to the response channel + // otherwise, let the original error propogate to the response channel + if hherr == nil && consistency == ConsistencyLevelAny { + ch <- nil + return + } + } + ch <- err + + }(shard.ID, nodeID, points) + } + + var wrote int + timeout := time.After(w.WriteTimeout) + var writeError error + for _, nodeID := range shard.OwnerIDs { + select { + case <-w.closing: + return ErrWriteFailed + case <-timeout: + // return timeout error to caller + return ErrTimeout + case err := <-ch: + // If the write returned an error, continue to the next response + if err != nil { + w.Logger.Printf("write failed for shard %d on node %d: %v", shard.ID, nodeID, err) + + // Keep track of the first error we see to return back to the client + if writeError == nil { + writeError = err + } + continue + } + + wrote += 1 + } + } + + // We wrote the required consistency level + if wrote >= required { + return nil + } + + if wrote > 0 { + return ErrPartialWrite + } + + if writeError != nil { + return fmt.Errorf("write failed: %v", writeError) + } + + return ErrWriteFailed +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/points_writer_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/points_writer_test.go new file mode 100644 index 000000000..321f731d0 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/points_writer_test.go @@ -0,0 +1,436 @@ +package cluster_test + +import ( + "fmt" + "sync" + "sync/atomic" + "testing" + "time" + + "github.com/influxdb/influxdb/cluster" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/tsdb" +) + +// Ensures the points writer maps a single point to a single shard. +func TestPointsWriter_MapShards_One(t *testing.T) { + ms := MetaStore{} + rp := NewRetentionPolicy("myp", time.Hour, 3) + + ms.NodeIDFn = func() uint64 { return 1 } + ms.RetentionPolicyFn = func(db, retentionPolicy string) (*meta.RetentionPolicyInfo, error) { + return rp, nil + } + + ms.CreateShardGroupIfNotExistsFn = func(database, policy string, timestamp time.Time) (*meta.ShardGroupInfo, error) { + return &rp.ShardGroups[0], nil + } + + c := cluster.PointsWriter{MetaStore: ms} + pr := &cluster.WritePointsRequest{ + Database: "mydb", + RetentionPolicy: "myrp", + ConsistencyLevel: cluster.ConsistencyLevelOne, + } + pr.AddPoint("cpu", 1.0, time.Now(), nil) + + var ( + shardMappings *cluster.ShardMapping + err error + ) + if shardMappings, err = c.MapShards(pr); err != nil { + t.Fatalf("unexpected an error: %v", err) + } + + if exp := 1; len(shardMappings.Points) != exp { + t.Errorf("MapShards() len mismatch. got %v, exp %v", len(shardMappings.Points), exp) + } +} + +// Ensures the points writer maps a multiple points across shard group boundaries. +func TestPointsWriter_MapShards_Multiple(t *testing.T) { + ms := MetaStore{} + rp := NewRetentionPolicy("myp", time.Hour, 3) + AttachShardGroupInfo(rp, []uint64{1, 2, 3}) + AttachShardGroupInfo(rp, []uint64{1, 2, 3}) + + ms.NodeIDFn = func() uint64 { return 1 } + ms.RetentionPolicyFn = func(db, retentionPolicy string) (*meta.RetentionPolicyInfo, error) { + return rp, nil + } + + ms.CreateShardGroupIfNotExistsFn = func(database, policy string, timestamp time.Time) (*meta.ShardGroupInfo, error) { + for i, sg := range rp.ShardGroups { + if timestamp.Equal(sg.StartTime) || timestamp.After(sg.StartTime) && timestamp.Before(sg.EndTime) { + return &rp.ShardGroups[i], nil + } + } + panic("should not get here") + } + + c := cluster.PointsWriter{MetaStore: ms} + pr := &cluster.WritePointsRequest{ + Database: "mydb", + RetentionPolicy: "myrp", + ConsistencyLevel: cluster.ConsistencyLevelOne, + } + + // Three points that range over the shardGroup duration (1h) and should map to two + // distinct shards + pr.AddPoint("cpu", 1.0, time.Unix(0, 0), nil) + pr.AddPoint("cpu", 2.0, time.Unix(0, 0).Add(time.Hour), nil) + pr.AddPoint("cpu", 3.0, time.Unix(0, 0).Add(time.Hour+time.Second), nil) + + var ( + shardMappings *cluster.ShardMapping + err error + ) + if shardMappings, err = c.MapShards(pr); err != nil { + t.Fatalf("unexpected an error: %v", err) + } + + if exp := 2; len(shardMappings.Points) != exp { + t.Errorf("MapShards() len mismatch. got %v, exp %v", len(shardMappings.Points), exp) + } + + for _, points := range shardMappings.Points { + // First shard shoud have 1 point w/ first point added + if len(points) == 1 && points[0].Time() != pr.Points[0].Time() { + t.Fatalf("MapShards() value mismatch. got %v, exp %v", points[0].Time(), pr.Points[0].Time()) + } + + // Second shard shoud have the last two points added + if len(points) == 2 && points[0].Time() != pr.Points[1].Time() { + t.Fatalf("MapShards() value mismatch. got %v, exp %v", points[0].Time(), pr.Points[1].Time()) + } + + if len(points) == 2 && points[1].Time() != pr.Points[2].Time() { + t.Fatalf("MapShards() value mismatch. got %v, exp %v", points[1].Time(), pr.Points[2].Time()) + } + } +} + +func TestPointsWriter_WritePoints(t *testing.T) { + tests := []struct { + name string + database string + retentionPolicy string + consistency cluster.ConsistencyLevel + + // the responses returned by each shard write call. node ID 1 = pos 0 + err []error + expErr error + }{ + // Consistency one + { + name: "write one success", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelOne, + err: []error{nil, nil, nil}, + expErr: nil, + }, + { + name: "write one error", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelOne, + err: []error{fmt.Errorf("a failure"), fmt.Errorf("a failure"), fmt.Errorf("a failure")}, + expErr: fmt.Errorf("write failed: a failure"), + }, + + // Consistency any + { + name: "write any success", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelAny, + err: []error{fmt.Errorf("a failure"), nil, fmt.Errorf("a failure")}, + expErr: nil, + }, + // Consistency all + { + name: "write all success", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelAll, + err: []error{nil, nil, nil}, + expErr: nil, + }, + { + name: "write all, 2/3, partial write", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelAll, + err: []error{nil, fmt.Errorf("a failure"), nil}, + expErr: cluster.ErrPartialWrite, + }, + { + name: "write all, 1/3 (failure)", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelAll, + err: []error{nil, fmt.Errorf("a failure"), fmt.Errorf("a failure")}, + expErr: cluster.ErrPartialWrite, + }, + + // Consistency quorum + { + name: "write quorum, 1/3 failure", + consistency: cluster.ConsistencyLevelQuorum, + database: "mydb", + retentionPolicy: "myrp", + err: []error{fmt.Errorf("a failure"), fmt.Errorf("a failure"), nil}, + expErr: cluster.ErrPartialWrite, + }, + { + name: "write quorum, 2/3 success", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelQuorum, + err: []error{nil, nil, fmt.Errorf("a failure")}, + expErr: nil, + }, + { + name: "write quorum, 3/3 success", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelQuorum, + err: []error{nil, nil, nil}, + expErr: nil, + }, + + // Error write error + { + name: "no writes succeed", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelOne, + err: []error{fmt.Errorf("a failure"), fmt.Errorf("a failure"), fmt.Errorf("a failure")}, + expErr: fmt.Errorf("write failed: a failure"), + }, + + // Hinted handoff w/ ANY + { + name: "hinted handoff write succeed", + database: "mydb", + retentionPolicy: "myrp", + consistency: cluster.ConsistencyLevelAny, + err: []error{fmt.Errorf("a failure"), fmt.Errorf("a failure"), fmt.Errorf("a failure")}, + expErr: nil, + }, + + // Write to non-existant database + { + name: "write to non-existant database", + database: "doesnt_exist", + retentionPolicy: "", + consistency: cluster.ConsistencyLevelAny, + err: []error{nil, nil, nil}, + expErr: fmt.Errorf("database not found: doesnt_exist"), + }, + } + + for _, test := range tests { + + pr := &cluster.WritePointsRequest{ + Database: test.database, + RetentionPolicy: test.retentionPolicy, + ConsistencyLevel: test.consistency, + } + + // Three points that range over the shardGroup duration (1h) and should map to two + // distinct shards + pr.AddPoint("cpu", 1.0, time.Unix(0, 0), nil) + pr.AddPoint("cpu", 2.0, time.Unix(0, 0).Add(time.Hour), nil) + pr.AddPoint("cpu", 3.0, time.Unix(0, 0).Add(time.Hour+time.Second), nil) + + // copy to prevent data race + theTest := test + sm := cluster.NewShardMapping() + sm.MapPoint( + &meta.ShardInfo{ID: uint64(1), OwnerIDs: []uint64{uint64(1), uint64(2), uint64(3)}}, + pr.Points[0]) + sm.MapPoint( + &meta.ShardInfo{ID: uint64(2), OwnerIDs: []uint64{uint64(1), uint64(2), uint64(3)}}, + pr.Points[1]) + sm.MapPoint( + &meta.ShardInfo{ID: uint64(2), OwnerIDs: []uint64{uint64(1), uint64(2), uint64(3)}}, + pr.Points[2]) + + // Local cluster.Node ShardWriter + // lock on the write increment since these functions get called in parallel + var mu sync.Mutex + sw := &fakeShardWriter{ + ShardWriteFn: func(shardID, nodeID uint64, points []tsdb.Point) error { + mu.Lock() + defer mu.Unlock() + return theTest.err[int(nodeID)-1] + }, + } + + store := &fakeStore{ + WriteFn: func(shardID uint64, points []tsdb.Point) error { + mu.Lock() + defer mu.Unlock() + return theTest.err[0] + }, + } + + hh := &fakeShardWriter{ + ShardWriteFn: func(shardID, nodeID uint64, points []tsdb.Point) error { + return nil + }, + } + + ms := NewMetaStore() + ms.DatabaseFn = func(database string) (*meta.DatabaseInfo, error) { + return nil, nil + } + ms.NodeIDFn = func() uint64 { return 1 } + c := cluster.NewPointsWriter() + c.MetaStore = ms + c.ShardWriter = sw + c.TSDBStore = store + c.HintedHandoff = hh + + err := c.WritePoints(pr) + if err == nil && test.expErr != nil { + t.Errorf("PointsWriter.WritePoints(): '%s' error: got %v, exp %v", test.name, err, test.expErr) + } + + if err != nil && test.expErr == nil { + t.Errorf("PointsWriter.WritePoints(): '%s' error: got %v, exp %v", test.name, err, test.expErr) + } + if err != nil && test.expErr != nil && err.Error() != test.expErr.Error() { + t.Errorf("PointsWriter.WritePoints(): '%s' error: got %v, exp %v", test.name, err, test.expErr) + } + } +} + +var shardID uint64 + +type fakeShardWriter struct { + ShardWriteFn func(shardID, nodeID uint64, points []tsdb.Point) error +} + +func (f *fakeShardWriter) WriteShard(shardID, nodeID uint64, points []tsdb.Point) error { + return f.ShardWriteFn(shardID, nodeID, points) +} + +type fakeStore struct { + WriteFn func(shardID uint64, points []tsdb.Point) error + CreateShardfn func(database, retentionPolicy string, shardID uint64) error +} + +func (f *fakeStore) WriteToShard(shardID uint64, points []tsdb.Point) error { + return f.WriteFn(shardID, points) +} + +func (f *fakeStore) CreateShard(database, retentionPolicy string, shardID uint64) error { + return f.CreateShardfn(database, retentionPolicy, shardID) +} + +func NewMetaStore() *MetaStore { + ms := &MetaStore{} + rp := NewRetentionPolicy("myp", time.Hour, 3) + AttachShardGroupInfo(rp, []uint64{1, 2, 3}) + AttachShardGroupInfo(rp, []uint64{1, 2, 3}) + + ms.RetentionPolicyFn = func(db, retentionPolicy string) (*meta.RetentionPolicyInfo, error) { + return rp, nil + } + + ms.CreateShardGroupIfNotExistsFn = func(database, policy string, timestamp time.Time) (*meta.ShardGroupInfo, error) { + for i, sg := range rp.ShardGroups { + if timestamp.Equal(sg.StartTime) || timestamp.After(sg.StartTime) && timestamp.Before(sg.EndTime) { + return &rp.ShardGroups[i], nil + } + } + panic("should not get here") + } + return ms +} + +type MetaStore struct { + NodeIDFn func() uint64 + RetentionPolicyFn func(database, name string) (*meta.RetentionPolicyInfo, error) + CreateShardGroupIfNotExistsFn func(database, policy string, timestamp time.Time) (*meta.ShardGroupInfo, error) + DatabaseFn func(database string) (*meta.DatabaseInfo, error) + ShardOwnerFn func(shardID uint64) (string, string, *meta.ShardGroupInfo) +} + +func (m MetaStore) NodeID() uint64 { return m.NodeIDFn() } + +func (m MetaStore) RetentionPolicy(database, name string) (*meta.RetentionPolicyInfo, error) { + return m.RetentionPolicyFn(database, name) +} + +func (m MetaStore) CreateShardGroupIfNotExists(database, policy string, timestamp time.Time) (*meta.ShardGroupInfo, error) { + return m.CreateShardGroupIfNotExistsFn(database, policy, timestamp) +} + +func (m MetaStore) Database(database string) (*meta.DatabaseInfo, error) { + return m.DatabaseFn(database) +} + +func (m MetaStore) ShardOwner(shardID uint64) (string, string, *meta.ShardGroupInfo) { + return m.ShardOwnerFn(shardID) +} + +func NewRetentionPolicy(name string, duration time.Duration, nodeCount int) *meta.RetentionPolicyInfo { + shards := []meta.ShardInfo{} + ownerIDs := []uint64{} + for i := 1; i <= nodeCount; i++ { + ownerIDs = append(ownerIDs, uint64(i)) + } + + // each node is fully replicated with each other + shards = append(shards, meta.ShardInfo{ + ID: nextShardID(), + OwnerIDs: ownerIDs, + }) + + rp := &meta.RetentionPolicyInfo{ + Name: "myrp", + ReplicaN: nodeCount, + Duration: duration, + ShardGroupDuration: duration, + ShardGroups: []meta.ShardGroupInfo{ + meta.ShardGroupInfo{ + ID: nextShardID(), + StartTime: time.Unix(0, 0), + EndTime: time.Unix(0, 0).Add(duration).Add(-1), + Shards: shards, + }, + }, + } + return rp +} + +func AttachShardGroupInfo(rp *meta.RetentionPolicyInfo, ownerIDs []uint64) { + var startTime, endTime time.Time + if len(rp.ShardGroups) == 0 { + startTime = time.Unix(0, 0) + } else { + startTime = rp.ShardGroups[len(rp.ShardGroups)-1].StartTime.Add(rp.ShardGroupDuration) + } + endTime = startTime.Add(rp.ShardGroupDuration).Add(-1) + + sh := meta.ShardGroupInfo{ + ID: uint64(len(rp.ShardGroups) + 1), + StartTime: startTime, + EndTime: endTime, + Shards: []meta.ShardInfo{ + meta.ShardInfo{ + ID: nextShardID(), + OwnerIDs: ownerIDs, + }, + }, + } + rp.ShardGroups = append(rp.ShardGroups, sh) +} + +func nextShardID() uint64 { + return atomic.AddUint64(&shardID, 1) +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/rpc.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/rpc.go new file mode 100644 index 000000000..f3eca790c --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/rpc.go @@ -0,0 +1,229 @@ +package cluster + +import ( + "time" + + "github.com/gogo/protobuf/proto" + "github.com/influxdb/influxdb/cluster/internal" + "github.com/influxdb/influxdb/tsdb" +) + +//go:generate protoc --gogo_out=. internal/data.proto + +// MapShardRequest represents the request to map a remote shard for a query. +type MapShardRequest struct { + pb internal.MapShardRequest +} + +func (m *MapShardRequest) ShardID() uint64 { return m.pb.GetShardID() } +func (m *MapShardRequest) Query() string { return m.pb.GetQuery() } +func (m *MapShardRequest) ChunkSize() int32 { return m.pb.GetChunkSize() } + +func (m *MapShardRequest) SetShardID(id uint64) { m.pb.ShardID = &id } +func (m *MapShardRequest) SetQuery(query string) { m.pb.Query = &query } +func (m *MapShardRequest) SetChunkSize(chunkSize int32) { m.pb.ChunkSize = &chunkSize } + +// MarshalBinary encodes the object to a binary format. +func (m *MapShardRequest) MarshalBinary() ([]byte, error) { + return proto.Marshal(&m.pb) +} + +// UnmarshalBinary populates MapShardRequest from a binary format. +func (m *MapShardRequest) UnmarshalBinary(buf []byte) error { + if err := proto.Unmarshal(buf, &m.pb); err != nil { + return err + } + return nil +} + +// MapShardResponse represents the response returned from a remote MapShardRequest call +type MapShardResponse struct { + pb internal.MapShardResponse +} + +func NewMapShardResponse(code int, message string) *MapShardResponse { + m := &MapShardResponse{} + m.SetCode(code) + m.SetMessage(message) + return m +} + +func (r *MapShardResponse) Code() int { return int(r.pb.GetCode()) } +func (r *MapShardResponse) Message() string { return r.pb.GetMessage() } +func (r *MapShardResponse) TagSets() []string { return r.pb.GetTagSets() } +func (r *MapShardResponse) Fields() []string { return r.pb.GetFields() } +func (r *MapShardResponse) Data() []byte { return r.pb.GetData() } + +func (r *MapShardResponse) SetCode(code int) { r.pb.Code = proto.Int32(int32(code)) } +func (r *MapShardResponse) SetMessage(message string) { r.pb.Message = &message } +func (r *MapShardResponse) SetTagSets(tagsets []string) { r.pb.TagSets = tagsets } +func (r *MapShardResponse) SetFields(fields []string) { r.pb.Fields = fields } +func (r *MapShardResponse) SetData(data []byte) { r.pb.Data = data } + +// MarshalBinary encodes the object to a binary format. +func (r *MapShardResponse) MarshalBinary() ([]byte, error) { + return proto.Marshal(&r.pb) +} + +// UnmarshalBinary populates WritePointRequest from a binary format. +func (r *MapShardResponse) UnmarshalBinary(buf []byte) error { + if err := proto.Unmarshal(buf, &r.pb); err != nil { + return err + } + return nil +} + +// WritePointsRequest represents a request to write point data to the cluster +type WritePointsRequest struct { + Database string + RetentionPolicy string + ConsistencyLevel ConsistencyLevel + Points []tsdb.Point +} + +// AddPoint adds a point to the WritePointRequest with field name 'value' +func (w *WritePointsRequest) AddPoint(name string, value interface{}, timestamp time.Time, tags map[string]string) { + w.Points = append(w.Points, tsdb.NewPoint( + name, tags, map[string]interface{}{"value": value}, timestamp, + )) +} + +// WriteShardRequest represents the a request to write a slice of points to a shard +type WriteShardRequest struct { + pb internal.WriteShardRequest +} + +// WriteShardResponse represents the response returned from a remote WriteShardRequest call +type WriteShardResponse struct { + pb internal.WriteShardResponse +} + +func (w *WriteShardRequest) SetShardID(id uint64) { w.pb.ShardID = &id } +func (w *WriteShardRequest) ShardID() uint64 { return w.pb.GetShardID() } + +func (w *WriteShardRequest) Points() []tsdb.Point { return w.unmarshalPoints() } + +func (w *WriteShardRequest) AddPoint(name string, value interface{}, timestamp time.Time, tags map[string]string) { + w.AddPoints([]tsdb.Point{tsdb.NewPoint( + name, tags, map[string]interface{}{"value": value}, timestamp, + )}) +} + +func (w *WriteShardRequest) AddPoints(points []tsdb.Point) { + w.pb.Points = append(w.pb.Points, w.marshalPoints(points)...) +} + +// MarshalBinary encodes the object to a binary format. +func (w *WriteShardRequest) MarshalBinary() ([]byte, error) { + return proto.Marshal(&w.pb) +} + +func (w *WriteShardRequest) marshalPoints(points []tsdb.Point) []*internal.Point { + pts := make([]*internal.Point, len(points)) + for i, p := range points { + fields := []*internal.Field{} + for k, v := range p.Fields() { + name := k + f := &internal.Field{ + Name: &name, + } + switch t := v.(type) { + case int: + f.Int64 = proto.Int64(int64(t)) + case int32: + f.Int32 = proto.Int32(t) + case int64: + f.Int64 = proto.Int64(t) + case float64: + f.Float64 = proto.Float64(t) + case bool: + f.Bool = proto.Bool(t) + case string: + f.String_ = proto.String(t) + case []byte: + f.Bytes = t + } + fields = append(fields, f) + } + + tags := []*internal.Tag{} + for k, v := range p.Tags() { + key := k + value := v + tags = append(tags, &internal.Tag{ + Key: &key, + Value: &value, + }) + } + name := p.Name() + pts[i] = &internal.Point{ + Name: &name, + Time: proto.Int64(p.Time().UnixNano()), + Fields: fields, + Tags: tags, + } + + } + return pts +} + +// UnmarshalBinary populates WritePointRequest from a binary format. +func (w *WriteShardRequest) UnmarshalBinary(buf []byte) error { + if err := proto.Unmarshal(buf, &w.pb); err != nil { + return err + } + return nil +} + +func (w *WriteShardRequest) unmarshalPoints() []tsdb.Point { + points := make([]tsdb.Point, len(w.pb.GetPoints())) + for i, p := range w.pb.GetPoints() { + pt := tsdb.NewPoint( + p.GetName(), map[string]string{}, + map[string]interface{}{}, time.Unix(0, p.GetTime())) + + for _, f := range p.GetFields() { + n := f.GetName() + if f.Int32 != nil { + pt.AddField(n, f.GetInt32()) + } else if f.Int64 != nil { + pt.AddField(n, f.GetInt64()) + } else if f.Float64 != nil { + pt.AddField(n, f.GetFloat64()) + } else if f.Bool != nil { + pt.AddField(n, f.GetBool()) + } else if f.String_ != nil { + pt.AddField(n, f.GetString_()) + } else { + pt.AddField(n, f.GetBytes()) + } + } + + tags := tsdb.Tags{} + for _, t := range p.GetTags() { + tags[t.GetKey()] = t.GetValue() + } + pt.SetTags(tags) + points[i] = pt + } + return points +} + +func (w *WriteShardResponse) SetCode(code int) { w.pb.Code = proto.Int32(int32(code)) } +func (w *WriteShardResponse) SetMessage(message string) { w.pb.Message = &message } + +func (w *WriteShardResponse) Code() int { return int(w.pb.GetCode()) } +func (w *WriteShardResponse) Message() string { return w.pb.GetMessage() } + +// MarshalBinary encodes the object to a binary format. +func (w *WriteShardResponse) MarshalBinary() ([]byte, error) { + return proto.Marshal(&w.pb) +} + +// UnmarshalBinary populates WritePointRequest from a binary format. +func (w *WriteShardResponse) UnmarshalBinary(buf []byte) error { + if err := proto.Unmarshal(buf, &w.pb); err != nil { + return err + } + return nil +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/rpc_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/rpc_test.go new file mode 100644 index 000000000..4e42cd5d6 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/rpc_test.go @@ -0,0 +1,110 @@ +package cluster + +import ( + "testing" + "time" +) + +func TestWriteShardRequestBinary(t *testing.T) { + sr := &WriteShardRequest{} + + sr.SetShardID(uint64(1)) + if exp := uint64(1); sr.ShardID() != exp { + t.Fatalf("ShardID mismatch: got %v, exp %v", sr.ShardID(), exp) + } + + sr.AddPoint("cpu", 1.0, time.Unix(0, 0), map[string]string{"host": "serverA"}) + sr.AddPoint("cpu", 2.0, time.Unix(0, 0).Add(time.Hour), nil) + sr.AddPoint("cpu_load", 3.0, time.Unix(0, 0).Add(time.Hour+time.Second), nil) + + b, err := sr.MarshalBinary() + if err != nil { + t.Fatalf("WritePointsRequest.MarshalBinary() failed: %v", err) + } + if len(b) == 0 { + t.Fatalf("WritePointsRequest.MarshalBinary() returned 0 bytes") + } + + got := &WriteShardRequest{} + if err := got.UnmarshalBinary(b); err != nil { + t.Fatalf("WritePointsRequest.UnmarshalMarshalBinary() failed: %v", err) + } + + if got.ShardID() != sr.ShardID() { + t.Errorf("ShardID mismatch: got %v, exp %v", got.ShardID(), sr.ShardID()) + } + + if len(got.Points()) != len(sr.Points()) { + t.Errorf("Points count mismatch: got %v, exp %v", len(got.Points()), len(sr.Points())) + } + + srPoints := sr.Points() + gotPoints := got.Points() + for i, p := range srPoints { + g := gotPoints[i] + + if g.Name() != p.Name() { + t.Errorf("Point %d name mismatch: got %v, exp %v", i, g.Name(), p.Name()) + } + + if !g.Time().Equal(p.Time()) { + t.Errorf("Point %d time mismatch: got %v, exp %v", i, g.Time(), p.Time()) + } + + if g.HashID() != p.HashID() { + t.Errorf("Point #%d HashID() mismatch: got %v, exp %v", i, g.HashID(), p.HashID()) + } + + for k, v := range p.Tags() { + if g.Tags()[k] != v { + t.Errorf("Point #%d tag mismatch: got %v, exp %v", i, k, v) + } + } + + if len(p.Fields()) != len(g.Fields()) { + t.Errorf("Point %d field count mismatch: got %v, exp %v", i, len(g.Fields()), len(p.Fields())) + } + + for j, f := range p.Fields() { + if g.Fields()[j] != f { + t.Errorf("Point %d field mismatch: got %v, exp %v", i, g.Fields()[j], f) + } + } + } +} + +func TestWriteShardResponseBinary(t *testing.T) { + sr := &WriteShardResponse{} + sr.SetCode(10) + sr.SetMessage("foo") + b, err := sr.MarshalBinary() + + if exp := 10; sr.Code() != exp { + t.Fatalf("Code mismatch: got %v, exp %v", sr.Code(), exp) + } + + if exp := "foo"; sr.Message() != exp { + t.Fatalf("Message mismatch: got %v, exp %v", sr.Message(), exp) + } + + if err != nil { + t.Fatalf("WritePointsResponse.MarshalBinary() failed: %v", err) + } + if len(b) == 0 { + t.Fatalf("WritePointsResponse.MarshalBinary() returned 0 bytes") + } + + got := &WriteShardResponse{} + if err := got.UnmarshalBinary(b); err != nil { + t.Fatalf("WritePointsResponse.UnmarshalMarshalBinary() failed: %v", err) + } + + if got.Code() != sr.Code() { + t.Errorf("Code mismatch: got %v, exp %v", got.Code(), sr.Code()) + } + + if got.Message() != sr.Message() { + t.Errorf("Message mismatch: got %v, exp %v", got.Message(), sr.Message()) + } + +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/service.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/service.go new file mode 100644 index 000000000..896b0ce1e --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/service.go @@ -0,0 +1,338 @@ +package cluster + +import ( + "encoding/binary" + "encoding/json" + "fmt" + "io" + "log" + "net" + "os" + "strings" + "sync" + + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/tsdb" +) + +// MaxMessageSize defines how large a message can be before we reject it +const MaxMessageSize = 1024 * 1024 * 1024 // 1GB + +// MuxHeader is the header byte used in the TCP mux. +const MuxHeader = 2 + +// Service processes data received over raw TCP connections. +type Service struct { + mu sync.RWMutex + + wg sync.WaitGroup + closing chan struct{} + + Listener net.Listener + + MetaStore interface { + ShardOwner(shardID uint64) (string, string, *meta.ShardGroupInfo) + } + + TSDBStore interface { + CreateShard(database, policy string, shardID uint64) error + WriteToShard(shardID uint64, points []tsdb.Point) error + CreateMapper(shardID uint64, query string, chunkSize int) (tsdb.Mapper, error) + } + + Logger *log.Logger +} + +// NewService returns a new instance of Service. +func NewService(c Config) *Service { + return &Service{ + closing: make(chan struct{}), + Logger: log.New(os.Stderr, "[tcp] ", log.LstdFlags), + } +} + +// Open opens the network listener and begins serving requests. +func (s *Service) Open() error { + + s.Logger.Println("Starting cluster service") + // Begin serving conections. + s.wg.Add(1) + go s.serve() + + return nil +} + +// SetLogger sets the internal logger to the logger passed in. +func (s *Service) SetLogger(l *log.Logger) { + s.Logger = l +} + +// serve accepts connections from the listener and handles them. +func (s *Service) serve() { + defer s.wg.Done() + + for { + // Check if the service is shutting down. + select { + case <-s.closing: + return + default: + } + + // Accept the next connection. + conn, err := s.Listener.Accept() + if err != nil { + if strings.Contains(err.Error(), "connection closed") { + s.Logger.Printf("cluster service accept error: %s", err) + return + } + s.Logger.Printf("accept error: %s", err) + continue + } + + // Delegate connection handling to a separate goroutine. + s.wg.Add(1) + go func() { + defer s.wg.Done() + s.handleConn(conn) + }() + } +} + +// Close shuts down the listener and waits for all connections to finish. +func (s *Service) Close() error { + if s.Listener != nil { + s.Listener.Close() + } + + // Shut down all handlers. + close(s.closing) + // s.wg.Wait() // FIXME(benbjohnson) + + return nil +} + +// handleConn services an individual TCP connection. +func (s *Service) handleConn(conn net.Conn) { + // Ensure connection is closed when service is closed. + closing := make(chan struct{}) + defer close(closing) + go func() { + select { + case <-closing: + case <-s.closing: + } + conn.Close() + }() + + s.Logger.Printf("accept remote write connection from %v\n", conn.RemoteAddr()) + defer func() { + s.Logger.Printf("close remote write connection from %v\n", conn.RemoteAddr()) + }() + for { + // Read type-length-value. + typ, buf, err := ReadTLV(conn) + if err != nil { + if strings.HasSuffix(err.Error(), "EOF") { + return + } + s.Logger.Printf("unable to read type-length-value %s", err) + return + } + + // Delegate message processing by type. + switch typ { + case writeShardRequestMessage: + err := s.processWriteShardRequest(buf) + if err != nil { + s.Logger.Printf("process write shard error: %s", err) + } + s.writeShardResponse(conn, err) + case mapShardRequestMessage: + err := s.processMapShardRequest(conn, buf) + if err != nil { + s.Logger.Printf("process map shard error: %s", err) + if err := writeMapShardResponseMessage(conn, NewMapShardResponse(1, err.Error())); err != nil { + s.Logger.Printf("process map shard error writing response: %s", err.Error()) + } + } + default: + s.Logger.Printf("cluster service message type not found: %d", typ) + } + } +} + +func (s *Service) processWriteShardRequest(buf []byte) error { + // Build request + var req WriteShardRequest + if err := req.UnmarshalBinary(buf); err != nil { + return err + } + + err := s.TSDBStore.WriteToShard(req.ShardID(), req.Points()) + + // We may have received a write for a shard that we don't have locally because the + // sending node may have just created the shard (via the metastore) and the write + // arrived before the local store could create the shard. In this case, we need + // to check the metastore to determine what database and retention policy this + // shard should reside within. + if err == tsdb.ErrShardNotFound { + + // Query the metastore for the owner of this shard + database, retentionPolicy, sgi := s.MetaStore.ShardOwner(req.ShardID()) + if sgi == nil { + // If we can't find it, then we need to drop this request + // as it is no longer valid. This could happen if writes were queued via + // hinted handoff and delivered after a shard group was deleted. + s.Logger.Printf("drop write request: shard=%d", req.ShardID()) + return nil + } + + err = s.TSDBStore.CreateShard(database, retentionPolicy, req.ShardID()) + if err != nil { + return err + } + return s.TSDBStore.WriteToShard(req.ShardID(), req.Points()) + } + + if err != nil { + return fmt.Errorf("write shard %d: %s", req.ShardID(), err) + } + + return nil +} + +func (s *Service) writeShardResponse(w io.Writer, e error) { + // Build response. + var resp WriteShardResponse + if e != nil { + resp.SetCode(1) + resp.SetMessage(e.Error()) + } else { + resp.SetCode(0) + } + + // Marshal response to binary. + buf, err := resp.MarshalBinary() + if err != nil { + s.Logger.Printf("error marshalling shard response: %s", err) + return + } + + // Write to connection. + if err := WriteTLV(w, writeShardResponseMessage, buf); err != nil { + s.Logger.Printf("write shard response error: %s", err) + } +} + +func (s *Service) processMapShardRequest(w io.Writer, buf []byte) error { + // Decode request + var req MapShardRequest + if err := req.UnmarshalBinary(buf); err != nil { + return err + } + + m, err := s.TSDBStore.CreateMapper(req.ShardID(), req.Query(), int(req.ChunkSize())) + if err != nil { + return fmt.Errorf("create mapper: %s", err) + } + if m == nil { + return writeMapShardResponseMessage(w, NewMapShardResponse(0, "")) + } + + if err := m.Open(); err != nil { + return fmt.Errorf("mapper open: %s", err) + } + defer m.Close() + + var metaSent bool + for { + var resp MapShardResponse + + if !metaSent { + resp.SetTagSets(m.TagSets()) + resp.SetFields(m.Fields()) + metaSent = true + } + + chunk, err := m.NextChunk() + if err != nil { + return fmt.Errorf("next chunk: %s", err) + } + if chunk != nil { + b, err := json.Marshal(chunk) + if err != nil { + return fmt.Errorf("encoding: %s", err) + } + resp.SetData(b) + } + + // Write to connection. + resp.SetCode(0) + if err := writeMapShardResponseMessage(w, &resp); err != nil { + return err + } + + if chunk == nil { + // All mapper data sent. + return nil + } + } +} + +func writeMapShardResponseMessage(w io.Writer, msg *MapShardResponse) error { + buf, err := msg.MarshalBinary() + if err != nil { + return err + } + return WriteTLV(w, mapShardResponseMessage, buf) +} + +// ReadTLV reads a type-length-value record from r. +func ReadTLV(r io.Reader) (byte, []byte, error) { + var typ [1]byte + if _, err := io.ReadFull(r, typ[:]); err != nil { + return 0, nil, fmt.Errorf("read message type: %s", err) + } + + // Read the size of the message. + var sz int64 + if err := binary.Read(r, binary.BigEndian, &sz); err != nil { + return 0, nil, fmt.Errorf("read message size: %s", err) + } + + if sz == 0 { + return 0, nil, fmt.Errorf("invalid message size: %d", sz) + } + + if sz >= MaxMessageSize { + return 0, nil, fmt.Errorf("max message size of %d exceeded: %d", MaxMessageSize, sz) + } + + // Read the value. + buf := make([]byte, sz) + if _, err := io.ReadFull(r, buf); err != nil { + return 0, nil, fmt.Errorf("read message value: %s", err) + } + + return typ[0], buf, nil +} + +// WriteTLV writes a type-length-value record to w. +func WriteTLV(w io.Writer, typ byte, buf []byte) error { + if _, err := w.Write([]byte{typ}); err != nil { + return fmt.Errorf("write message type: %s", err) + } + + // Write the size of the message. + if err := binary.Write(w, binary.BigEndian, int64(len(buf))); err != nil { + return fmt.Errorf("write message size: %s", err) + } + + // Write the value. + if _, err := w.Write(buf); err != nil { + return fmt.Errorf("write message value: %s", err) + } + + return nil +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/service_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/service_test.go new file mode 100644 index 000000000..114135db8 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/service_test.go @@ -0,0 +1,103 @@ +package cluster_test + +import ( + "fmt" + "net" + "time" + + "github.com/influxdb/influxdb/cluster" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/tcp" + "github.com/influxdb/influxdb/tsdb" +) + +type metaStore struct { + host string +} + +func (m *metaStore) Node(nodeID uint64) (*meta.NodeInfo, error) { + return &meta.NodeInfo{ + ID: nodeID, + Host: m.host, + }, nil +} + +type testService struct { + nodeID uint64 + ln net.Listener + muxln net.Listener + writeShardFunc func(shardID uint64, points []tsdb.Point) error + createShardFunc func(database, policy string, shardID uint64) error + createMapperFunc func(shardID uint64, query string, chunkSize int) (tsdb.Mapper, error) +} + +func newTestWriteService(f func(shardID uint64, points []tsdb.Point) error) testService { + ln, err := net.Listen("tcp", "127.0.0.1:0") + if err != nil { + panic(err) + } + + mux := tcp.NewMux() + muxln := mux.Listen(cluster.MuxHeader) + go mux.Serve(ln) + + return testService{ + writeShardFunc: f, + ln: ln, + muxln: muxln, + } +} + +func (ts *testService) Close() { + if ts.ln != nil { + ts.ln.Close() + } +} + +type serviceResponses []serviceResponse +type serviceResponse struct { + shardID uint64 + ownerID uint64 + points []tsdb.Point +} + +func (t testService) WriteToShard(shardID uint64, points []tsdb.Point) error { + return t.writeShardFunc(shardID, points) +} + +func (t testService) CreateShard(database, policy string, shardID uint64) error { + return t.createShardFunc(database, policy, shardID) +} + +func (t testService) CreateMapper(shardID uint64, query string, chunkSize int) (tsdb.Mapper, error) { + return t.createMapperFunc(shardID, query, chunkSize) +} + +func writeShardSuccess(shardID uint64, points []tsdb.Point) error { + responses <- &serviceResponse{ + shardID: shardID, + points: points, + } + return nil +} + +func writeShardFail(shardID uint64, points []tsdb.Point) error { + return fmt.Errorf("failed to write") +} + +var responses = make(chan *serviceResponse, 1024) + +func (testService) ResponseN(n int) ([]*serviceResponse, error) { + var a []*serviceResponse + for { + select { + case r := <-responses: + a = append(a, r) + if len(a) == n { + return a, nil + } + case <-time.After(time.Second): + return a, fmt.Errorf("unexpected response count: expected: %d, actual: %d", n, len(a)) + } + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_mapper.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_mapper.go new file mode 100644 index 000000000..8756002d2 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_mapper.go @@ -0,0 +1,207 @@ +package cluster + +import ( + "encoding/json" + "fmt" + "io" + "math/rand" + "net" + "time" + + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/tsdb" + "gopkg.in/fatih/pool.v2" +) + +// ShardMapper is responsible for providing mappers for requested shards. It is +// responsible for creating those mappers from the local store, or reaching +// out to another node on the cluster. +type ShardMapper struct { + ForceRemoteMapping bool // All shards treated as remote. Useful for testing. + + MetaStore interface { + NodeID() uint64 + Node(id uint64) (ni *meta.NodeInfo, err error) + } + + TSDBStore interface { + CreateMapper(shardID uint64, query string, chunkSize int) (tsdb.Mapper, error) + } + + timeout time.Duration + pool *clientPool +} + +// NewShardMapper returns a mapper of local and remote shards. +func NewShardMapper(timeout time.Duration) *ShardMapper { + return &ShardMapper{ + pool: newClientPool(), + timeout: timeout, + } +} + +// CreateMapper returns a Mapper for the given shard ID. +func (s *ShardMapper) CreateMapper(sh meta.ShardInfo, stmt string, chunkSize int) (tsdb.Mapper, error) { + var err error + var m tsdb.Mapper + if sh.OwnedBy(s.MetaStore.NodeID()) && !s.ForceRemoteMapping { + m, err = s.TSDBStore.CreateMapper(sh.ID, stmt, chunkSize) + if err != nil { + return nil, err + } + } else { + // Pick a node in a pseudo-random manner. + conn, err := s.dial(sh.OwnerIDs[rand.Intn(len(sh.OwnerIDs))]) + if err != nil { + return nil, err + } + conn.SetDeadline(time.Now().Add(s.timeout)) + + rm := NewRemoteMapper(conn.(*pool.PoolConn), sh.ID, stmt, chunkSize) + m = rm + } + + return m, nil +} + +func (s *ShardMapper) dial(nodeID uint64) (net.Conn, error) { + // If we don't have a connection pool for that addr yet, create one + _, ok := s.pool.getPool(nodeID) + if !ok { + factory := &connFactory{nodeID: nodeID, clientPool: s.pool, timeout: s.timeout} + factory.metaStore = s.MetaStore + + p, err := pool.NewChannelPool(1, 3, factory.dial) + if err != nil { + return nil, err + } + s.pool.setPool(nodeID, p) + } + return s.pool.conn(nodeID) +} + +type remoteShardConn interface { + io.ReadWriter + Close() error + MarkUnusable() +} + +// RemoteMapper implements the tsdb.Mapper interface. It connects to a remote node, +// sends a query, and interprets the stream of data that comes back. +type RemoteMapper struct { + shardID uint64 + stmt string + chunkSize int + + tagsets []string + fields []string + + conn remoteShardConn + bufferedResponse *MapShardResponse +} + +// NewRemoteMapper returns a new remote mapper using the given connection. +func NewRemoteMapper(c remoteShardConn, shardID uint64, stmt string, chunkSize int) *RemoteMapper { + return &RemoteMapper{ + conn: c, + shardID: shardID, + stmt: stmt, + chunkSize: chunkSize, + } +} + +// Open connects to the remote node and starts receiving data. +func (r *RemoteMapper) Open() (err error) { + defer func() { + if err != nil { + r.conn.Close() + } + }() + // Build Map request. + var request MapShardRequest + request.SetShardID(r.shardID) + request.SetQuery(r.stmt) + request.SetChunkSize(int32(r.chunkSize)) + + // Marshal into protocol buffers. + buf, err := request.MarshalBinary() + if err != nil { + return err + } + + // Write request. + if err := WriteTLV(r.conn, mapShardRequestMessage, buf); err != nil { + r.conn.MarkUnusable() + return err + } + + // Read the response. + _, buf, err = ReadTLV(r.conn) + if err != nil { + r.conn.MarkUnusable() + return err + } + + // Unmarshal response. + r.bufferedResponse = &MapShardResponse{} + if err := r.bufferedResponse.UnmarshalBinary(buf); err != nil { + return err + } + + if r.bufferedResponse.Code() != 0 { + return fmt.Errorf("error code %d: %s", r.bufferedResponse.Code(), r.bufferedResponse.Message()) + } + + // Decode the first response to get the TagSets. + r.tagsets = r.bufferedResponse.TagSets() + + return nil +} + +func (r *RemoteMapper) TagSets() []string { + return r.tagsets +} + +func (r *RemoteMapper) Fields() []string { + return r.fields +} + +// NextChunk returns the next chunk read from the remote node to the client. +func (r *RemoteMapper) NextChunk() (chunk interface{}, err error) { + output := &tsdb.MapperOutput{} + var response *MapShardResponse + + if r.bufferedResponse != nil { + response = r.bufferedResponse + r.bufferedResponse = nil + } else { + response = &MapShardResponse{} + + // Read the response. + _, buf, err := ReadTLV(r.conn) + if err != nil { + r.conn.MarkUnusable() + return nil, err + } + + // Unmarshal response. + if err := response.UnmarshalBinary(buf); err != nil { + return nil, err + } + + if response.Code() != 0 { + return nil, fmt.Errorf("error code %d: %s", response.Code(), response.Message()) + } + } + + if response.Data() == nil { + return nil, nil + } + err = json.Unmarshal(response.Data(), output) + return output, err +} + +// Close the Mapper +func (r *RemoteMapper) Close() { + r.conn.Close() +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_mapper_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_mapper_test.go new file mode 100644 index 000000000..afb35ef15 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_mapper_test.go @@ -0,0 +1,96 @@ +package cluster + +import ( + "bytes" + "encoding/json" + "io" + "testing" + + "github.com/influxdb/influxdb/tsdb" +) + +// remoteShardResponder implements the remoteShardConn interface. +type remoteShardResponder struct { + t *testing.T + rxBytes []byte + + buffer *bytes.Buffer +} + +func newRemoteShardResponder(outputs []*tsdb.MapperOutput, tagsets []string) *remoteShardResponder { + r := &remoteShardResponder{} + a := make([]byte, 0, 1024) + r.buffer = bytes.NewBuffer(a) + + // Pump the outputs in the buffer for later reading. + for _, o := range outputs { + resp := &MapShardResponse{} + resp.SetCode(0) + if o != nil { + d, _ := json.Marshal(o) + resp.SetData(d) + resp.SetTagSets(tagsets) + } + + g, _ := resp.MarshalBinary() + WriteTLV(r.buffer, mapShardResponseMessage, g) + } + + return r +} + +func (r remoteShardResponder) MarkUnusable() { return } +func (r remoteShardResponder) Close() error { return nil } +func (r remoteShardResponder) Read(p []byte) (n int, err error) { + return io.ReadFull(r.buffer, p) +} + +func (r remoteShardResponder) Write(p []byte) (n int, err error) { + if r.rxBytes == nil { + r.rxBytes = make([]byte, 0) + } + r.rxBytes = append(r.rxBytes, p...) + return len(p), nil +} + +// Ensure a RemoteMapper can process valid responses from a remote shard. +func TestShardWriter_RemoteMapper_Success(t *testing.T) { + expTagSets := []string{"tagsetA"} + expOutput := &tsdb.MapperOutput{ + Name: "cpu", + Tags: map[string]string{"host": "serverA"}, + } + + c := newRemoteShardResponder([]*tsdb.MapperOutput{expOutput, nil}, expTagSets) + + r := NewRemoteMapper(c, 1234, "SELECT * FROM CPU", 10) + if err := r.Open(); err != nil { + t.Fatalf("failed to open remote mapper: %s", err.Error()) + } + + if r.TagSets()[0] != expTagSets[0] { + t.Fatalf("incorrect tagsets received, exp %v, got %v", expTagSets, r.TagSets()) + } + + // Get first chunk from mapper. + chunk, err := r.NextChunk() + if err != nil { + t.Fatalf("failed to get next chunk from mapper: %s", err.Error()) + } + output, ok := chunk.(*tsdb.MapperOutput) + if !ok { + t.Fatal("chunk is not of expected type") + } + if output.Name != "cpu" { + t.Fatalf("received output incorrect, exp: %v, got %v", expOutput, output) + } + + // Next chunk should be nil, indicating no more data. + chunk, err = r.NextChunk() + if err != nil { + t.Fatalf("failed to get next chunk from mapper: %s", err.Error()) + } + if chunk != nil { + t.Fatal("received more chunks when none expected") + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_writer.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_writer.go new file mode 100644 index 000000000..a0f317be9 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_writer.go @@ -0,0 +1,163 @@ +package cluster + +import ( + "fmt" + "net" + "time" + + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/tsdb" + "gopkg.in/fatih/pool.v2" +) + +const ( + writeShardRequestMessage byte = iota + 1 + writeShardResponseMessage + mapShardRequestMessage + mapShardResponseMessage +) + +// ShardWriter writes a set of points to a shard. +type ShardWriter struct { + pool *clientPool + timeout time.Duration + + MetaStore interface { + Node(id uint64) (ni *meta.NodeInfo, err error) + } +} + +// NewShardWriter returns a new instance of ShardWriter. +func NewShardWriter(timeout time.Duration) *ShardWriter { + return &ShardWriter{ + pool: newClientPool(), + timeout: timeout, + } +} + +func (w *ShardWriter) WriteShard(shardID, ownerID uint64, points []tsdb.Point) error { + c, err := w.dial(ownerID) + if err != nil { + return err + } + + conn, ok := c.(*pool.PoolConn) + if !ok { + panic("wrong connection type") + } + defer func(conn net.Conn) { + conn.Close() // return to pool + }(conn) + + // Build write request. + var request WriteShardRequest + request.SetShardID(shardID) + request.AddPoints(points) + + // Marshal into protocol buffers. + buf, err := request.MarshalBinary() + if err != nil { + return err + } + + // Write request. + conn.SetWriteDeadline(time.Now().Add(w.timeout)) + if err := WriteTLV(conn, writeShardRequestMessage, buf); err != nil { + conn.MarkUnusable() + return err + } + + // Read the response. + conn.SetReadDeadline(time.Now().Add(w.timeout)) + _, buf, err = ReadTLV(conn) + if err != nil { + conn.MarkUnusable() + return err + } + + // Unmarshal response. + var response WriteShardResponse + if err := response.UnmarshalBinary(buf); err != nil { + return err + } + + if response.Code() != 0 { + return fmt.Errorf("error code %d: %s", response.Code(), response.Message()) + } + + return nil +} + +func (c *ShardWriter) dial(nodeID uint64) (net.Conn, error) { + // If we don't have a connection pool for that addr yet, create one + _, ok := c.pool.getPool(nodeID) + if !ok { + factory := &connFactory{nodeID: nodeID, clientPool: c.pool, timeout: c.timeout} + factory.metaStore = c.MetaStore + + p, err := pool.NewChannelPool(1, 3, factory.dial) + if err != nil { + return nil, err + } + c.pool.setPool(nodeID, p) + } + return c.pool.conn(nodeID) +} + +func (w *ShardWriter) Close() error { + if w.pool == nil { + return fmt.Errorf("client already closed") + } + w.pool.close() + w.pool = nil + return nil +} + +const ( + maxConnections = 500 + maxRetries = 3 +) + +var errMaxConnectionsExceeded = fmt.Errorf("can not exceed max connections of %d", maxConnections) + +type connFactory struct { + nodeID uint64 + timeout time.Duration + + clientPool interface { + size() int + } + + metaStore interface { + Node(id uint64) (ni *meta.NodeInfo, err error) + } +} + +func (c *connFactory) dial() (net.Conn, error) { + if c.clientPool.size() > maxConnections { + return nil, errMaxConnectionsExceeded + } + + ni, err := c.metaStore.Node(c.nodeID) + if err != nil { + return nil, err + } + + if ni == nil { + return nil, fmt.Errorf("node %d does not exist", c.nodeID) + } + + conn, err := net.DialTimeout("tcp", ni.Host, c.timeout) + if err != nil { + return nil, err + } + + // Write a marker byte for cluster messages. + _, err = conn.Write([]byte{MuxHeader}) + if err != nil { + conn.Close() + return nil, err + } + + return conn, nil +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_writer_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_writer_test.go new file mode 100644 index 000000000..d994315ca --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cluster/shard_writer_test.go @@ -0,0 +1,186 @@ +package cluster_test + +import ( + "net" + "strings" + "testing" + "time" + + "github.com/influxdb/influxdb/cluster" + "github.com/influxdb/influxdb/tsdb" +) + +// Ensure the shard writer can successful write a single request. +func TestShardWriter_WriteShard_Success(t *testing.T) { + ts := newTestWriteService(writeShardSuccess) + s := cluster.NewService(cluster.Config{}) + s.Listener = ts.muxln + s.TSDBStore = ts + if err := s.Open(); err != nil { + t.Fatal(err) + } + defer s.Close() + defer ts.Close() + + w := cluster.NewShardWriter(time.Minute) + w.MetaStore = &metaStore{host: ts.ln.Addr().String()} + + // Build a single point. + now := time.Now() + var points []tsdb.Point + points = append(points, tsdb.NewPoint("cpu", tsdb.Tags{"host": "server01"}, map[string]interface{}{"value": int64(100)}, now)) + + // Write to shard and close. + if err := w.WriteShard(1, 2, points); err != nil { + t.Fatal(err) + } else if err := w.Close(); err != nil { + t.Fatal(err) + } + + // Validate response. + responses, err := ts.ResponseN(1) + if err != nil { + t.Fatal(err) + } else if responses[0].shardID != 1 { + t.Fatalf("unexpected shard id: %d", responses[0].shardID) + } + + // Validate point. + if p := responses[0].points[0]; p.Name() != "cpu" { + t.Fatalf("unexpected name: %s", p.Name()) + } else if p.Fields()["value"] != int64(100) { + t.Fatalf("unexpected 'value' field: %d", p.Fields()["value"]) + } else if p.Tags()["host"] != "server01" { + t.Fatalf("unexpected 'host' tag: %s", p.Tags()["host"]) + } else if p.Time().UnixNano() != now.UnixNano() { + t.Fatalf("unexpected time: %s", p.Time()) + } +} + +// Ensure the shard writer can successful write a multiple requests. +func TestShardWriter_WriteShard_Multiple(t *testing.T) { + ts := newTestWriteService(writeShardSuccess) + s := cluster.NewService(cluster.Config{}) + s.Listener = ts.muxln + s.TSDBStore = ts + if err := s.Open(); err != nil { + t.Fatal(err) + } + defer s.Close() + defer ts.Close() + + w := cluster.NewShardWriter(time.Minute) + w.MetaStore = &metaStore{host: ts.ln.Addr().String()} + + // Build a single point. + now := time.Now() + var points []tsdb.Point + points = append(points, tsdb.NewPoint("cpu", tsdb.Tags{"host": "server01"}, map[string]interface{}{"value": int64(100)}, now)) + + // Write to shard twice and close. + if err := w.WriteShard(1, 2, points); err != nil { + t.Fatal(err) + } else if err := w.WriteShard(1, 2, points); err != nil { + t.Fatal(err) + } else if err := w.Close(); err != nil { + t.Fatal(err) + } + + // Validate response. + responses, err := ts.ResponseN(1) + if err != nil { + t.Fatal(err) + } else if responses[0].shardID != 1 { + t.Fatalf("unexpected shard id: %d", responses[0].shardID) + } + + // Validate point. + if p := responses[0].points[0]; p.Name() != "cpu" { + t.Fatalf("unexpected name: %s", p.Name()) + } else if p.Fields()["value"] != int64(100) { + t.Fatalf("unexpected 'value' field: %d", p.Fields()["value"]) + } else if p.Tags()["host"] != "server01" { + t.Fatalf("unexpected 'host' tag: %s", p.Tags()["host"]) + } else if p.Time().UnixNano() != now.UnixNano() { + t.Fatalf("unexpected time: %s", p.Time()) + } +} + +// Ensure the shard writer returns an error when the server fails to accept the write. +func TestShardWriter_WriteShard_Error(t *testing.T) { + ts := newTestWriteService(writeShardFail) + s := cluster.NewService(cluster.Config{}) + s.Listener = ts.muxln + s.TSDBStore = ts + if err := s.Open(); err != nil { + t.Fatal(err) + } + defer s.Close() + defer ts.Close() + + w := cluster.NewShardWriter(time.Minute) + w.MetaStore = &metaStore{host: ts.ln.Addr().String()} + now := time.Now() + + shardID := uint64(1) + ownerID := uint64(2) + var points []tsdb.Point + points = append(points, tsdb.NewPoint( + "cpu", tsdb.Tags{"host": "server01"}, map[string]interface{}{"value": int64(100)}, now, + )) + + if err := w.WriteShard(shardID, ownerID, points); err == nil || err.Error() != "error code 1: write shard 1: failed to write" { + t.Fatalf("unexpected error: %v", err) + } +} + +// Ensure the shard writer returns an error when dialing times out. +func TestShardWriter_Write_ErrDialTimeout(t *testing.T) { + ts := newTestWriteService(writeShardSuccess) + s := cluster.NewService(cluster.Config{}) + s.Listener = ts.muxln + s.TSDBStore = ts + if err := s.Open(); err != nil { + t.Fatal(err) + } + defer s.Close() + defer ts.Close() + + w := cluster.NewShardWriter(time.Nanosecond) + w.MetaStore = &metaStore{host: ts.ln.Addr().String()} + now := time.Now() + + shardID := uint64(1) + ownerID := uint64(2) + var points []tsdb.Point + points = append(points, tsdb.NewPoint( + "cpu", tsdb.Tags{"host": "server01"}, map[string]interface{}{"value": int64(100)}, now, + )) + + if err, exp := w.WriteShard(shardID, ownerID, points), "i/o timeout"; err == nil || !strings.Contains(err.Error(), exp) { + t.Fatalf("expected error %v, to contain %s", err, exp) + } +} + +// Ensure the shard writer returns an error when reading times out. +func TestShardWriter_Write_ErrReadTimeout(t *testing.T) { + ln, err := net.Listen("tcp", "127.0.0.1:0") + if err != nil { + t.Fatal(err) + } + + w := cluster.NewShardWriter(time.Millisecond) + w.MetaStore = &metaStore{host: ln.Addr().String()} + now := time.Now() + + shardID := uint64(1) + ownerID := uint64(2) + var points []tsdb.Point + points = append(points, tsdb.NewPoint( + "cpu", tsdb.Tags{"host": "server01"}, map[string]interface{}{"value": int64(100)}, now, + )) + + if err := w.WriteShard(shardID, ownerID, points); err == nil || !strings.Contains(err.Error(), "i/o timeout") { + t.Fatalf("unexpected error: %s", err) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx/main.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx/main.go new file mode 100644 index 000000000..24af8efa4 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx/main.go @@ -0,0 +1,724 @@ +package main + +import ( + "encoding/csv" + "encoding/json" + "flag" + "fmt" + "io" + "net" + "net/url" + "os" + "os/user" + "path/filepath" + "sort" + "strconv" + "strings" + "text/tabwriter" + + "github.com/influxdb/influxdb/client" + "github.com/influxdb/influxdb/importer/v8" + "github.com/peterh/liner" +) + +// These variables are populated via the Go linker. +var ( + version string = "0.9" +) + +const ( + // defaultFormat is the default format of the results when issuing queries + defaultFormat = "column" + + // defaultPPS is the default points per second that the import will throttle at + // by default it's 0, which means it will not throttle + defaultPPS = 0 +) + +type CommandLine struct { + Client *client.Client + Line *liner.State + Host string + Port int + Username string + Password string + Database string + Ssl bool + RetentionPolicy string + Version string + Pretty bool // controls pretty print for json + Format string // controls the output format. Valid values are json, csv, or column + Execute string + ShowVersion bool + Import bool + PPS int // Controls how many points per second the import will allow via throttling + Path string + Compressed bool +} + +func main() { + c := CommandLine{} + + fs := flag.NewFlagSet("InfluxDB shell version "+version, flag.ExitOnError) + fs.StringVar(&c.Host, "host", client.DefaultHost, "Influxdb host to connect to.") + fs.IntVar(&c.Port, "port", client.DefaultPort, "Influxdb port to connect to.") + fs.StringVar(&c.Username, "username", c.Username, "Username to connect to the server.") + fs.StringVar(&c.Password, "password", c.Password, `Password to connect to the server. Leaving blank will prompt for password (--password="").`) + fs.StringVar(&c.Database, "database", c.Database, "Database to connect to the server.") + fs.BoolVar(&c.Ssl, "ssl", false, "Use https for connecting to cluster.") + fs.StringVar(&c.Format, "format", defaultFormat, "Format specifies the format of the server responses: json, csv, or column.") + fs.BoolVar(&c.Pretty, "pretty", false, "Turns on pretty print for the json format.") + fs.StringVar(&c.Execute, "execute", c.Execute, "Execute command and quit.") + fs.BoolVar(&c.ShowVersion, "version", false, "Displays the InfluxDB version.") + fs.BoolVar(&c.Import, "import", false, "Import a previous database.") + fs.IntVar(&c.PPS, "pps", defaultPPS, "How many points per second the import will allow. By default it is zero and will not throttle importing.") + fs.StringVar(&c.Path, "path", "", "path to the file to import") + fs.BoolVar(&c.Compressed, "compressed", false, "set to true if the import file is compressed") + + // Define our own custom usage to print + fs.Usage = func() { + fmt.Println(`Usage of influx: + -version + Display the version and exit. + -host 'host name' + Host to connect to. + -port 'port #' + Port to connect to. + -database 'database name' + Database to connect to the server. + -password 'password' + Password to connect to the server. Leaving blank will prompt for password (--password ''). + -username 'username' + Username to connect to the server. + -ssl + Use https for requests. + -execute 'command' + Execute command and quit. + -format 'json|csv|column' + Format specifies the format of the server responses: json, csv, or column. + -pretty + Turns on pretty print for the json format. + -import + Import a previous database export from file + -pps + How many points per second the import will allow. By default it is zero and will not throttle importing. + -path + Path to file to import + -compressed + Set to true if the import file is compressed + +Examples: + + # Use influx in a non-interactive mode to query the database "metrics" and pretty print json: + $ influx -database 'metrics' -execute 'select * from cpu' -format 'json' -pretty + + # Connect to a specific database on startup and set database context: + $ influx -database 'metrics' -host 'localhost' -port '8086' +`) + } + fs.Parse(os.Args[1:]) + + if c.ShowVersion { + showVersion() + os.Exit(0) + } + + var promptForPassword bool + // determine if they set the password flag but provided no value + for _, v := range os.Args { + v = strings.ToLower(v) + if (strings.HasPrefix(v, "-password") || strings.HasPrefix(v, "--password")) && c.Password == "" { + promptForPassword = true + break + } + } + + c.Line = liner.NewLiner() + defer c.Line.Close() + + if promptForPassword { + p, e := c.Line.PasswordPrompt("password: ") + if e != nil { + fmt.Println("Unable to parse password.") + } else { + c.Password = p + } + } + + if err := c.connect(""); err != nil { + + } + if c.Execute == "" && !c.Import { + fmt.Printf("Connected to %s version %s\n", c.Client.Addr(), c.Version) + } + + if c.Execute != "" { + if err := c.ExecuteQuery(c.Execute); err != nil { + c.Line.Close() + os.Exit(1) + } + c.Line.Close() + os.Exit(0) + } + + if c.Import { + path := net.JoinHostPort(c.Host, strconv.Itoa(c.Port)) + u, e := client.ParseConnectionString(path, c.Ssl) + if e != nil { + fmt.Println(e) + return + } + + config := v8.NewConfig() + config.Username = c.Username + config.Password = c.Password + config.Precision = "ns" + config.WriteConsistency = "any" + config.Path = c.Path + config.Version = version + config.URL = u + config.Compressed = c.Compressed + config.PPS = c.PPS + + i := v8.NewImporter(config) + if err := i.Import(); err != nil { + fmt.Printf("ERROR: %s\n", err) + c.Line.Close() + os.Exit(1) + } + c.Line.Close() + os.Exit(0) + } + + showVersion() + + var historyFile string + usr, err := user.Current() + // Only load history if we can get the user + if err == nil { + historyFile = filepath.Join(usr.HomeDir, ".influx_history") + + if f, err := os.Open(historyFile); err == nil { + c.Line.ReadHistory(f) + f.Close() + } + } + + for { + l, e := c.Line.Prompt("> ") + if e != nil { + break + } + if c.ParseCommand(l) { + // write out the history + if len(historyFile) > 0 { + c.Line.AppendHistory(l) + if f, err := os.Create(historyFile); err == nil { + c.Line.WriteHistory(f) + f.Close() + } + } + } else { + break // exit main loop + } + } +} + +func showVersion() { + fmt.Println("InfluxDB shell " + version) +} + +func (c *CommandLine) ParseCommand(cmd string) bool { + lcmd := strings.TrimSpace(strings.ToLower(cmd)) + switch { + case strings.HasPrefix(lcmd, "exit"): + // signal the program to exit + return false + case strings.HasPrefix(lcmd, "gopher"): + c.gopher() + case strings.HasPrefix(lcmd, "connect"): + c.connect(cmd) + case strings.HasPrefix(lcmd, "auth"): + c.SetAuth(cmd) + case strings.HasPrefix(lcmd, "help"): + c.help() + case strings.HasPrefix(lcmd, "format"): + c.SetFormat(cmd) + case strings.HasPrefix(lcmd, "settings"): + c.Settings() + case strings.HasPrefix(lcmd, "pretty"): + c.Pretty = !c.Pretty + if c.Pretty { + fmt.Println("Pretty print enabled") + } else { + fmt.Println("Pretty print disabled") + } + case strings.HasPrefix(lcmd, "use"): + c.use(cmd) + case strings.HasPrefix(lcmd, "insert"): + c.Insert(cmd) + case lcmd == "": + break + default: + c.ExecuteQuery(cmd) + } + return true +} + +func (c *CommandLine) connect(cmd string) error { + var cl *client.Client + var u url.URL + + // Remove the "connect" keyword if it exists + path := strings.TrimSpace(strings.Replace(cmd, "connect", "", -1)) + + // If they didn't provide a connection string, use the current settings + if path == "" { + path = net.JoinHostPort(c.Host, strconv.Itoa(c.Port)) + } + + var e error + u, e = client.ParseConnectionString(path, c.Ssl) + if e != nil { + return e + } + + config := client.NewConfig() + config.URL = u + config.Username = c.Username + config.Password = c.Password + config.UserAgent = "InfluxDBShell/" + version + cl, err := client.NewClient(config) + if err != nil { + return fmt.Errorf("Could not create client %s", err) + } + c.Client = cl + if _, v, e := c.Client.Ping(); e != nil { + return fmt.Errorf("Failed to connect to %s\n", c.Client.Addr()) + } else { + c.Version = v + } + return nil +} + +func (c *CommandLine) SetAuth(cmd string) { + // If they pass in the entire command, we should parse it + // auth + args := strings.Fields(cmd) + if len(args) == 3 { + args = args[1:] + } else { + args = []string{} + } + + if len(args) == 2 { + c.Username = args[0] + c.Password = args[1] + } else { + u, e := c.Line.Prompt("username: ") + if e != nil { + fmt.Printf("Unable to process input: %s", e) + return + } + c.Username = strings.TrimSpace(u) + p, e := c.Line.PasswordPrompt("password: ") + if e != nil { + fmt.Printf("Unable to process input: %s", e) + return + } + c.Password = p + } + + // Update the client as well + c.Client.SetAuth(c.Username, c.Password) +} + +func (c *CommandLine) use(cmd string) { + args := strings.Split(strings.TrimSuffix(strings.TrimSpace(cmd), ";"), " ") + if len(args) != 2 { + fmt.Printf("Could not parse database name from %q.\n", cmd) + return + } + d := args[1] + c.Database = d + fmt.Printf("Using database %s\n", d) +} + +func (c *CommandLine) SetFormat(cmd string) { + // Remove the "format" keyword if it exists + cmd = strings.TrimSpace(strings.Replace(cmd, "format", "", -1)) + // normalize cmd + cmd = strings.ToLower(cmd) + + switch cmd { + case "json", "csv", "column": + c.Format = cmd + default: + fmt.Printf("Unknown format %q. Please use json, csv, or column.\n", cmd) + } +} + +// isWhitespace returns true if the rune is a space, tab, or newline. +func isWhitespace(ch rune) bool { return ch == ' ' || ch == '\t' || ch == '\n' } + +// isLetter returns true if the rune is a letter. +func isLetter(ch rune) bool { return (ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z') } + +// isDigit returns true if the rune is a digit. +func isDigit(ch rune) bool { return (ch >= '0' && ch <= '9') } + +// isIdentFirstChar returns true if the rune can be used as the first char in an unquoted identifer. +func isIdentFirstChar(ch rune) bool { return isLetter(ch) || ch == '_' } + +// isIdentChar returns true if the rune can be used in an unquoted identifier. +func isNotIdentChar(ch rune) bool { return !(isLetter(ch) || isDigit(ch) || ch == '_') } + +func parseUnquotedIdentifier(stmt string) (string, string) { + if fields := strings.FieldsFunc(stmt, isNotIdentChar); len(fields) > 0 { + return fields[0], strings.TrimPrefix(stmt, fields[0]) + } + return "", stmt +} + +func parseDoubleQuotedIdentifier(stmt string) (string, string) { + escapeNext := false + fields := strings.FieldsFunc(stmt, func(ch rune) bool { + if ch == '\\' { + escapeNext = true + } else if ch == '"' { + if !escapeNext { + return true + } + escapeNext = false + } + return false + }) + if len(fields) > 0 { + return fields[0], strings.TrimPrefix(stmt, "\""+fields[0]+"\"") + } + return "", stmt +} + +func parseNextIdentifier(stmt string) (ident, remainder string) { + if len(stmt) > 0 { + switch { + case isWhitespace(rune(stmt[0])): + return parseNextIdentifier(stmt[1:]) + case isIdentFirstChar(rune(stmt[0])): + return parseUnquotedIdentifier(stmt) + case stmt[0] == '"': + return parseDoubleQuotedIdentifier(stmt) + } + } + return "", stmt +} + +func (c *CommandLine) parseInto(stmt string) string { + ident, stmt := parseNextIdentifier(stmt) + if strings.HasPrefix(stmt, ".") { + c.Database = ident + fmt.Printf("Using database %s\n", c.Database) + ident, stmt = parseNextIdentifier(stmt[1:]) + } + if strings.HasPrefix(stmt, " ") { + c.RetentionPolicy = ident + fmt.Printf("Using retention policy %s\n", c.RetentionPolicy) + return stmt[1:] + } + return stmt +} + +func (c *CommandLine) Insert(stmt string) error { + i, point := parseNextIdentifier(stmt) + if !strings.EqualFold(i, "insert") { + fmt.Printf("ERR: found %s, expected INSERT\n", i) + return nil + } + if i, r := parseNextIdentifier(point); strings.EqualFold(i, "into") { + point = c.parseInto(r) + } + _, err := c.Client.Write(client.BatchPoints{ + Points: []client.Point{ + client.Point{Raw: point}, + }, + Database: c.Database, + RetentionPolicy: c.RetentionPolicy, + Precision: "n", + WriteConsistency: client.ConsistencyAny, + }) + if err != nil { + fmt.Printf("ERR: %s\n", err) + if c.Database == "" { + fmt.Println("Note: error may be due to not setting a database or retention policy.") + fmt.Println(`Please set a database with the command "use " or`) + fmt.Println("INSERT INTO . ") + } + return err + } + return nil +} + +func (c *CommandLine) ExecuteQuery(query string) error { + response, err := c.Client.Query(client.Query{Command: query, Database: c.Database}) + if err != nil { + fmt.Printf("ERR: %s\n", err) + return err + } + c.FormatResponse(response, os.Stdout) + if err := response.Error(); err != nil { + fmt.Printf("ERR: %s\n", response.Error()) + if c.Database == "" { + fmt.Println("Warning: It is possible this error is due to not setting a database.") + fmt.Println(`Please set a database with the command "use ".`) + } + return err + } + return nil +} + +func (c *CommandLine) FormatResponse(response *client.Response, w io.Writer) { + switch c.Format { + case "json": + c.writeJSON(response, w) + case "csv": + c.writeCSV(response, w) + case "column": + c.writeColumns(response, w) + default: + fmt.Fprintf(w, "Unknown output format %q.\n", c.Format) + } +} + +func (c *CommandLine) writeJSON(response *client.Response, w io.Writer) { + var data []byte + var err error + if c.Pretty { + data, err = json.MarshalIndent(response, "", " ") + } else { + data, err = json.Marshal(response) + } + if err != nil { + fmt.Fprintf(w, "Unable to parse json: %s\n", err) + return + } + fmt.Fprintln(w, string(data)) +} + +func (c *CommandLine) writeCSV(response *client.Response, w io.Writer) { + csvw := csv.NewWriter(w) + for _, result := range response.Results { + // Create a tabbed writer for each result as they won't always line up + rows := c.formatResults(result, "\t") + for _, r := range rows { + csvw.Write(strings.Split(r, "\t")) + } + csvw.Flush() + } +} + +func (c *CommandLine) writeColumns(response *client.Response, w io.Writer) { + for _, result := range response.Results { + // Create a tabbed writer for each result a they won't always line up + w := new(tabwriter.Writer) + w.Init(os.Stdout, 0, 8, 1, '\t', 0) + csv := c.formatResults(result, "\t") + for _, r := range csv { + fmt.Fprintln(w, r) + } + w.Flush() + } +} + +// formatResults will behave differently if you are formatting for columns or csv +func (c *CommandLine) formatResults(result client.Result, separator string) []string { + rows := []string{} + // Create a tabbed writer for each result a they won't always line up + for i, row := range result.Series { + // gather tags + tags := []string{} + for k, v := range row.Tags { + tags = append(tags, fmt.Sprintf("%s=%s", k, v)) + sort.Strings(tags) + } + + columnNames := []string{} + + // Only put name/tags in a column if format is csv + if c.Format == "csv" { + if len(tags) > 0 { + columnNames = append([]string{"tags"}, columnNames...) + } + + if row.Name != "" { + columnNames = append([]string{"name"}, columnNames...) + } + } + + for _, column := range row.Columns { + columnNames = append(columnNames, column) + } + + // Output a line separator if we have more than one set or results and format is column + if i > 0 && c.Format == "column" { + rows = append(rows, "") + } + + // If we are column format, we break out the name/tag to seperate lines + if c.Format == "column" { + if row.Name != "" { + n := fmt.Sprintf("name: %s", row.Name) + rows = append(rows, n) + if len(tags) == 0 { + l := strings.Repeat("-", len(n)) + rows = append(rows, l) + } + } + if len(tags) > 0 { + t := fmt.Sprintf("tags: %s", (strings.Join(tags, ", "))) + rows = append(rows, t) + } + } + + rows = append(rows, strings.Join(columnNames, separator)) + + // if format is column, break tags to their own line/format + if c.Format == "column" && len(tags) > 0 { + lines := []string{} + for _, columnName := range columnNames { + lines = append(lines, strings.Repeat("-", len(columnName))) + } + rows = append(rows, strings.Join(lines, separator)) + } + + for _, v := range row.Values { + var values []string + if c.Format == "csv" { + if row.Name != "" { + values = append(values, row.Name) + } + if len(tags) > 0 { + values = append(values, strings.Join(tags, ",")) + } + } + + for _, vv := range v { + values = append(values, interfaceToString(vv)) + } + rows = append(rows, strings.Join(values, separator)) + } + // Outout a line separator if in column format + if c.Format == "column" { + rows = append(rows, "") + } + } + return rows +} + +func interfaceToString(v interface{}) string { + switch t := v.(type) { + case nil: + return "" + case bool: + return fmt.Sprintf("%v", v) + case int, int8, int16, int32, int64, uint, uint8, uint16, uint32, uint64, uintptr: + return fmt.Sprintf("%d", t) + case float32, float64: + return fmt.Sprintf("%v", t) + default: + return fmt.Sprintf("%v", t) + } +} + +func (c *CommandLine) Settings() { + w := new(tabwriter.Writer) + w.Init(os.Stdout, 0, 8, 1, '\t', 0) + if c.Port > 0 { + fmt.Fprintf(w, "Host\t%s:%d\n", c.Host, c.Port) + } else { + fmt.Fprintf(w, "Host\t%s\n", c.Host) + } + fmt.Fprintf(w, "Username\t%s\n", c.Username) + fmt.Fprintf(w, "Database\t%s\n", c.Database) + fmt.Fprintf(w, "Pretty\t%v\n", c.Pretty) + fmt.Fprintf(w, "Format\t%s\n", c.Format) + fmt.Fprintln(w) + w.Flush() +} + +func (c *CommandLine) help() { + fmt.Println(`Usage: + connect connect to another node + auth prompt for username and password + pretty toggle pretty print + use set current databases + format set the output format: json, csv, or column + settings output the current settings for the shell + exit quit the influx shell + + show databases show database names + show series show series information + show measurements show measurement information + show tag keys show tag key information + show tag values show tag value information + + a full list of influxql commands can be found at: + https://influxdb.com/docs/v0.9/query_language/spec.html +`) +} + +func (c *CommandLine) gopher() { + fmt.Println(` + .-::-::://:-::- .:/++/' + '://:-''/oo+//++o+/.://o- ./+: + .:-. '++- .o/ '+yydhy' o- + .:/. .h: :osoys .smMN- :/ + -/:.' s- /MMMymh. '/y/ s' + -+s:'''' d -mMMms// '-/o: + -/++/++/////:. o: '... s- :s. + :+-+s-' ':/' 's- /+ 'o: + '+-'o: /ydhsh. '//. '-o- o- + .y. o: .MMMdm+y ':+++:::/+:.' s: + .-h/ y- 'sdmds'h -+ydds:::-.' 'h. + .//-.d' o: '.' 'dsNMMMNh:.:++' :y + +y. 'd 's. .s:mddds: ++ o/ + 'N- odd 'o/. './o-s-' .---+++' o- + 'N' yNd .://:/:::::. -s -+/s/./s' 'o/' + so' .h '''' ////s: '+. .s +y' + os/-.y' 's' 'y::+ +d' + '.:o/ -+:-:.' so.---.' + o' 'd-.''/s' + .s' :y.''.y + -s mo:::' + :: yh + // '''' /M' + o+ .s///:/. 'N: + :+ /: -s' ho + 's- -/s/:+/.+h' +h + ys' ':' '-. -d + oh .h + /o .s + s. .h + -y .d + m/ -h + +d /o + 'N- y: + h: m. + s- -d + o- s+ + +- 'm' + s/ oo--. + y- /s ':+' + s' 'od--' .d: + -+ ':o: ':+-/+ + y- .:+- ' + //o- '.:+/. + .-:+/' ''-/+/. + ./:' ''.:o+/-' + .+o:/:/+-' ''.-+ooo/-' + o: -h///++////-. + /: .o/ + //+ 'y + ./sooy. + +`) +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx/main_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx/main_test.go new file mode 100644 index 000000000..eb58cb1e8 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx/main_test.go @@ -0,0 +1,194 @@ +package main_test + +import ( + "encoding/json" + "net/http" + "net/http/httptest" + "net/url" + "testing" + + "github.com/influxdb/influxdb/client" + main "github.com/influxdb/influxdb/cmd/influx" +) + +func TestParseCommand_CommandsExist(t *testing.T) { + t.Parallel() + c := main.CommandLine{} + tests := []struct { + cmd string + }{ + {cmd: "gopher"}, + {cmd: "connect"}, + {cmd: "help"}, + {cmd: "pretty"}, + {cmd: "use"}, + {cmd: ""}, // test that a blank command just returns + } + for _, test := range tests { + if !c.ParseCommand(test.cmd) { + t.Fatalf(`Command failed for %q.`, test.cmd) + } + } +} + +func TestParseCommand_TogglePretty(t *testing.T) { + t.Parallel() + c := main.CommandLine{} + if c.Pretty { + t.Fatalf(`Pretty should be false.`) + } + c.ParseCommand("pretty") + if !c.Pretty { + t.Fatalf(`Pretty should be true.`) + } + c.ParseCommand("pretty") + if c.Pretty { + t.Fatalf(`Pretty should be false.`) + } +} + +func TestParseCommand_Exit(t *testing.T) { + t.Parallel() + c := main.CommandLine{} + tests := []struct { + cmd string + }{ + {cmd: "exit"}, + {cmd: " exit"}, + {cmd: "exit "}, + {cmd: "Exit "}, + } + + for _, test := range tests { + if c.ParseCommand(test.cmd) { + t.Fatalf(`Command "exit" failed for %q.`, test.cmd) + } + } +} + +func TestParseCommand_Use(t *testing.T) { + t.Parallel() + c := main.CommandLine{} + tests := []struct { + cmd string + }{ + {cmd: "use db"}, + {cmd: " use db"}, + {cmd: "use db "}, + {cmd: "use db;"}, + {cmd: "use db; "}, + {cmd: "Use db"}, + } + + for _, test := range tests { + if !c.ParseCommand(test.cmd) { + t.Fatalf(`Command "use" failed for %q.`, test.cmd) + } + + if c.Database != "db" { + t.Fatalf(`Command "use" changed database to %q. Expected db`, c.Database) + } + } +} + +func TestParseCommand_Insert(t *testing.T) { + t.Parallel() + ts := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { + var data client.Response + w.WriteHeader(http.StatusNoContent) + _ = json.NewEncoder(w).Encode(data) + })) + defer ts.Close() + + u, _ := url.Parse(ts.URL) + config := client.Config{URL: *u} + c, err := client.NewClient(config) + if err != nil { + t.Fatalf("unexpected error. expected %v, actual %v", nil, err) + } + m := main.CommandLine{Client: c} + + tests := []struct { + cmd string + }{ + {cmd: "INSERT cpu,host=serverA,region=us-west value=1.0"}, + {cmd: " INSERT cpu,host=serverA,region=us-west value=1.0"}, + {cmd: "INSERT cpu,host=serverA,region=us-west value=1.0"}, + {cmd: "insert cpu,host=serverA,region=us-west value=1.0 "}, + {cmd: "insert"}, + {cmd: "Insert "}, + {cmd: "insert c"}, + {cmd: "insert int"}, + } + + for _, test := range tests { + if !m.ParseCommand(test.cmd) { + t.Fatalf(`Command "insert" failed for %q.`, test.cmd) + } + } +} + +func TestParseCommand_InsertInto(t *testing.T) { + t.Parallel() + ts := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { + var data client.Response + w.WriteHeader(http.StatusNoContent) + _ = json.NewEncoder(w).Encode(data) + })) + defer ts.Close() + + u, _ := url.Parse(ts.URL) + config := client.Config{URL: *u} + c, err := client.NewClient(config) + if err != nil { + t.Fatalf("unexpected error. expected %v, actual %v", nil, err) + } + m := main.CommandLine{Client: c} + + tests := []struct { + cmd, db, rp string + }{ + { + cmd: `INSERT INTO test cpu,host=serverA,region=us-west value=1.0`, + db: "", + rp: "test", + }, + { + cmd: ` INSERT INTO .test cpu,host=serverA,region=us-west value=1.0`, + db: "", + rp: "test", + }, + { + cmd: `INSERT INTO "test test" cpu,host=serverA,region=us-west value=1.0`, + db: "", + rp: "test test", + }, + { + cmd: `Insert iNTO test.test cpu,host=serverA,region=us-west value=1.0`, + db: "test", + rp: "test", + }, + { + cmd: `insert into "test test" cpu,host=serverA,region=us-west value=1.0`, + db: "test", + rp: "test test", + }, + { + cmd: `insert into "d b"."test test" cpu,host=serverA,region=us-west value=1.0`, + db: "d b", + rp: "test test", + }, + } + + for _, test := range tests { + if !m.ParseCommand(test.cmd) { + t.Fatalf(`Command "insert into" failed for %q.`, test.cmd) + } + if m.Database != test.db { + t.Fatalf(`Command "insert into" db parsing failed, expected: %q, actual: %q`, test.db, m.Database) + } + if m.RetentionPolicy != test.rp { + t.Fatalf(`Command "insert into" rp parsing failed, expected: %q, actual: %q`, test.rp, m.RetentionPolicy) + } + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx_stress/influx_stress.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx_stress/influx_stress.go new file mode 100644 index 000000000..7b9bc9873 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influx_stress/influx_stress.go @@ -0,0 +1,154 @@ +package main + +import ( + "flag" + "fmt" + "math/rand" + "net/url" + "runtime" + "sort" + "sync" + "time" + + "github.com/influxdb/influxdb/client" +) + +var ( + batchSize = flag.Int("batchsize", 5000, "number of points per batch") + seriesCount = flag.Int("series", 100000, "number of unique series to create") + pointCount = flag.Int("points", 100, "number of points per series to create") + concurrency = flag.Int("concurrency", 10, "number of simultaneous writes to run") + batchInterval = flag.Duration("batchinterval", 0*time.Second, "duration between batches") + database = flag.String("database", "stress", "name of database") + address = flag.String("addr", "localhost:8086", "IP address and port of database (e.g., localhost:8086)") +) + +func main() { + flag.Parse() + runtime.GOMAXPROCS(runtime.NumCPU()) + + startTime := time.Now() + counter := NewConcurrencyLimiter(*concurrency) + + u, _ := url.Parse(fmt.Sprintf("http://%s", *address)) + c, err := client.NewClient(client.Config{URL: *u}) + if err != nil { + panic(err) + } + + var mu sync.Mutex + var wg sync.WaitGroup + responseTimes := make([]int, 0) + + totalPoints := 0 + + batch := &client.BatchPoints{ + Database: *database, + WriteConsistency: "any", + Time: time.Now(), + Precision: "n", + } + for i := 1; i <= *pointCount; i++ { + for j := 1; j <= *seriesCount; j++ { + p := client.Point{ + Measurement: "cpu", + Tags: map[string]string{"region": "uswest", "host": fmt.Sprintf("host-%d", j)}, + Fields: map[string]interface{}{"value": rand.Float64()}, + } + batch.Points = append(batch.Points, p) + if len(batch.Points) >= *batchSize { + wg.Add(1) + counter.Increment() + totalPoints += len(batch.Points) + go func(b *client.BatchPoints, total int) { + st := time.Now() + if _, err := c.Write(*b); err != nil { + fmt.Println("ERROR: ", err.Error()) + } else { + mu.Lock() + responseTimes = append(responseTimes, int(time.Since(st).Nanoseconds())) + mu.Unlock() + } + wg.Done() + counter.Decrement() + if total%500000 == 0 { + fmt.Printf("%d total points. %d in %s\n", total, *batchSize, time.Since(st)) + } + }(batch, totalPoints) + + batch = &client.BatchPoints{ + Database: *database, + WriteConsistency: "any", + Precision: "n", + Time: time.Now(), + } + } + } + } + + wg.Wait() + sort.Sort(sort.Reverse(sort.IntSlice(responseTimes))) + + total := int64(0) + for _, t := range responseTimes { + total += int64(t) + } + mean := total / int64(len(responseTimes)) + + fmt.Printf("Wrote %d points at average rate of %.0f\n", totalPoints, float64(totalPoints)/time.Since(startTime).Seconds()) + fmt.Println("Average response time: ", time.Duration(mean)) + fmt.Println("Slowest response times:") + for _, r := range responseTimes[:100] { + fmt.Println(time.Duration(r)) + } +} + +// ConcurrencyLimiter is a go routine safe struct that can be used to +// ensure that no more than a specifid max number of goroutines are +// executing. +type ConcurrencyLimiter struct { + inc chan chan struct{} + dec chan struct{} + max int + count int +} + +// NewConcurrencyLimiter returns a configured limiter that will +// ensure that calls to Increment will block if the max is hit. +func NewConcurrencyLimiter(max int) *ConcurrencyLimiter { + c := &ConcurrencyLimiter{ + inc: make(chan chan struct{}), + dec: make(chan struct{}, max), + max: max, + } + go c.handleLimits() + return c +} + +// Increment will increase the count of running goroutines by 1. +// if the number is currently at the max, the call to Increment +// will block until another goroutine decrements. +func (c *ConcurrencyLimiter) Increment() { + r := make(chan struct{}) + c.inc <- r + <-r +} + +// Decrement will reduce the count of running goroutines by 1 +func (c *ConcurrencyLimiter) Decrement() { + c.dec <- struct{}{} +} + +// handleLimits runs in a goroutine to manage the count of +// running goroutines. +func (c *ConcurrencyLimiter) handleLimits() { + for { + r := <-c.inc + if c.count >= c.max { + <-c.dec + c.count-- + } + c.count++ + r <- struct{}{} + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/backup/backup.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/backup/backup.go new file mode 100644 index 000000000..c88652f75 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/backup/backup.go @@ -0,0 +1,170 @@ +package backup + +import ( + "encoding/json" + "errors" + "flag" + "fmt" + "io" + "log" + "net" + "os" + + "github.com/influxdb/influxdb/services/snapshotter" + "github.com/influxdb/influxdb/snapshot" +) + +// Suffix is a suffix added to the backup while it's in-process. +const Suffix = ".pending" + +// Command represents the program execution for "influxd backup". +type Command struct { + // The logger passed to the ticker during execution. + Logger *log.Logger + + // Standard input/output, overridden for testing. + Stderr io.Writer +} + +// NewCommand returns a new instance of Command with default settings. +func NewCommand() *Command { + return &Command{ + Stderr: os.Stderr, + } +} + +// Run executes the program. +func (cmd *Command) Run(args ...string) error { + // Set up logger. + cmd.Logger = log.New(cmd.Stderr, "", log.LstdFlags) + cmd.Logger.Printf("influxdb backup") + + // Parse command line arguments. + host, path, err := cmd.parseFlags(args) + if err != nil { + return err + } + + // Retrieve snapshot from local file. + m, err := snapshot.ReadFileManifest(path) + if err != nil && !os.IsNotExist(err) { + return fmt.Errorf("read file snapshot: %s", err) + } + + // Determine temporary path to download to. + tmppath := path + Suffix + + // Calculate path of next backup file. + // This uses the path if it doesn't exist. + // Otherwise it appends an autoincrementing number. + path, err = cmd.nextPath(path) + if err != nil { + return fmt.Errorf("next path: %s", err) + } + + // Retrieve snapshot. + if err := cmd.download(host, m, tmppath); err != nil { + return fmt.Errorf("download: %s", err) + } + + // Rename temporary file to final path. + if err := os.Rename(tmppath, path); err != nil { + return fmt.Errorf("rename: %s", err) + } + + // TODO: Check file integrity. + + // Notify user of completion. + cmd.Logger.Println("backup complete") + + return nil +} + +// parseFlags parses and validates the command line arguments. +func (cmd *Command) parseFlags(args []string) (host string, path string, err error) { + fs := flag.NewFlagSet("", flag.ContinueOnError) + fs.StringVar(&host, "host", "localhost:8088", "") + fs.SetOutput(cmd.Stderr) + fs.Usage = cmd.printUsage + if err := fs.Parse(args); err != nil { + return "", "", err + } + + // Ensure that only one arg is specified. + if fs.NArg() == 0 { + return "", "", errors.New("snapshot path required") + } else if fs.NArg() != 1 { + return "", "", errors.New("only one snapshot path allowed") + } + path = fs.Arg(0) + + return host, path, nil +} + +// nextPath returns the next file to write to. +func (cmd *Command) nextPath(path string) (string, error) { + // Use base path if it doesn't exist. + if _, err := os.Stat(path); os.IsNotExist(err) { + return path, nil + } else if err != nil { + return "", err + } + + // Otherwise iterate through incremental files until one is available. + for i := 0; ; i++ { + s := fmt.Sprintf(path+".%d", i) + if _, err := os.Stat(s); os.IsNotExist(err) { + return s, nil + } else if err != nil { + return "", err + } + } +} + +// download downloads a snapshot from a host to a given path. +func (cmd *Command) download(host string, m *snapshot.Manifest, path string) error { + // Create local file to write to. + f, err := os.Create(path) + if err != nil { + return fmt.Errorf("open temp file: %s", err) + } + defer f.Close() + + // Connect to snapshotter service. + conn, err := net.Dial("tcp", host) + if err != nil { + return err + } + defer conn.Close() + + // Send snapshotter marker byte. + if _, err := conn.Write([]byte{snapshotter.MuxHeader}); err != nil { + return fmt.Errorf("write snapshot header byte: %s", err) + } + + // Write the manifest we currently have. + if err := json.NewEncoder(conn).Encode(m); err != nil { + return fmt.Errorf("encode snapshot manifest: %s", err) + } + + // Read snapshot from the connection. + if _, err := io.Copy(f, conn); err != nil { + return fmt.Errorf("copy snapshot to file: %s", err) + } + + // FIXME(benbjohnson): Verify integrity of snapshot. + + return nil +} + +// printUsage prints the usage message to STDERR. +func (cmd *Command) printUsage() { + fmt.Fprintf(cmd.Stderr, `usage: influxd backup [flags] PATH + +backup downloads a snapshot of a data node and saves it to disk. + + -host + The host to connect to snapshot. + Defaults to 127.0.0.1:8088. +`) +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/backup/backup_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/backup/backup_test.go new file mode 100644 index 000000000..15db96449 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/backup/backup_test.go @@ -0,0 +1,125 @@ +package backup_test + +/* +import ( + "bytes" + "net/http" + "net/http/httptest" + "os" + "strings" + "testing" + + "github.com/influxdb/influxdb" + "github.com/influxdb/influxdb/cmd/influxd" +) + +// Ensure the backup can download from the server and save to disk. +func TestBackupCommand(t *testing.T) { + // Mock the backup endpoint. + s := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { + if r.URL.Path != "/data/snapshot" { + t.Fatalf("unexpected url path: %s", r.URL.Path) + } + + // Write a simple snapshot to the buffer. + sw := influxdb.NewSnapshotWriter() + sw.Snapshot = &influxdb.Snapshot{Files: []influxdb.SnapshotFile{ + {Name: "meta", Size: 5, Index: 10}, + }} + sw.FileWriters["meta"] = influxdb.NopWriteToCloser(bytes.NewBufferString("55555")) + if _, err := sw.WriteTo(w); err != nil { + t.Fatal(err) + } + })) + defer s.Close() + + // Create a temp path and remove incremental backups at the end. + path := tempfile() + defer os.Remove(path) + defer os.Remove(path + ".0") + defer os.Remove(path + ".1") + + // Execute the backup against the mock server. + for i := 0; i < 3; i++ { + if err := NewBackupCommand().Run("-host", s.URL, path); err != nil { + t.Fatal(err) + } + } + + // Verify snapshot and two incremental snapshots were written. + if _, err := os.Stat(path); err != nil { + t.Fatalf("snapshot not found: %s", err) + } else if _, err = os.Stat(path + ".0"); err != nil { + t.Fatalf("incremental snapshot(0) not found: %s", err) + } else if _, err = os.Stat(path + ".1"); err != nil { + t.Fatalf("incremental snapshot(1) not found: %s", err) + } +} + +// Ensure the backup command returns an error if flags cannot be parsed. +func TestBackupCommand_ErrFlagParse(t *testing.T) { + cmd := NewBackupCommand() + if err := cmd.Run("-bad-flag"); err == nil || err.Error() != `flag provided but not defined: -bad-flag` { + t.Fatal(err) + } else if !strings.Contains(cmd.Stderr.String(), "usage") { + t.Fatal("usage message not displayed") + } +} + +// Ensure the backup command returns an error if the host cannot be parsed. +func TestBackupCommand_ErrInvalidHostURL(t *testing.T) { + if err := NewBackupCommand().Run("-host", "http://%f"); err == nil || err.Error() != `parse host url: parse http://%f: hexadecimal escape in host` { + t.Fatal(err) + } +} + +// Ensure the backup command returns an error if the output path is not specified. +func TestBackupCommand_ErrPathRequired(t *testing.T) { + if err := NewBackupCommand().Run("-host", "//localhost"); err == nil || err.Error() != `snapshot path required` { + t.Fatal(err) + } +} + +// Ensure the backup returns an error if it cannot connect to the server. +func TestBackupCommand_ErrConnectionRefused(t *testing.T) { + // Start and immediately stop a server so we have a dead port. + s := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {})) + s.Close() + + // Execute the backup command. + path := tempfile() + defer os.Remove(path) + if err := NewBackupCommand().Run("-host", s.URL, path); err == nil || + !(strings.Contains(err.Error(), `connection refused`) || strings.Contains(err.Error(), `No connection could be made`)) { + t.Fatal(err) + } +} + +// Ensure the backup returns any non-200 status codes. +func TestBackupCommand_ErrServerError(t *testing.T) { + s := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) { + w.WriteHeader(http.StatusInternalServerError) + })) + defer s.Close() + + // Execute the backup command. + path := tempfile() + defer os.Remove(path) + if err := NewBackupCommand().Run("-host", s.URL, path); err == nil || err.Error() != `download: snapshot error: status=500` { + t.Fatal(err) + } +} + +// BackupCommand is a test wrapper for main.BackupCommand. +type BackupCommand struct { + *main.BackupCommand + Stderr bytes.Buffer +} + +// NewBackupCommand returns a new instance of BackupCommand. +func NewBackupCommand() *BackupCommand { + cmd := &BackupCommand{BackupCommand: main.NewBackupCommand()} + cmd.BackupCommand.Stderr = &cmd.Stderr + return cmd +} +*/ diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/help/help.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/help/help.go new file mode 100644 index 000000000..3f6bbfb08 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/help/help.go @@ -0,0 +1,46 @@ +package help + +import ( + "fmt" + "io" + "os" + "strings" +) + +// Command displays help for command-line sub-commands. +type Command struct { + Stdout io.Writer +} + +// NewCommand returns a new instance of Command. +func NewCommand() *Command { + return &Command{ + Stdout: os.Stdout, + } +} + +// Run executes the command. +func (cmd *Command) Run(args ...string) error { + fmt.Fprintln(cmd.Stdout, strings.TrimSpace(usage)) + return nil +} + +const usage = ` +Configure and start an InfluxDB server. + +Usage: + + influxd [[command] [arguments]] + +The commands are: + + backup downloads a snapshot of a data node and saves it to disk + config display the default configuration + restore uses a snapshot of a data node to rebuild a cluster + run run node with existing configuration + version displays the InfluxDB version + +"run" is the default command. + +Use "influxd help [command]" for more information about a command. +` diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/main.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/main.go new file mode 100644 index 000000000..9748493a5 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/main.go @@ -0,0 +1,200 @@ +package main + +import ( + "flag" + "fmt" + "io" + "log" + "math/rand" + "os" + "os/signal" + "strings" + "syscall" + "time" + + "github.com/influxdb/influxdb/cmd/influxd/backup" + "github.com/influxdb/influxdb/cmd/influxd/help" + "github.com/influxdb/influxdb/cmd/influxd/restore" + "github.com/influxdb/influxdb/cmd/influxd/run" +) + +// These variables are populated via the Go linker. +var ( + version string = "0.9" + commit string + branch string +) + +func init() { + // If commit or branch are not set, make that clear. + if commit == "" { + commit = "unknown" + } + if branch == "" { + branch = "unknown" + } +} + +func main() { + rand.Seed(time.Now().UnixNano()) + + m := NewMain() + if err := m.Run(os.Args[1:]...); err != nil { + fmt.Fprintln(os.Stderr, err) + os.Exit(1) + } +} + +// Main represents the program execution. +type Main struct { + Logger *log.Logger + + Stdin io.Reader + Stdout io.Writer + Stderr io.Writer +} + +// NewMain return a new instance of Main. +func NewMain() *Main { + return &Main{ + Logger: log.New(os.Stderr, "[run] ", log.LstdFlags), + Stdin: os.Stdin, + Stdout: os.Stdout, + Stderr: os.Stderr, + } +} + +// Run determines and runs the command specified by the CLI args. +func (m *Main) Run(args ...string) error { + name, args := ParseCommandName(args) + + // Extract name from args. + switch name { + case "", "run": + cmd := run.NewCommand() + + // Tell the server the build details. + cmd.Version = version + cmd.Commit = commit + cmd.Branch = branch + + if err := cmd.Run(args...); err != nil { + return fmt.Errorf("run: %s", err) + } + + signalCh := make(chan os.Signal, 1) + signal.Notify(signalCh, os.Interrupt, syscall.SIGTERM) + m.Logger.Println("Listening for signals") + + // Block until one of the signals above is received + select { + case <-signalCh: + m.Logger.Println("Signal received, initializing clean shutdown...") + go func() { + cmd.Close() + }() + } + + // Block again until another signal is received, a shutdown timeout elapses, + // or the Command is gracefully closed + m.Logger.Println("Waiting for clean shutdown...") + select { + case <-signalCh: + m.Logger.Println("second signal received, initializing hard shutdown") + case <-time.After(time.Second * 30): + m.Logger.Println("time limit reached, initializing hard shutdown") + case <-cmd.Closed: + m.Logger.Println("server shutdown completed") + } + + // goodbye. + + case "backup": + name := backup.NewCommand() + if err := name.Run(args...); err != nil { + return fmt.Errorf("backup: %s", err) + } + case "restore": + name := restore.NewCommand() + if err := name.Run(args...); err != nil { + return fmt.Errorf("restore: %s", err) + } + case "config": + if err := run.NewPrintConfigCommand().Run(args...); err != nil { + return fmt.Errorf("config: %s", err) + } + case "version": + if err := NewVersionCommand().Run(args...); err != nil { + return fmt.Errorf("version: %s", err) + } + case "help": + if err := help.NewCommand().Run(args...); err != nil { + return fmt.Errorf("help: %s", err) + } + default: + return fmt.Errorf(`unknown command "%s"`+"\n"+`Run 'influxd help' for usage`+"\n\n", name) + } + + return nil +} + +// ParseCommandName extracts the command name and args from the args list. +func ParseCommandName(args []string) (string, []string) { + // Retrieve command name as first argument. + var name string + if len(args) > 0 && !strings.HasPrefix(args[0], "-") { + name = args[0] + } + + // Special case -h immediately following binary name + if len(args) > 0 && args[0] == "-h" { + name = "help" + } + + // If command is "help" and has an argument then rewrite args to use "-h". + if name == "help" && len(args) > 1 { + args[0], args[1] = args[1], "-h" + name = args[0] + } + + // If a named command is specified then return it with its arguments. + if name != "" { + return name, args[1:] + } + return "", args +} + +// Command represents the command executed by "influxd version". +type VersionCommand struct { + Stdout io.Writer + Stderr io.Writer +} + +// NewVersionCommand return a new instance of VersionCommand. +func NewVersionCommand() *VersionCommand { + return &VersionCommand{ + Stdout: os.Stdout, + Stderr: os.Stderr, + } +} + +// Run prints the current version and commit info. +func (cmd *VersionCommand) Run(args ...string) error { + // Parse flags in case -h is specified. + fs := flag.NewFlagSet("", flag.ContinueOnError) + fs.Usage = func() { fmt.Fprintln(cmd.Stderr, strings.TrimSpace(versionUsage)) } + if err := fs.Parse(args); err != nil { + return err + } + + // Print version info. + fmt.Fprintf(cmd.Stdout, "InfluxDB v%s (git: %s %s)\n", version, branch, commit) + + return nil +} + +var versionUsage = ` +usage: version + + version displays the InfluxDB version, build branch and git commit hash +` diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/restore/restore.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/restore/restore.go new file mode 100644 index 000000000..5a95f8726 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/restore/restore.go @@ -0,0 +1,250 @@ +package restore + +import ( + "bytes" + "errors" + "flag" + "fmt" + "io" + "net" + "os" + "path/filepath" + + "github.com/BurntSushi/toml" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/snapshot" + "github.com/influxdb/influxdb/tsdb" +) + +// Command represents the program execution for "influxd restore". +type Command struct { + Stdout io.Writer + Stderr io.Writer +} + +// NewCommand returns a new instance of Command with default settings. +func NewCommand() *Command { + return &Command{ + Stdout: os.Stdout, + Stderr: os.Stderr, + } +} + +// Run executes the program. +func (cmd *Command) Run(args ...string) error { + config, path, err := cmd.parseFlags(args) + if err != nil { + return err + } + + return cmd.Restore(config, path) +} + +func (cmd *Command) Restore(config *Config, path string) error { + // Remove meta and data directories. + if err := os.RemoveAll(config.Meta.Dir); err != nil { + return fmt.Errorf("remove meta dir: %s", err) + } else if err := os.RemoveAll(config.Data.Dir); err != nil { + return fmt.Errorf("remove data dir: %s", err) + } + + // Open snapshot file and all incremental backups. + mr, files, err := snapshot.OpenFileMultiReader(path) + if err != nil { + return fmt.Errorf("open multireader: %s", err) + } + defer closeAll(files) + + // Unpack files from archive. + if err := cmd.unpack(mr, config); err != nil { + return fmt.Errorf("unpack: %s", err) + } + + // Notify user of completion. + fmt.Fprintf(os.Stdout, "restore complete using %s", path) + return nil +} + +// parseFlags parses and validates the command line arguments. +func (cmd *Command) parseFlags(args []string) (*Config, string, error) { + fs := flag.NewFlagSet("", flag.ContinueOnError) + configPath := fs.String("config", "", "") + fs.SetOutput(cmd.Stderr) + fs.Usage = cmd.printUsage + if err := fs.Parse(args); err != nil { + return nil, "", err + } + + // Parse configuration file from disk. + if *configPath == "" { + return nil, "", fmt.Errorf("config required") + } + + // Parse config. + config := Config{ + Meta: meta.NewConfig(), + Data: tsdb.NewConfig(), + } + if _, err := toml.DecodeFile(*configPath, &config); err != nil { + return nil, "", err + } + + // Require output path. + path := fs.Arg(0) + if path == "" { + return nil, "", fmt.Errorf("snapshot path required") + } + + return &config, path, nil +} + +func closeAll(a []io.Closer) { + for _, c := range a { + _ = c.Close() + } +} + +// unpack expands the files in the snapshot archive into a directory. +func (cmd *Command) unpack(mr *snapshot.MultiReader, config *Config) error { + // Loop over files and extract. + for { + // Read entry header. + sf, err := mr.Next() + if err == io.EOF { + break + } else if err != nil { + return fmt.Errorf("next: entry=%s, err=%s", sf.Name, err) + } + + // Log progress. + fmt.Fprintf(os.Stdout, "unpacking: %s (%d bytes)\n", sf.Name, sf.Size) + + // Handle meta and tsdb files separately. + switch sf.Name { + case "meta": + if err := cmd.unpackMeta(mr, sf, config); err != nil { + return fmt.Errorf("meta: %s", err) + } + default: + if err := cmd.unpackData(mr, sf, config); err != nil { + return fmt.Errorf("data: %s", err) + } + } + } + + return nil +} + +// unpackMeta reads the metadata from the snapshot and initializes a raft +// cluster and replaces the root metadata. +func (cmd *Command) unpackMeta(mr *snapshot.MultiReader, sf snapshot.File, config *Config) error { + // Read meta into buffer. + var buf bytes.Buffer + if _, err := io.CopyN(&buf, mr, sf.Size); err != nil { + return fmt.Errorf("copy: %s", err) + } + + // Unpack into metadata. + var data meta.Data + if err := data.UnmarshalBinary(buf.Bytes()); err != nil { + return fmt.Errorf("unmarshal: %s", err) + } + + // Copy meta config and remove peers so it starts in single mode. + c := config.Meta + c.Peers = nil + + // Initialize meta store. + store := meta.NewStore(config.Meta) + store.RaftListener = newNopListener() + store.ExecListener = newNopListener() + + // Determine advertised address. + _, port, err := net.SplitHostPort(config.Meta.BindAddress) + if err != nil { + return fmt.Errorf("split bind address: %s", err) + } + hostport := net.JoinHostPort(config.Meta.Hostname, port) + + // Resolve address. + addr, err := net.ResolveTCPAddr("tcp", hostport) + if err != nil { + return fmt.Errorf("resolve tcp: addr=%s, err=%s", hostport, err) + } + store.Addr = addr + + // Open the meta store. + if err := store.Open(); err != nil { + return fmt.Errorf("open store: %s", err) + } + defer store.Close() + + // Wait for the store to be ready or error. + select { + case <-store.Ready(): + case err := <-store.Err(): + return err + } + + // Force set the full metadata. + if err := store.SetData(&data); err != nil { + return fmt.Errorf("set data: %s", err) + } + + return nil +} + +func (cmd *Command) unpackData(mr *snapshot.MultiReader, sf snapshot.File, config *Config) error { + path := filepath.Join(config.Data.Dir, sf.Name) + // Create parent directory for output file. + if err := os.MkdirAll(filepath.Dir(path), 0777); err != nil { + return fmt.Errorf("mkdir: entry=%s, err=%s", sf.Name, err) + } + + // Create output file. + f, err := os.Create(path) + if err != nil { + return fmt.Errorf("create: entry=%s, err=%s", sf.Name, err) + } + defer f.Close() + + // Copy contents from reader. + if _, err := io.CopyN(f, mr, sf.Size); err != nil { + return fmt.Errorf("copy: entry=%s, err=%s", sf.Name, err) + } + + return nil +} + +// printUsage prints the usage message to STDERR. +func (cmd *Command) printUsage() { + fmt.Fprintf(cmd.Stderr, `usage: influxd restore [flags] PATH + +restore uses a snapshot of a data node to rebuild a cluster. + + -config + Set the path to the configuration file. +`) +} + +// Config represents a partial config for rebuilding the server. +type Config struct { + Meta *meta.Config `toml:"meta"` + Data tsdb.Config `toml:"data"` +} + +type nopListener struct { + closing chan struct{} +} + +func newNopListener() *nopListener { + return &nopListener{make(chan struct{})} +} + +func (ln *nopListener) Accept() (net.Conn, error) { + <-ln.closing + return nil, errors.New("listener closing") +} + +func (ln *nopListener) Close() error { close(ln.closing); return nil } +func (ln *nopListener) Addr() net.Addr { return nil } diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/restore/restore_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/restore/restore_test.go new file mode 100644 index 000000000..6e3143f25 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/restore/restore_test.go @@ -0,0 +1,155 @@ +package restore_test + +/* +import ( + "bytes" + "io/ioutil" + "os" + "path/filepath" + "reflect" + "testing" + "time" + + main "github.com/influxdb/influxdb/cmd/influxd" + "github.com/influxdb/influxdb/tsdb" +) + +func newConfig(path string, port int) main.Config { + config := main.NewConfig() + config.Port = port + config.Broker.Enabled = true + config.Broker.Dir = filepath.Join(path, "broker") + + config.Data.Enabled = true + config.Data.Dir = filepath.Join(path, "data") + return *config +} + +// Ensure the restore command can expand a snapshot and bootstrap a broker. +func TestRestoreCommand(t *testing.T) { + if testing.Short() { + t.Skip("skipping TestRestoreCommand") + } + + now := time.Now() + + // Create root path to server. + path := tempfile() + defer os.Remove(path) + + // Parse configuration. + config := newConfig(path, 8900) + + // Start server. + cmd := main.NewRunCommand() + node := cmd.Open(&config, "") + if node.Broker == nil { + t.Fatal("cannot run broker") + } else if node.DataNode == nil { + t.Fatal("cannot run server") + } + b := node.Broker + s := node.DataNode + + // Create data. + if err := s.CreateDatabase("db"); err != nil { + t.Fatalf("cannot create database: %s", err) + } + if index, err := s.WriteSeries("db", "default", []tsdb.Point{tsdb.NewPoint("cpu", nil, map[string]interface{}{"value": float64(100)}, now)}); err != nil { + t.Fatalf("cannot write series: %s", err) + } else if err = s.Sync(1, index); err != nil { + t.Fatalf("shard sync: %s", err) + } + + // Create snapshot writer. + sw, err := s.CreateSnapshotWriter() + if err != nil { + t.Fatalf("create snapshot writer: %s", err) + } + + // Snapshot to file. + sspath := tempfile() + f, err := os.Create(sspath) + if err != nil { + t.Fatal(err) + } + sw.WriteTo(f) + f.Close() + + // Stop server. + node.Close() + + // Remove data & broker directories. + if err := os.RemoveAll(path); err != nil { + t.Fatalf("remove: %s", err) + } + + // Execute the restore. + if err := NewRestoreCommand().Restore(&config, sspath); err != nil { + t.Fatal(err) + } + + // Rewrite config to a new port and re-parse. + config = newConfig(path, 8910) + + // Restart server. + cmd = main.NewRunCommand() + node = cmd.Open(&config, "") + if b == nil { + t.Fatal("cannot run broker") + } else if s == nil { + t.Fatal("cannot run server") + } + b = node.Broker + s = node.DataNode + + // Write new data. + if err := s.CreateDatabase("newdb"); err != nil { + t.Fatalf("cannot create new database: %s", err) + } + if index, err := s.WriteSeries("newdb", "default", []tsdb.Point{tsdb.NewPoint("mem", nil, map[string]interface{}{"value": float64(1000)}, now)}); err != nil { + t.Fatalf("cannot write new series: %s", err) + } else if err = s.Sync(2, index); err != nil { + t.Fatalf("shard sync: %s", err) + } + + // Read series data. + if v, err := s.ReadSeries("db", "default", "cpu", nil, now); err != nil { + t.Fatal(err) + } else if !reflect.DeepEqual(v, map[string]interface{}{"value": float64(100)}) { + t.Fatalf("read series(0) mismatch: %#v", v) + } + + // Read new series data. + if v, err := s.ReadSeries("newdb", "default", "mem", nil, now); err != nil { + t.Fatal(err) + } else if !reflect.DeepEqual(v, map[string]interface{}{"value": float64(1000)}) { + t.Fatalf("read series(1) mismatch: %#v", v) + } + + // Stop server. + node.Close() +} + +// RestoreCommand is a test wrapper for main.RestoreCommand. +type RestoreCommand struct { + *main.RestoreCommand + Stderr bytes.Buffer +} + +// NewRestoreCommand returns a new instance of RestoreCommand. +func NewRestoreCommand() *RestoreCommand { + cmd := &RestoreCommand{RestoreCommand: main.NewRestoreCommand()} + cmd.RestoreCommand.Stderr = &cmd.Stderr + return cmd +} + +// MustReadFile reads data from a file. Panic on error. +func MustReadFile(filename string) []byte { + b, err := ioutil.ReadFile(filename) + if err != nil { + panic(err.Error()) + } + return b +} +*/ diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/command.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/command.go new file mode 100644 index 000000000..629fd3eec --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/command.go @@ -0,0 +1,235 @@ +package run + +import ( + "flag" + "fmt" + "io" + "io/ioutil" + "log" + "os" + "path/filepath" + "runtime" + "strconv" + "strings" + "time" + + "github.com/BurntSushi/toml" +) + +const logo = ` + 8888888 .d888 888 8888888b. 888888b. + 888 d88P" 888 888 "Y88b 888 "88b + 888 888 888 888 888 888 .88P + 888 88888b. 888888 888 888 888 888 888 888 888 8888888K. + 888 888 "88b 888 888 888 888 Y8bd8P' 888 888 888 "Y88b + 888 888 888 888 888 888 888 X88K 888 888 888 888 + 888 888 888 888 888 Y88b 888 .d8""8b. 888 .d88P 888 d88P + 8888888 888 888 888 888 "Y88888 888 888 8888888P" 8888888P" + +` + +// Command represents the command executed by "influxd run". +type Command struct { + Version string + Branch string + Commit string + + closing chan struct{} + Closed chan struct{} + + Stdin io.Reader + Stdout io.Writer + Stderr io.Writer + + Server *Server +} + +// NewCommand return a new instance of Command. +func NewCommand() *Command { + return &Command{ + closing: make(chan struct{}), + Closed: make(chan struct{}), + Stdin: os.Stdin, + Stdout: os.Stdout, + Stderr: os.Stderr, + } +} + +// Run parses the config from args and runs the server. +func (cmd *Command) Run(args ...string) error { + // Parse the command line flags. + options, err := cmd.ParseFlags(args...) + if err != nil { + return err + } + + // Print sweet InfluxDB logo. + fmt.Print(logo) + + // Write the PID file. + if err := cmd.writePIDFile(options.PIDFile); err != nil { + return fmt.Errorf("write pid file: %s", err) + } + + // Set parallelism. + runtime.GOMAXPROCS(runtime.NumCPU()) + + // Turn on block profiling to debug stuck databases + runtime.SetBlockProfileRate(int(1 * time.Second)) + + // Parse config + config, err := cmd.ParseConfig(options.ConfigPath) + if err != nil { + return fmt.Errorf("parse config: %s", err) + } + + // Apply any environment variables on top of the parsed config + if err := config.ApplyEnvOverrides(); err != nil { + return fmt.Errorf("apply env config: %v", err) + } + + // Override config hostname if specified in the command line args. + if options.Hostname != "" { + config.Meta.Hostname = options.Hostname + } + + if options.Join != "" { + config.Meta.Peers = strings.Split(options.Join, ",") + } + + // Validate the configuration. + if err := config.Validate(); err != nil { + return fmt.Errorf("%s. To generate a valid configuration file run `influxd config > influxdb.generated.conf`.", err) + } + + // Create server from config and start it. + s, err := NewServer(config, cmd.Version) + if err != nil { + return fmt.Errorf("create server: %s", err) + } + s.CPUProfile = options.CPUProfile + s.MemProfile = options.MemProfile + if err := s.Open(); err != nil { + return fmt.Errorf("open server: %s", err) + } + cmd.Server = s + + // Mark start-up in log. + log.Printf("InfluxDB starting, version %s, branch %s, commit %s", cmd.Version, cmd.Branch, cmd.Commit) + log.Println("GOMAXPROCS set to", runtime.GOMAXPROCS(0)) + + // Begin monitoring the server's error channel. + go cmd.monitorServerErrors() + + return nil +} + +// Close shuts down the server. +func (cmd *Command) Close() error { + defer close(cmd.Closed) + close(cmd.closing) + if cmd.Server != nil { + return cmd.Server.Close() + } + return nil +} + +func (cmd *Command) monitorServerErrors() { + logger := log.New(cmd.Stderr, "", log.LstdFlags) + for { + select { + case err := <-cmd.Server.Err(): + logger.Println(err) + case <-cmd.closing: + return + } + } +} + +// ParseFlags parses the command line flags from args and returns an options set. +func (cmd *Command) ParseFlags(args ...string) (Options, error) { + var options Options + fs := flag.NewFlagSet("", flag.ContinueOnError) + fs.StringVar(&options.ConfigPath, "config", "", "") + fs.StringVar(&options.PIDFile, "pidfile", "", "") + fs.StringVar(&options.Hostname, "hostname", "", "") + fs.StringVar(&options.Join, "join", "", "") + fs.StringVar(&options.CPUProfile, "cpuprofile", "", "") + fs.StringVar(&options.MemProfile, "memprofile", "", "") + fs.Usage = func() { fmt.Fprintln(cmd.Stderr, usage) } + if err := fs.Parse(args); err != nil { + return Options{}, err + } + return options, nil +} + +// writePIDFile writes the process ID to path. +func (cmd *Command) writePIDFile(path string) error { + // Ignore if path is not set. + if path == "" { + return nil + } + + // Ensure the required directory structure exists. + err := os.MkdirAll(filepath.Dir(path), 0777) + if err != nil { + return fmt.Errorf("mkdir: %s", err) + } + + // Retrieve the PID and write it. + pid := strconv.Itoa(os.Getpid()) + if err := ioutil.WriteFile(path, []byte(pid), 0666); err != nil { + return fmt.Errorf("write file: %s", err) + } + + return nil +} + +// ParseConfig parses the config at path. +// Returns a demo configuration if path is blank. +func (cmd *Command) ParseConfig(path string) (*Config, error) { + // Use demo configuration if no config path is specified. + if path == "" { + fmt.Fprintln(cmd.Stdout, "no configuration provided, using default settings") + return NewDemoConfig() + } + + fmt.Fprintf(cmd.Stdout, "Using configuration at: %s\n", path) + + config := NewConfig() + if _, err := toml.DecodeFile(path, &config); err != nil { + return nil, err + } + + return config, nil +} + +var usage = `usage: run [flags] + +run starts the broker and data node server. If this is the first time running +the command then a new cluster will be initialized unless the -join argument +is used. + + -config + Set the path to the configuration file. + + -hostname + Override the hostname, the 'hostname' configuration + option will be overridden. + + -join + Joins the server to an existing cluster. + + -pidfile + Write process ID to a file. +` + +// Options represents the command line options that can be parsed. +type Options struct { + ConfigPath string + PIDFile string + Hostname string + Join string + CPUProfile string + MemProfile string +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config.go new file mode 100644 index 000000000..77515ca60 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config.go @@ -0,0 +1,227 @@ +package run + +import ( + "errors" + "fmt" + "os" + "os/user" + "path/filepath" + "reflect" + "strconv" + "strings" + "time" + + "github.com/influxdb/influxdb/cluster" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/services/admin" + "github.com/influxdb/influxdb/services/collectd" + "github.com/influxdb/influxdb/services/continuous_querier" + "github.com/influxdb/influxdb/services/graphite" + "github.com/influxdb/influxdb/services/hh" + "github.com/influxdb/influxdb/services/httpd" + "github.com/influxdb/influxdb/services/monitor" + "github.com/influxdb/influxdb/services/opentsdb" + "github.com/influxdb/influxdb/services/precreator" + "github.com/influxdb/influxdb/services/retention" + "github.com/influxdb/influxdb/services/udp" + "github.com/influxdb/influxdb/tsdb" +) + +// Config represents the configuration format for the influxd binary. +type Config struct { + Meta *meta.Config `toml:"meta"` + Data tsdb.Config `toml:"data"` + Cluster cluster.Config `toml:"cluster"` + Retention retention.Config `toml:"retention"` + Precreator precreator.Config `toml:"shard-precreation"` + + Admin admin.Config `toml:"admin"` + HTTPD httpd.Config `toml:"http"` + Graphites []graphite.Config `toml:"graphite"` + Collectd collectd.Config `toml:"collectd"` + OpenTSDB opentsdb.Config `toml:"opentsdb"` + UDPs []udp.Config `toml:"udp"` + + // Snapshot SnapshotConfig `toml:"snapshot"` + Monitoring monitor.Config `toml:"monitoring"` + ContinuousQuery continuous_querier.Config `toml:"continuous_queries"` + + HintedHandoff hh.Config `toml:"hinted-handoff"` + + // Server reporting + ReportingDisabled bool `toml:"reporting-disabled"` +} + +// NewConfig returns an instance of Config with reasonable defaults. +func NewConfig() *Config { + c := &Config{} + c.Meta = meta.NewConfig() + c.Data = tsdb.NewConfig() + c.Cluster = cluster.NewConfig() + c.Precreator = precreator.NewConfig() + + c.Admin = admin.NewConfig() + c.HTTPD = httpd.NewConfig() + c.Collectd = collectd.NewConfig() + c.OpenTSDB = opentsdb.NewConfig() + c.Graphites = append(c.Graphites, graphite.NewConfig()) + + c.Monitoring = monitor.NewConfig() + c.ContinuousQuery = continuous_querier.NewConfig() + c.Retention = retention.NewConfig() + c.HintedHandoff = hh.NewConfig() + + return c +} + +// NewDemoConfig returns the config that runs when no config is specified. +func NewDemoConfig() (*Config, error) { + c := NewConfig() + + var homeDir string + // By default, store meta and data files in current users home directory + u, err := user.Current() + if err == nil { + homeDir = u.HomeDir + } else if os.Getenv("HOME") != "" { + homeDir = os.Getenv("HOME") + } else { + return nil, fmt.Errorf("failed to determine current user for storage") + } + + c.Meta.Dir = filepath.Join(homeDir, ".influxdb/meta") + c.Data.Dir = filepath.Join(homeDir, ".influxdb/data") + c.HintedHandoff.Dir = filepath.Join(homeDir, ".influxdb/hh") + c.Data.WALDir = filepath.Join(homeDir, ".influxdb/wal") + + c.Admin.Enabled = true + c.Monitoring.Enabled = false + + return c, nil +} + +// Validate returns an error if the config is invalid. +func (c *Config) Validate() error { + if c.Meta.Dir == "" { + return errors.New("Meta.Dir must be specified") + } else if c.Data.Dir == "" { + return errors.New("Data.Dir must be specified") + } else if c.HintedHandoff.Dir == "" { + return errors.New("HintedHandoff.Dir must be specified") + } else if c.Data.WALDir == "" { + return errors.New("Data.WALDir must be specified") + } + + for _, g := range c.Graphites { + if err := g.Validate(); err != nil { + return fmt.Errorf("invalid graphite config: %v", err) + } + } + return nil +} + +func (c *Config) ApplyEnvOverrides() error { + return c.applyEnvOverrides("INFLUXDB", reflect.ValueOf(c)) +} + +func (c *Config) applyEnvOverrides(prefix string, spec reflect.Value) error { + // If we have a pointer, dereference it + s := spec + if spec.Kind() == reflect.Ptr { + s = spec.Elem() + } + + // Make sure we have struct + if s.Kind() != reflect.Struct { + return nil + } + + typeOfSpec := s.Type() + for i := 0; i < s.NumField(); i++ { + f := s.Field(i) + // Get the toml tag to determine what env var name to use + configName := typeOfSpec.Field(i).Tag.Get("toml") + // Replace hyphens with underscores to avoid issues with shells + configName = strings.Replace(configName, "-", "_", -1) + fieldName := typeOfSpec.Field(i).Name + + // Skip any fields that we cannot set + if f.CanSet() || f.Kind() == reflect.Slice { + + // Use the upper-case prefix and toml name for the env var + key := strings.ToUpper(configName) + if prefix != "" { + key = strings.ToUpper(fmt.Sprintf("%s_%s", prefix, configName)) + } + value := os.Getenv(key) + + // If the type is s slice, apply to each using the index as a suffix + // e.g. GRAPHITE_0 + if f.Kind() == reflect.Slice || f.Kind() == reflect.Array { + for i := 0; i < f.Len(); i++ { + if err := c.applyEnvOverrides(fmt.Sprintf("%s_%d", key, i), f.Index(i)); err != nil { + return err + } + } + continue + } + + // If it's a sub-config, recursively apply + if f.Kind() == reflect.Struct || f.Kind() == reflect.Ptr { + if err := c.applyEnvOverrides(key, f); err != nil { + return err + } + continue + } + + // Skip any fields we don't have a value to set + if value == "" { + continue + } + + switch f.Kind() { + case reflect.String: + f.SetString(value) + case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64: + + var intValue int64 + + // Handle toml.Duration + if f.Type().Name() == "Duration" { + dur, err := time.ParseDuration(value) + if err != nil { + return fmt.Errorf("failed to apply %v to %v using type %v and value '%v'", key, fieldName, f.Type().String(), value) + } + intValue = dur.Nanoseconds() + } else { + var err error + intValue, err = strconv.ParseInt(value, 0, f.Type().Bits()) + if err != nil { + return fmt.Errorf("failed to apply %v to %v using type %v and value '%v'", key, fieldName, f.Type().String(), value) + } + } + + f.SetInt(intValue) + case reflect.Bool: + boolValue, err := strconv.ParseBool(value) + if err != nil { + return fmt.Errorf("failed to apply %v to %v using type %v and value '%v'", key, fieldName, f.Type().String(), value) + + } + f.SetBool(boolValue) + case reflect.Float32, reflect.Float64: + floatValue, err := strconv.ParseFloat(value, f.Type().Bits()) + if err != nil { + return fmt.Errorf("failed to apply %v to %v using type %v and value '%v'", key, fieldName, f.Type().String(), value) + + } + f.SetFloat(floatValue) + default: + if err := c.applyEnvOverrides(key, f); err != nil { + return err + } + } + } + } + return nil +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config_command.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config_command.go new file mode 100644 index 000000000..8135aa014 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config_command.go @@ -0,0 +1,73 @@ +package run + +import ( + "flag" + "fmt" + "io" + "os" + + "github.com/BurntSushi/toml" +) + +// PrintConfigCommand represents the command executed by "influxd config". +type PrintConfigCommand struct { + Stdin io.Reader + Stdout io.Writer + Stderr io.Writer +} + +// NewPrintConfigCommand return a new instance of PrintConfigCommand. +func NewPrintConfigCommand() *PrintConfigCommand { + return &PrintConfigCommand{ + Stdin: os.Stdin, + Stdout: os.Stdout, + Stderr: os.Stderr, + } +} + +// Run parses and prints the current config loaded. +func (cmd *PrintConfigCommand) Run(args ...string) error { + // Parse command flags. + fs := flag.NewFlagSet("", flag.ContinueOnError) + configPath := fs.String("config", "", "") + hostname := fs.String("hostname", "", "") + fs.Usage = func() { fmt.Fprintln(cmd.Stderr, printConfigUsage) } + if err := fs.Parse(args); err != nil { + return err + } + + // Parse config from path. + config, err := cmd.parseConfig(*configPath) + if err != nil { + return fmt.Errorf("parse config: %s", err) + } + + // Override config properties. + if *hostname != "" { + config.Meta.Hostname = *hostname + } + + toml.NewEncoder(cmd.Stdout).Encode(config) + fmt.Fprint(cmd.Stdout, "\n") + + return nil +} + +// ParseConfig parses the config at path. +// Returns a demo configuration if path is blank. +func (cmd *PrintConfigCommand) parseConfig(path string) (*Config, error) { + if path == "" { + return NewDemoConfig() + } + + config := NewConfig() + if _, err := toml.DecodeFile(path, &config); err != nil { + return nil, err + } + return config, nil +} + +var printConfigUsage = `usage: config + + config displays the default configuration +` diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config_test.go new file mode 100644 index 000000000..ffcaf2533 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/config_test.go @@ -0,0 +1,144 @@ +package run_test + +import ( + "os" + "testing" + + "github.com/BurntSushi/toml" + "github.com/influxdb/influxdb/cmd/influxd/run" +) + +// Ensure the configuration can be parsed. +func TestConfig_Parse(t *testing.T) { + // Parse configuration. + var c run.Config + if _, err := toml.Decode(` +[meta] +dir = "/tmp/meta" + +[data] +dir = "/tmp/data" + +[cluster] + +[admin] +bind-address = ":8083" + +[http] +bind-address = ":8087" + +[[graphite]] +protocol = "udp" + +[[graphite]] +protocol = "tcp" + +[collectd] +bind-address = ":1000" + +[opentsdb] +bind-address = ":2000" + +[[udp]] +bind-address = ":4444" + +[monitoring] +enabled = true + +[continuous_queries] +enabled = true +`, &c); err != nil { + t.Fatal(err) + } + + // Validate configuration. + if c.Meta.Dir != "/tmp/meta" { + t.Fatalf("unexpected meta dir: %s", c.Meta.Dir) + } else if c.Data.Dir != "/tmp/data" { + t.Fatalf("unexpected data dir: %s", c.Data.Dir) + } else if c.Admin.BindAddress != ":8083" { + t.Fatalf("unexpected admin bind address: %s", c.Admin.BindAddress) + } else if c.HTTPD.BindAddress != ":8087" { + t.Fatalf("unexpected api bind address: %s", c.HTTPD.BindAddress) + } else if len(c.Graphites) != 2 { + t.Fatalf("unexpected graphites count: %d", len(c.Graphites)) + } else if c.Graphites[0].Protocol != "udp" { + t.Fatalf("unexpected graphite protocol(0): %s", c.Graphites[0].Protocol) + } else if c.Graphites[1].Protocol != "tcp" { + t.Fatalf("unexpected graphite protocol(1): %s", c.Graphites[1].Protocol) + } else if c.Collectd.BindAddress != ":1000" { + t.Fatalf("unexpected collectd bind address: %s", c.Collectd.BindAddress) + } else if c.OpenTSDB.BindAddress != ":2000" { + t.Fatalf("unexpected opentsdb bind address: %s", c.OpenTSDB.BindAddress) + } else if c.UDPs[0].BindAddress != ":4444" { + t.Fatalf("unexpected udp bind address: %s", c.UDPs[0].BindAddress) + } else if c.Monitoring.Enabled != true { + t.Fatalf("unexpected monitoring enabled: %v", c.Monitoring.Enabled) + } else if c.ContinuousQuery.Enabled != true { + t.Fatalf("unexpected continuous query enabled: %v", c.ContinuousQuery.Enabled) + } +} + +// Ensure the configuration can be parsed. +func TestConfig_Parse_EnvOverride(t *testing.T) { + // Parse configuration. + var c run.Config + if _, err := toml.Decode(` +[meta] +dir = "/tmp/meta" + +[data] +dir = "/tmp/data" + +[cluster] + +[admin] +bind-address = ":8083" + +[http] +bind-address = ":8087" + +[[graphite]] +protocol = "udp" + +[[graphite]] +protocol = "tcp" + +[collectd] +bind-address = ":1000" + +[opentsdb] +bind-address = ":2000" + +[[udp]] +bind-address = ":4444" + +[monitoring] +enabled = true + +[continuous_queries] +enabled = true +`, &c); err != nil { + t.Fatal(err) + } + + if err := os.Setenv("INFLUXDB_UDP_BIND_ADDRESS", ":1234"); err != nil { + t.Fatalf("failed to set env var: %v", err) + } + + if err := os.Setenv("INFLUXDB_GRAPHITE_1_PROTOCOL", "udp"); err != nil { + t.Fatalf("failed to set env var: %v", err) + } + + if err := c.ApplyEnvOverrides(); err != nil { + t.Fatalf("failed to apply env overrides: %v", err) + } + + if c.UDPs[0].BindAddress != ":4444" { + t.Fatalf("unexpected udp bind address: %s", c.UDPs[0].BindAddress) + } + + if c.Graphites[1].Protocol != "udp" { + t.Fatalf("unexpected graphite protocol(0): %s", c.Graphites[0].Protocol) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server.go new file mode 100644 index 000000000..ab65817d7 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server.go @@ -0,0 +1,536 @@ +package run + +import ( + "bytes" + "fmt" + "log" + "net" + "net/http" + "os" + "runtime" + "runtime/pprof" + "strings" + "time" + + "github.com/influxdb/influxdb/cluster" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/services/admin" + "github.com/influxdb/influxdb/services/collectd" + "github.com/influxdb/influxdb/services/continuous_querier" + "github.com/influxdb/influxdb/services/graphite" + "github.com/influxdb/influxdb/services/hh" + "github.com/influxdb/influxdb/services/httpd" + "github.com/influxdb/influxdb/services/opentsdb" + "github.com/influxdb/influxdb/services/precreator" + "github.com/influxdb/influxdb/services/retention" + "github.com/influxdb/influxdb/services/snapshotter" + "github.com/influxdb/influxdb/services/udp" + "github.com/influxdb/influxdb/tcp" + "github.com/influxdb/influxdb/tsdb" + _ "github.com/influxdb/influxdb/tsdb/engine" +) + +// Server represents a container for the metadata and storage data and services. +// It is built using a Config and it manages the startup and shutdown of all +// services in the proper order. +type Server struct { + version string // Build version + + err chan error + closing chan struct{} + + Hostname string + BindAddress string + Listener net.Listener + + MetaStore *meta.Store + TSDBStore *tsdb.Store + QueryExecutor *tsdb.QueryExecutor + PointsWriter *cluster.PointsWriter + ShardWriter *cluster.ShardWriter + ShardMapper *cluster.ShardMapper + HintedHandoff *hh.Service + + Services []Service + + // These references are required for the tcp muxer. + ClusterService *cluster.Service + SnapshotterService *snapshotter.Service + + // Server reporting + reportingDisabled bool + + // Profiling + CPUProfile string + MemProfile string +} + +// NewServer returns a new instance of Server built from a config. +func NewServer(c *Config, version string) (*Server, error) { + // Construct base meta store and data store. + tsdbStore := tsdb.NewStore(c.Data.Dir) + tsdbStore.EngineOptions.Config = c.Data + + s := &Server{ + version: version, + err: make(chan error), + closing: make(chan struct{}), + + Hostname: c.Meta.Hostname, + BindAddress: c.Meta.BindAddress, + + MetaStore: meta.NewStore(c.Meta), + TSDBStore: tsdbStore, + + reportingDisabled: c.ReportingDisabled, + } + + // Copy TSDB configuration. + s.TSDBStore.EngineOptions.MaxWALSize = c.Data.MaxWALSize + s.TSDBStore.EngineOptions.WALFlushInterval = time.Duration(c.Data.WALFlushInterval) + s.TSDBStore.EngineOptions.WALPartitionFlushDelay = time.Duration(c.Data.WALPartitionFlushDelay) + + // Set the shard mapper + s.ShardMapper = cluster.NewShardMapper(time.Duration(c.Cluster.ShardMapperTimeout)) + s.ShardMapper.ForceRemoteMapping = c.Cluster.ForceRemoteShardMapping + s.ShardMapper.MetaStore = s.MetaStore + s.ShardMapper.TSDBStore = s.TSDBStore + + // Initialize query executor. + s.QueryExecutor = tsdb.NewQueryExecutor(s.TSDBStore) + s.QueryExecutor.MetaStore = s.MetaStore + s.QueryExecutor.MetaStatementExecutor = &meta.StatementExecutor{Store: s.MetaStore} + s.QueryExecutor.ShardMapper = s.ShardMapper + + // Set the shard writer + s.ShardWriter = cluster.NewShardWriter(time.Duration(c.Cluster.ShardWriterTimeout)) + s.ShardWriter.MetaStore = s.MetaStore + + // Create the hinted handoff service + s.HintedHandoff = hh.NewService(c.HintedHandoff, s.ShardWriter) + + // Initialize points writer. + s.PointsWriter = cluster.NewPointsWriter() + s.PointsWriter.WriteTimeout = time.Duration(c.Cluster.WriteTimeout) + s.PointsWriter.MetaStore = s.MetaStore + s.PointsWriter.TSDBStore = s.TSDBStore + s.PointsWriter.ShardWriter = s.ShardWriter + s.PointsWriter.HintedHandoff = s.HintedHandoff + + // Append services. + s.appendClusterService(c.Cluster) + s.appendPrecreatorService(c.Precreator) + s.appendSnapshotterService() + s.appendAdminService(c.Admin) + s.appendContinuousQueryService(c.ContinuousQuery) + s.appendHTTPDService(c.HTTPD) + s.appendCollectdService(c.Collectd) + if err := s.appendOpenTSDBService(c.OpenTSDB); err != nil { + return nil, err + } + for _, g := range c.UDPs { + s.appendUDPService(g) + } + s.appendRetentionPolicyService(c.Retention) + for _, g := range c.Graphites { + if err := s.appendGraphiteService(g); err != nil { + return nil, err + } + } + + return s, nil +} + +func (s *Server) appendClusterService(c cluster.Config) { + srv := cluster.NewService(c) + srv.TSDBStore = s.TSDBStore + srv.MetaStore = s.MetaStore + s.Services = append(s.Services, srv) + s.ClusterService = srv +} + +func (s *Server) appendSnapshotterService() { + srv := snapshotter.NewService() + srv.TSDBStore = s.TSDBStore + srv.MetaStore = s.MetaStore + s.Services = append(s.Services, srv) + s.SnapshotterService = srv +} + +func (s *Server) appendRetentionPolicyService(c retention.Config) { + if !c.Enabled { + return + } + srv := retention.NewService(c) + srv.MetaStore = s.MetaStore + srv.TSDBStore = s.TSDBStore + s.Services = append(s.Services, srv) +} + +func (s *Server) appendAdminService(c admin.Config) { + if !c.Enabled { + return + } + srv := admin.NewService(c) + s.Services = append(s.Services, srv) +} + +func (s *Server) appendHTTPDService(c httpd.Config) { + if !c.Enabled { + return + } + srv := httpd.NewService(c) + srv.Handler.MetaStore = s.MetaStore + srv.Handler.QueryExecutor = s.QueryExecutor + srv.Handler.PointsWriter = s.PointsWriter + srv.Handler.Version = s.version + + // If a ContinuousQuerier service has been started, attach it. + for _, srvc := range s.Services { + if cqsrvc, ok := srvc.(continuous_querier.ContinuousQuerier); ok { + srv.Handler.ContinuousQuerier = cqsrvc + } + } + + s.Services = append(s.Services, srv) +} + +func (s *Server) appendCollectdService(c collectd.Config) { + if !c.Enabled { + return + } + srv := collectd.NewService(c) + srv.MetaStore = s.MetaStore + srv.PointsWriter = s.PointsWriter + s.Services = append(s.Services, srv) +} + +func (s *Server) appendOpenTSDBService(c opentsdb.Config) error { + if !c.Enabled { + return nil + } + srv, err := opentsdb.NewService(c) + if err != nil { + return err + } + srv.PointsWriter = s.PointsWriter + srv.MetaStore = s.MetaStore + s.Services = append(s.Services, srv) + return nil +} + +func (s *Server) appendGraphiteService(c graphite.Config) error { + if !c.Enabled { + return nil + } + srv, err := graphite.NewService(c) + if err != nil { + return err + } + + srv.PointsWriter = s.PointsWriter + srv.MetaStore = s.MetaStore + s.Services = append(s.Services, srv) + return nil +} + +func (s *Server) appendPrecreatorService(c precreator.Config) error { + if !c.Enabled { + return nil + } + srv, err := precreator.NewService(c) + if err != nil { + return err + } + + srv.MetaStore = s.MetaStore + s.Services = append(s.Services, srv) + return nil +} + +func (s *Server) appendUDPService(c udp.Config) { + if !c.Enabled { + return + } + srv := udp.NewService(c) + srv.PointsWriter = s.PointsWriter + s.Services = append(s.Services, srv) +} + +func (s *Server) appendContinuousQueryService(c continuous_querier.Config) { + if !c.Enabled { + return + } + srv := continuous_querier.NewService(c) + srv.MetaStore = s.MetaStore + srv.QueryExecutor = s.QueryExecutor + srv.PointsWriter = s.PointsWriter + s.Services = append(s.Services, srv) +} + +// Err returns an error channel that multiplexes all out of band errors received from all services. +func (s *Server) Err() <-chan error { return s.err } + +// Open opens the meta and data store and all services. +func (s *Server) Open() error { + if err := func() error { + // Start profiling, if set. + startProfile(s.CPUProfile, s.MemProfile) + + host, port, err := s.hostAddr() + if err != nil { + return err + } + + hostport := net.JoinHostPort(host, port) + addr, err := net.ResolveTCPAddr("tcp", hostport) + if err != nil { + return fmt.Errorf("resolve tcp: addr=%s, err=%s", hostport, err) + } + s.MetaStore.Addr = addr + s.MetaStore.RemoteAddr = &tcpaddr{hostport} + + // Open shared TCP connection. + ln, err := net.Listen("tcp", s.BindAddress) + if err != nil { + return fmt.Errorf("listen: %s", err) + } + s.Listener = ln + + // The port 0 is used, we need to retrieve the port assigned by the kernel + if strings.HasSuffix(s.BindAddress, ":0") { + s.MetaStore.Addr = ln.Addr() + } + + // Multiplex listener. + mux := tcp.NewMux() + s.MetaStore.RaftListener = mux.Listen(meta.MuxRaftHeader) + s.MetaStore.ExecListener = mux.Listen(meta.MuxExecHeader) + s.MetaStore.RPCListener = mux.Listen(meta.MuxRPCHeader) + + s.ClusterService.Listener = mux.Listen(cluster.MuxHeader) + s.SnapshotterService.Listener = mux.Listen(snapshotter.MuxHeader) + go mux.Serve(ln) + + // Open meta store. + if err := s.MetaStore.Open(); err != nil { + return fmt.Errorf("open meta store: %s", err) + } + go s.monitorErrorChan(s.MetaStore.Err()) + + // Wait for the store to initialize. + <-s.MetaStore.Ready() + + // Open TSDB store. + if err := s.TSDBStore.Open(); err != nil { + return fmt.Errorf("open tsdb store: %s", err) + } + + // Open the hinted handoff service + if err := s.HintedHandoff.Open(); err != nil { + return fmt.Errorf("open hinted handoff: %s", err) + } + + for _, service := range s.Services { + if err := service.Open(); err != nil { + return fmt.Errorf("open service: %s", err) + } + } + + // Start the reporting service, if not disabled. + if !s.reportingDisabled { + go s.startServerReporting() + } + + return nil + + }(); err != nil { + s.Close() + return err + } + + return nil +} + +// Close shuts down the meta and data stores and all services. +func (s *Server) Close() error { + stopProfile() + + if s.Listener != nil { + s.Listener.Close() + } + if s.MetaStore != nil { + s.MetaStore.Close() + } + if s.TSDBStore != nil { + s.TSDBStore.Close() + } + if s.HintedHandoff != nil { + s.HintedHandoff.Close() + } + for _, service := range s.Services { + service.Close() + } + + close(s.closing) + return nil +} + +// startServerReporting starts periodic server reporting. +func (s *Server) startServerReporting() { + for { + select { + case <-s.closing: + return + default: + } + if err := s.MetaStore.WaitForLeader(30 * time.Second); err != nil { + log.Printf("no leader available for reporting: %s", err.Error()) + time.Sleep(time.Second) + continue + } + s.reportServer() + <-time.After(24 * time.Hour) + } +} + +// reportServer reports anonymous statistics about the system. +func (s *Server) reportServer() { + dis, err := s.MetaStore.Databases() + if err != nil { + log.Printf("failed to retrieve databases for reporting: %s", err.Error()) + return + } + numDatabases := len(dis) + + numMeasurements := 0 + numSeries := 0 + for _, di := range dis { + d := s.TSDBStore.DatabaseIndex(di.Name) + if d == nil { + // No data in this store for this database. + continue + } + m, s := d.MeasurementSeriesCounts() + numMeasurements += m + numSeries += s + } + + clusterID, err := s.MetaStore.ClusterID() + if err != nil { + log.Printf("failed to retrieve cluster ID for reporting: %s", err.Error()) + return + } + + json := fmt.Sprintf(`[{ + "name":"reports", + "columns":["os", "arch", "version", "server_id", "cluster_id", "num_series", "num_measurements", "num_databases"], + "points":[["%s", "%s", "%s", "%x", "%x", "%d", "%d", "%d"]] + }]`, runtime.GOOS, runtime.GOARCH, s.version, s.MetaStore.NodeID(), clusterID, numSeries, numMeasurements, numDatabases) + + data := bytes.NewBufferString(json) + + log.Printf("Sending anonymous usage statistics to m.influxdb.com") + + client := http.Client{Timeout: time.Duration(5 * time.Second)} + go client.Post("http://m.influxdb.com:8086/db/reporting/series?u=reporter&p=influxdb", "application/json", data) +} + +// monitorErrorChan reads an error channel and resends it through the server. +func (s *Server) monitorErrorChan(ch <-chan error) { + for { + select { + case err, ok := <-ch: + if !ok { + return + } + s.err <- err + case <-s.closing: + return + } + } +} + +// hostAddr returns the host and port that remote nodes will use to reach this +// node. +func (s *Server) hostAddr() (string, string, error) { + // Resolve host to address. + _, port, err := net.SplitHostPort(s.BindAddress) + if err != nil { + return "", "", fmt.Errorf("split bind address: %s", err) + } + + host := s.Hostname + + // See if we might have a port that will override the BindAddress port + if host != "" && host[len(host)-1] >= '0' && host[len(host)-1] <= '9' && strings.Contains(host, ":") { + hostArg, portArg, err := net.SplitHostPort(s.Hostname) + if err != nil { + return "", "", err + } + + if hostArg != "" { + host = hostArg + } + + if portArg != "" { + port = portArg + } + } + return host, port, nil +} + +// Service represents a service attached to the server. +type Service interface { + Open() error + Close() error +} + +// prof stores the file locations of active profiles. +var prof struct { + cpu *os.File + mem *os.File +} + +// StartProfile initializes the cpu and memory profile, if specified. +func startProfile(cpuprofile, memprofile string) { + if cpuprofile != "" { + f, err := os.Create(cpuprofile) + if err != nil { + log.Fatalf("cpuprofile: %v", err) + } + log.Printf("writing CPU profile to: %s\n", cpuprofile) + prof.cpu = f + pprof.StartCPUProfile(prof.cpu) + } + + if memprofile != "" { + f, err := os.Create(memprofile) + if err != nil { + log.Fatalf("memprofile: %v", err) + } + log.Printf("writing mem profile to: %s\n", memprofile) + prof.mem = f + runtime.MemProfileRate = 4096 + } + +} + +// StopProfile closes the cpu and memory profiles if they are running. +func stopProfile() { + if prof.cpu != nil { + pprof.StopCPUProfile() + prof.cpu.Close() + log.Println("CPU profile stopped") + } + if prof.mem != nil { + pprof.Lookup("heap").WriteTo(prof.mem, 0) + prof.mem.Close() + log.Println("mem profile stopped") + } +} + +type tcpaddr struct{ host string } + +func (a *tcpaddr) Network() string { return "tcp" } +func (a *tcpaddr) String() string { return a.host } diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_helpers_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_helpers_test.go new file mode 100644 index 000000000..adabdc19e --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_helpers_test.go @@ -0,0 +1,312 @@ +// This package is a set of convenience helpers and structs to make integration testing easier +package run_test + +import ( + "encoding/json" + "fmt" + "io" + "io/ioutil" + "log" + "math" + "net/http" + "net/url" + "os" + "regexp" + "strings" + "testing" + "time" + + "github.com/influxdb/influxdb/cmd/influxd/run" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/services/httpd" + "github.com/influxdb/influxdb/toml" +) + +// Server represents a test wrapper for run.Server. +type Server struct { + *run.Server + Config *run.Config +} + +// NewServer returns a new instance of Server. +func NewServer(c *run.Config) *Server { + srv, _ := run.NewServer(c, "testServer") + s := Server{ + Server: srv, + Config: c, + } + s.TSDBStore.EngineOptions.Config = c.Data + configureLogging(&s) + return &s +} + +// OpenServer opens a test server. +func OpenServer(c *run.Config, joinURLs string) *Server { + s := NewServer(c) + configureLogging(s) + if err := s.Open(); err != nil { + panic(err.Error()) + } + + return s +} + +// OpenServerWithVersion opens a test server with a specific version. +func OpenServerWithVersion(c *run.Config, version string) *Server { + srv, _ := run.NewServer(c, version) + s := Server{ + Server: srv, + Config: c, + } + configureLogging(&s) + if err := s.Open(); err != nil { + panic(err.Error()) + } + + return &s +} + +// Close shuts down the server and removes all temporary paths. +func (s *Server) Close() { + os.RemoveAll(s.Config.Meta.Dir) + os.RemoveAll(s.Config.Data.Dir) + os.RemoveAll(s.Config.HintedHandoff.Dir) + s.Server.Close() +} + +// URL returns the base URL for the httpd endpoint. +func (s *Server) URL() string { + for _, service := range s.Services { + if service, ok := service.(*httpd.Service); ok { + return "http://" + service.Addr().String() + } + } + panic("httpd server not found in services") +} + +// CreateDatabaseAndRetentionPolicy will create the database and retention policy. +func (s *Server) CreateDatabaseAndRetentionPolicy(db string, rp *meta.RetentionPolicyInfo) error { + if _, err := s.MetaStore.CreateDatabase(db); err != nil { + return err + } else if _, err := s.MetaStore.CreateRetentionPolicy(db, rp); err != nil { + return err + } + return nil +} + +// Query executes a query against the server and returns the results. +func (s *Server) Query(query string) (results string, err error) { + return s.QueryWithParams(query, nil) +} + +// Query executes a query against the server and returns the results. +func (s *Server) QueryWithParams(query string, values url.Values) (results string, err error) { + if values == nil { + values = url.Values{} + } + values.Set("q", query) + resp, err := http.Get(s.URL() + "/query?" + values.Encode()) + if err != nil { + return "", err + //} else if resp.StatusCode != http.StatusOK && resp.StatusCode != http.StatusBadRequest { + } + body := string(MustReadAll(resp.Body)) + switch resp.StatusCode { + case http.StatusBadRequest: + if !expectPattern(".*error parsing query*.", body) { + return "", fmt.Errorf("unexpected status code: code=%d, body=%s", resp.StatusCode, body) + } + return body, nil + case http.StatusOK: + return body, nil + default: + return "", fmt.Errorf("unexpected status code: code=%d, body=%s", resp.StatusCode, body) + } +} + +// Write executes a write against the server and returns the results. +func (s *Server) Write(db, rp, body string, params url.Values) (results string, err error) { + if params == nil { + params = url.Values{} + } + if params.Get("db") == "" { + params.Set("db", db) + } + if params.Get("rp") == "" { + params.Set("rp", rp) + } + resp, err := http.Post(s.URL()+"/write?"+params.Encode(), "", strings.NewReader(body)) + if err != nil { + return "", err + } else if resp.StatusCode != http.StatusOK && resp.StatusCode != http.StatusNoContent { + return "", fmt.Errorf("invalid status code: code=%d, body=%s", resp.StatusCode, MustReadAll(resp.Body)) + } + return string(MustReadAll(resp.Body)), nil +} + +// NewConfig returns the default config with temporary paths. +func NewConfig() *run.Config { + c := run.NewConfig() + c.ReportingDisabled = true + c.Meta.Dir = MustTempFile() + c.Meta.BindAddress = "127.0.0.1:0" + c.Meta.HeartbeatTimeout = toml.Duration(50 * time.Millisecond) + c.Meta.ElectionTimeout = toml.Duration(50 * time.Millisecond) + c.Meta.LeaderLeaseTimeout = toml.Duration(50 * time.Millisecond) + c.Meta.CommitTimeout = toml.Duration(5 * time.Millisecond) + + c.Data.Dir = MustTempFile() + c.Data.WALDir = MustTempFile() + + c.HintedHandoff.Dir = MustTempFile() + + c.HTTPD.Enabled = true + c.HTTPD.BindAddress = "127.0.0.1:0" + c.HTTPD.LogEnabled = testing.Verbose() + + return c +} + +func newRetentionPolicyInfo(name string, rf int, duration time.Duration) *meta.RetentionPolicyInfo { + return &meta.RetentionPolicyInfo{Name: name, ReplicaN: rf, Duration: duration} +} + +func maxFloat64() string { + maxFloat64, _ := json.Marshal(math.MaxFloat64) + return string(maxFloat64) +} + +func maxInt64() string { + maxInt64, _ := json.Marshal(^int64(0)) + return string(maxInt64) +} + +func now() time.Time { + return time.Now().UTC() +} + +func yesterday() time.Time { + return now().Add(-1 * time.Hour * 24) +} + +func mustParseTime(layout, value string) time.Time { + tm, err := time.Parse(layout, value) + if err != nil { + panic(err) + } + return tm +} + +// MustReadAll reads r. Panic on error. +func MustReadAll(r io.Reader) []byte { + b, err := ioutil.ReadAll(r) + if err != nil { + panic(err) + } + return b +} + +// MustTempFile returns a path to a temporary file. +func MustTempFile() string { + f, err := ioutil.TempFile("", "influxd-") + if err != nil { + panic(err) + } + f.Close() + os.Remove(f.Name()) + return f.Name() +} + +func expectPattern(exp, act string) bool { + re := regexp.MustCompile(exp) + if !re.MatchString(act) { + return false + } + return true +} + +type Query struct { + name string + command string + params url.Values + exp, act string + pattern bool + skip bool +} + +// Execute runs the command and returns an err if it fails +func (q *Query) Execute(s *Server) (err error) { + if q.params == nil { + q.act, err = s.Query(q.command) + return + } + q.act, err = s.QueryWithParams(q.command, q.params) + return +} + +func (q *Query) success() bool { + if q.pattern { + return expectPattern(q.exp, q.act) + } + return q.exp == q.act +} + +func (q *Query) Error(err error) string { + return fmt.Sprintf("%s: %v", q.name, err) +} + +func (q *Query) failureMessage() string { + return fmt.Sprintf("%s: unexpected results\nquery: %s\nexp: %s\nactual: %s\n", q.name, q.command, q.exp, q.act) +} + +type Test struct { + initialized bool + write string + params url.Values + db string + rp string + exp string + queries []*Query +} + +func NewTest(db, rp string) Test { + return Test{ + db: db, + rp: rp, + } +} + +func (t *Test) addQueries(q ...*Query) { + t.queries = append(t.queries, q...) +} + +func (t *Test) init(s *Server) error { + if t.write == "" || t.initialized { + return nil + } + t.initialized = true + if res, err := s.Write(t.db, t.rp, t.write, t.params); err != nil { + return err + } else if t.exp != res { + return fmt.Errorf("unexpected results\nexp: %s\ngot: %s\n", t.exp, res) + } + return nil +} + +func configureLogging(s *Server) { + // Set the logger to discard unless verbose is on + if !testing.Verbose() { + type logSetter interface { + SetLogger(*log.Logger) + } + nullLogger := log.New(ioutil.Discard, "", 0) + s.MetaStore.Logger = nullLogger + s.TSDBStore.Logger = nullLogger + s.HintedHandoff.SetLogger(nullLogger) + for _, service := range s.Services { + if service, ok := service.(logSetter); ok { + service.SetLogger(nullLogger) + } + } + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_test.go new file mode 100644 index 000000000..10d67c7de --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_test.go @@ -0,0 +1,3719 @@ +package run_test + +import ( + "fmt" + "net/http" + "net/url" + "strconv" + "strings" + "testing" + "time" +) + +// Ensure that HTTP responses include the InfluxDB version. +func TestServer_HTTPResponseVersion(t *testing.T) { + version := "v1234" + s := OpenServerWithVersion(NewConfig(), version) + defer s.Close() + + resp, _ := http.Get(s.URL() + "/query") + got := resp.Header.Get("X-Influxdb-Version") + if got != version { + t.Errorf("Server responded with incorrect version, exp %s, got %s", version, got) + } +} + +// Ensure the database commands work. +func TestServer_DatabaseCommands(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + test := Test{ + queries: []*Query{ + &Query{ + name: "create database should succeed", + command: `CREATE DATABASE db0`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "create database should error with bad name", + command: `CREATE DATABASE 0xdb0`, + exp: `{"error":"error parsing query: found 0, expected identifier at line 1, char 17"}`, + }, + &Query{ + name: "show database should succeed", + command: `SHOW DATABASES`, + exp: `{"results":[{"series":[{"name":"databases","columns":["name"],"values":[["db0"]]}]}]}`, + }, + &Query{ + name: "create database should error if it already exists", + command: `CREATE DATABASE db0`, + exp: `{"results":[{"error":"database already exists"}]}`, + }, + &Query{ + name: "drop database should succeed", + command: `DROP DATABASE db0`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "show database should have no results", + command: `SHOW DATABASES`, + exp: `{"results":[{"series":[{"name":"databases","columns":["name"]}]}]}`, + }, + &Query{ + name: "drop database should error if it doesn't exist", + command: `DROP DATABASE db0`, + exp: `{"results":[{"error":"database not found: db0"}]}`, + }, + }, + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_DropAndRecreateDatabase(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=serverA,region=uswest val=23.2 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "Drop database after data write", + command: `DROP DATABASE db0`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "Recreate database", + command: `CREATE DATABASE db0`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "Recreate retention policy", + command: `CREATE RETENTION POLICY rp0 ON db0 DURATION 365d REPLICATION 1 DEFAULT`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "Show measurements after recreate", + command: `SHOW MEASUREMENTS`, + exp: `{"results":[{}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "Query data after recreate", + command: `SELECT * FROM cpu`, + exp: `{"results":[{}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_DropDatabaseIsolated(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + if err := s.CreateDatabaseAndRetentionPolicy("db1", newRetentionPolicyInfo("rp1", 1, 0)); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=serverA,region=uswest val=23.2 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "Query data from 1st database", + command: `SELECT * FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","host","region","val"],"values":[["2000-01-01T00:00:00Z","serverA","uswest",23.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "Query data from 1st database with GROUP BY *", + command: `SELECT * FROM cpu GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"serverA","region":"uswest"},"columns":["time","val"],"values":[["2000-01-01T00:00:00Z",23.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "Drop other database", + command: `DROP DATABASE db1`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "Query data from 1st database and ensure it's still there", + command: `SELECT * FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","host","region","val"],"values":[["2000-01-01T00:00:00Z","serverA","uswest",23.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "Query data from 1st database and ensure it's still there with GROUP BY *", + command: `SELECT * FROM cpu GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"serverA","region":"uswest"},"columns":["time","val"],"values":[["2000-01-01T00:00:00Z",23.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_DropAndRecreateSeries(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=serverA,region=uswest val=23.2 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "Show series is present", + command: `SHOW SERIES`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=serverA,region=uswest","serverA","uswest"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "Drop series after data write", + command: `DROP SERIES FROM cpu`, + exp: `{"results":[{}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "Show series is gone", + command: `SHOW SERIES`, + exp: `{"results":[{}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } + + // Re-write data and test again. + reTest := NewTest("db0", "rp0") + reTest.write = strings.Join(writes, "\n") + + reTest.addQueries([]*Query{ + &Query{ + name: "Show series is present again after re-write", + command: `SHOW SERIES`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=serverA,region=uswest","serverA","uswest"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range reTest.queries { + if i == 0 { + if err := reTest.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure retention policy commands work. +func TestServer_RetentionPolicyCommands(t *testing.T) { + t.Parallel() + c := NewConfig() + c.Meta.RetentionAutoCreate = false + s := OpenServer(c, "") + defer s.Close() + + // Create a database. + if _, err := s.MetaStore.CreateDatabase("db0"); err != nil { + t.Fatal(err) + } + + test := Test{ + queries: []*Query{ + &Query{ + name: "create retention policy should succeed", + command: `CREATE RETENTION POLICY rp0 ON db0 DURATION 1h REPLICATION 1`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "create retention policy should error if it already exists", + command: `CREATE RETENTION POLICY rp0 ON db0 DURATION 1h REPLICATION 1`, + exp: `{"results":[{"error":"retention policy already exists"}]}`, + }, + &Query{ + name: "show retention policy should succeed", + command: `SHOW RETENTION POLICIES ON db0`, + exp: `{"results":[{"series":[{"columns":["name","duration","replicaN","default"],"values":[["rp0","1h0m0s",1,false]]}]}]}`, + }, + &Query{ + name: "alter retention policy should succeed", + command: `ALTER RETENTION POLICY rp0 ON db0 DURATION 2h REPLICATION 3 DEFAULT`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "show retention policy should have new altered information", + command: `SHOW RETENTION POLICIES ON db0`, + exp: `{"results":[{"series":[{"columns":["name","duration","replicaN","default"],"values":[["rp0","2h0m0s",3,true]]}]}]}`, + }, + &Query{ + name: "drop retention policy should succeed", + command: `DROP RETENTION POLICY rp0 ON db0`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "show retention policy should be empty after dropping them", + command: `SHOW RETENTION POLICIES ON db0`, + exp: `{"results":[{"series":[{"columns":["name","duration","replicaN","default"]}]}]}`, + }, + &Query{ + name: "Ensure retention policy with unacceptable retention cannot be created", + command: `CREATE RETENTION POLICY rp3 ON db0 DURATION 1s REPLICATION 1`, + exp: `{"results":[{"error":"retention policy duration must be at least 1h0m0s"}]}`, + }, + &Query{ + name: "Check error when deleting retention policy on non-existent database", + command: `DROP RETENTION POLICY rp1 ON mydatabase`, + exp: `{"results":[{"error":"database not found"}]}`, + }, + }, + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the autocreation of retention policy works. +func TestServer_DatabaseRetentionPolicyAutoCreate(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + test := Test{ + queries: []*Query{ + &Query{ + name: "create database should succeed", + command: `CREATE DATABASE db0`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "show retention policies should return auto-created policy", + command: `SHOW RETENTION POLICIES ON db0`, + exp: `{"results":[{"series":[{"columns":["name","duration","replicaN","default"],"values":[["default","0",1,true]]}]}]}`, + }, + }, + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure user commands work. +func TestServer_UserCommands(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + // Create a database. + if _, err := s.MetaStore.CreateDatabase("db0"); err != nil { + t.Fatal(err) + } + + test := Test{ + queries: []*Query{ + &Query{ + name: "show users, no actual users", + command: `SHOW USERS`, + exp: `{"results":[{"series":[{"columns":["user","admin"]}]}]}`, + }, + &Query{ + name: `create user`, + command: "CREATE USER jdoe WITH PASSWORD '1337'", + exp: `{"results":[{}]}`, + }, + &Query{ + name: "show users, 1 existing user", + command: `SHOW USERS`, + exp: `{"results":[{"series":[{"columns":["user","admin"],"values":[["jdoe",false]]}]}]}`, + }, + &Query{ + name: "grant all priviledges to jdoe", + command: `GRANT ALL PRIVILEGES TO jdoe`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "show users, existing user as admin", + command: `SHOW USERS`, + exp: `{"results":[{"series":[{"columns":["user","admin"],"values":[["jdoe",true]]}]}]}`, + }, + &Query{ + name: "grant DB privileges to user", + command: `GRANT READ ON db0 TO jdoe`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "revoke all privileges", + command: `REVOKE ALL PRIVILEGES FROM jdoe`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "bad create user request", + command: `CREATE USER 0xBAD WITH PASSWORD pwd1337`, + exp: `{"error":"error parsing query: found 0, expected identifier at line 1, char 13"}`, + }, + &Query{ + name: "bad create user request, no name", + command: `CREATE USER WITH PASSWORD pwd1337`, + exp: `{"error":"error parsing query: found WITH, expected identifier at line 1, char 13"}`, + }, + &Query{ + name: "bad create user request, no password", + command: `CREATE USER jdoe`, + exp: `{"error":"error parsing query: found EOF, expected WITH at line 1, char 18"}`, + }, + &Query{ + name: "drop user", + command: `DROP USER jdoe`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "make sure user was dropped", + command: `SHOW USERS`, + exp: `{"results":[{"series":[{"columns":["user","admin"]}]}]}`, + }, + &Query{ + name: "delete non existing user", + command: `DROP USER noone`, + exp: `{"results":[{"error":"user not found"}]}`, + }, + }, + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(fmt.Sprintf("command: %s - err: %s", query.command, query.Error(err))) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can create a single point via json protocol and read it back. +func TestServer_Write_JSON(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + if res, err := s.Write("", "", fmt.Sprintf(`{"database" : "db0", "retentionPolicy" : "rp0", "points": [{"measurement": "cpu", "tags": {"host": "server02"},"fields": {"value": 1.0}}],"time":"%s"} `, now.Format(time.RFC3339Nano)), nil); err != nil { + t.Fatal(err) + } else if exp := ``; exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } + + // Verify the data was written. + if res, err := s.Query(`SELECT * FROM db0.rp0.cpu GROUP BY *`); err != nil { + t.Fatal(err) + } else if exp := fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server02"},"columns":["time","value"],"values":[["%s",1]]}]}]}`, now.Format(time.RFC3339Nano)); exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } +} + +// Ensure the server can create a single point via line protocol with float type and read it back. +func TestServer_Write_LineProtocol_Float(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + if res, err := s.Write("db0", "rp0", `cpu,host=server01 value=1.0 `+strconv.FormatInt(now.UnixNano(), 10), nil); err != nil { + t.Fatal(err) + } else if exp := ``; exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } + + // Verify the data was written. + if res, err := s.Query(`SELECT * FROM db0.rp0.cpu GROUP BY *`); err != nil { + t.Fatal(err) + } else if exp := fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",1]]}]}]}`, now.Format(time.RFC3339Nano)); exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } +} + +// Ensure the server can create a single point via line protocol with bool type and read it back. +func TestServer_Write_LineProtocol_Bool(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + if res, err := s.Write("db0", "rp0", `cpu,host=server01 value=true `+strconv.FormatInt(now.UnixNano(), 10), nil); err != nil { + t.Fatal(err) + } else if exp := ``; exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } + + // Verify the data was written. + if res, err := s.Query(`SELECT * FROM db0.rp0.cpu GROUP BY *`); err != nil { + t.Fatal(err) + } else if exp := fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",true]]}]}]}`, now.Format(time.RFC3339Nano)); exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } +} + +// Ensure the server can create a single point via line protocol with string type and read it back. +func TestServer_Write_LineProtocol_String(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + if res, err := s.Write("db0", "rp0", `cpu,host=server01 value="disk full" `+strconv.FormatInt(now.UnixNano(), 10), nil); err != nil { + t.Fatal(err) + } else if exp := ``; exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } + + // Verify the data was written. + if res, err := s.Query(`SELECT * FROM db0.rp0.cpu GROUP BY *`); err != nil { + t.Fatal(err) + } else if exp := fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s","disk full"]]}]}]}`, now.Format(time.RFC3339Nano)); exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } +} + +// Ensure the server can create a single point via line protocol with integer type and read it back. +func TestServer_Write_LineProtocol_Integer(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + if res, err := s.Write("db0", "rp0", `cpu,host=server01 value=100 `+strconv.FormatInt(now.UnixNano(), 10), nil); err != nil { + t.Fatal(err) + } else if exp := ``; exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } + + // Verify the data was written. + if res, err := s.Query(`SELECT * FROM db0.rp0.cpu GROUP BY *`); err != nil { + t.Fatal(err) + } else if exp := fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",100]]}]}]}`, now.Format(time.RFC3339Nano)); exp != res { + t.Fatalf("unexpected results\nexp: %s\ngot: %s\n", exp, res) + } +} + +// Ensure the server can query with default databases (via param) and default retention policy +func TestServer_Query_DefaultDBAndRP(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + test := NewTest("db0", "rp0") + test.write = fmt.Sprintf(`cpu value=1.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T01:00:00Z").UnixNano()) + + test.addQueries([]*Query{ + &Query{ + name: "default db and rp", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT * FROM cpu GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["2000-01-01T01:00:00Z",1]]}]}]}`, + }, + &Query{ + name: "default rp exists", + command: `show retention policies ON db0`, + exp: `{"results":[{"series":[{"columns":["name","duration","replicaN","default"],"values":[["default","0",1,false],["rp0","1h0m0s",1,true]]}]}]}`, + }, + &Query{ + name: "default rp", + command: `SELECT * FROM db0..cpu GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["2000-01-01T01:00:00Z",1]]}]}]}`, + }, + &Query{ + name: "default dp", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT * FROM rp0.cpu GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["2000-01-01T01:00:00Z",1]]}]}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can have a database with multiple measurements. +func TestServer_Query_Multiple_Measurements(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + // Make sure we do writes for measurements that will span across shards + writes := []string{ + fmt.Sprintf("cpu,host=server01 value=100,core=4 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf("cpu1,host=server02 value=50,core=2 %d", mustParseTime(time.RFC3339Nano, "2015-01-01T00:00:00Z").UnixNano()), + } + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "measurement in one shard but not another shouldn't panic server", + command: `SELECT host,value FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","host","value"],"values":[["2000-01-01T00:00:00Z","server01",100]]}]}]}`, + }, + &Query{ + name: "measurement in one shard but not another shouldn't panic server", + command: `SELECT host,value FROM db0.rp0.cpu GROUP BY host`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["2000-01-01T00:00:00Z",100]]}]}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server correctly supports data with identical tag values. +func TestServer_Query_IdenticalTagValues(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf("cpu,t1=val1 value=1 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf("cpu,t2=val2 value=2 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:01:00Z").UnixNano()), + fmt.Sprintf("cpu,t1=val2 value=3 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:02:00Z").UnixNano()), + } + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "measurements with identical tag values - SELECT *, no GROUP BY", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"t1":"val1","t2":""},"columns":["time","value"],"values":[["2000-01-01T00:00:00Z",1]]},{"name":"cpu","tags":{"t1":"val2","t2":""},"columns":["time","value"],"values":[["2000-01-01T00:02:00Z",3]]},{"name":"cpu","tags":{"t1":"","t2":"val2"},"columns":["time","value"],"values":[["2000-01-01T00:01:00Z",2]]}]}]}`, + }, + &Query{ + name: "measurements with identical tag values - SELECT *, with GROUP BY", + command: `SELECT value FROM db0.rp0.cpu GROUP BY t1,t2`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"t1":"val1","t2":""},"columns":["time","value"],"values":[["2000-01-01T00:00:00Z",1]]},{"name":"cpu","tags":{"t1":"val2","t2":""},"columns":["time","value"],"values":[["2000-01-01T00:02:00Z",3]]},{"name":"cpu","tags":{"t1":"","t2":"val2"},"columns":["time","value"],"values":[["2000-01-01T00:01:00Z",2]]}]}]}`, + }, + &Query{ + name: "measurements with identical tag values - SELECT value no GROUP BY", + command: `SELECT value FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["2000-01-01T00:00:00Z",1],["2000-01-01T00:01:00Z",2],["2000-01-01T00:02:00Z",3]]}]}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can handle a query that involves accessing no shards. +func TestServer_Query_NoShards(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = `cpu,host=server01 value=1 ` + strconv.FormatInt(now.UnixNano(), 10) + + test.addQueries([]*Query{ + &Query{ + name: "selecting value should succeed", + command: `SELECT value FROM db0.rp0.cpu WHERE time < now() - 1d`, + exp: `{"results":[{}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can query a non-existent field +func TestServer_Query_NonExistent(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = `cpu,host=server01 value=1 ` + strconv.FormatInt(now.UnixNano(), 10) + + test.addQueries([]*Query{ + &Query{ + name: "selecting value should succeed", + command: `SELECT value FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["%s",1]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "selecting non-existent should succeed", + command: `SELECT foo FROM db0.rp0.cpu`, + exp: `{"results":[{}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can perform basic math +func TestServer_Query_Math(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db", newRetentionPolicyInfo("rp", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + writes := []string{ + "float value=42 " + strconv.FormatInt(now.UnixNano(), 10), + "integer value=42i " + strconv.FormatInt(now.UnixNano(), 10), + } + + test := NewTest("db", "rp") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "SELECT multiple of float value", + command: `SELECT value * 2 from db.rp.float`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"float","columns":["time",""],"values":[["%s",84]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "SELECT multiple of float value", + command: `SELECT 2 * value from db.rp.float`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"float","columns":["time",""],"values":[["%s",84]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "SELECT multiple of integer value", + command: `SELECT value * 2 from db.rp.integer`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"integer","columns":["time",""],"values":[["%s",84]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "SELECT float multiple of integer value", + command: `SELECT value * 2.0 from db.rp.integer`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"integer","columns":["time",""],"values":[["%s",84]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can query with the count aggregate function +func TestServer_Query_Count(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = `cpu,host=server01 value=1.0 ` + strconv.FormatInt(now.UnixNano(), 10) + + hour_ago := now.Add(-time.Hour).UTC() + + test.addQueries([]*Query{ + &Query{ + name: "selecting count(value) should succeed", + command: `SELECT count(value) FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",1]]}]}]}`, + }, + &Query{ + name: "selecting count(value) with where time should return result", + command: fmt.Sprintf(`SELECT count(value) FROM db0.rp0.cpu WHERE time >= '%s'`, hour_ago.Format(time.RFC3339Nano)), + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","count"],"values":[["%s",1]]}]}]}`, hour_ago.Format(time.RFC3339Nano)), + }, + &Query{ + name: "selecting count(*) should error", + command: `SELECT count(*) FROM db0.rp0.cpu`, + exp: `{"results":[{"error":"expected field argument in count()"}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can query with Now(). +func TestServer_Query_Now(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = `cpu,host=server01 value=1.0 ` + strconv.FormatInt(now.UnixNano(), 10) + + test.addQueries([]*Query{ + &Query{ + name: "where with time < now() should work", + command: `SELECT * FROM db0.rp0.cpu where time < now()`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","host","value"],"values":[["%s","server01",1]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "where with time < now() and GROUP BY * should work", + command: `SELECT * FROM db0.rp0.cpu where time < now() GROUP BY *`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",1]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "where with time > now() should return an empty result", + command: `SELECT * FROM db0.rp0.cpu where time > now()`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "where with time > now() with GROUP BY * should return an empty result", + command: `SELECT * FROM db0.rp0.cpu where time > now() GROUP BY *`, + exp: `{"results":[{}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can query with epoch precisions. +func TestServer_Query_EpochPrecision(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = `cpu,host=server01 value=1.0 ` + strconv.FormatInt(now.UnixNano(), 10) + + test.addQueries([]*Query{ + &Query{ + name: "nanosecond precision", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + params: url.Values{"epoch": []string{"n"}}, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[[%d,1]]}]}]}`, now.UnixNano()), + }, + &Query{ + name: "microsecond precision", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + params: url.Values{"epoch": []string{"u"}}, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[[%d,1]]}]}]}`, now.UnixNano()/int64(time.Microsecond)), + }, + &Query{ + name: "millisecond precision", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + params: url.Values{"epoch": []string{"ms"}}, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[[%d,1]]}]}]}`, now.UnixNano()/int64(time.Millisecond)), + }, + &Query{ + name: "second precision", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + params: url.Values{"epoch": []string{"s"}}, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[[%d,1]]}]}]}`, now.UnixNano()/int64(time.Second)), + }, + &Query{ + name: "minute precision", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + params: url.Values{"epoch": []string{"m"}}, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[[%d,1]]}]}]}`, now.UnixNano()/int64(time.Minute)), + }, + &Query{ + name: "hour precision", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + params: url.Values{"epoch": []string{"h"}}, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[[%d,1]]}]}]}`, now.UnixNano()/int64(time.Hour)), + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server works with tag queries. +func TestServer_Query_Tags(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + writes := []string{ + fmt.Sprintf("cpu,host=server01 value=100,core=4 %d", now.UnixNano()), + fmt.Sprintf("cpu,host=server02 value=50,core=2 %d", now.Add(1).UnixNano()), + + fmt.Sprintf("cpu1,host=server01,region=us-west value=100 %d", mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + fmt.Sprintf("cpu1,host=server02 value=200 %d", mustParseTime(time.RFC3339Nano, "2010-02-28T01:03:37.703820946Z").UnixNano()), + fmt.Sprintf("cpu1,host=server03 value=300 %d", mustParseTime(time.RFC3339Nano, "2012-02-28T01:03:38.703820946Z").UnixNano()), + + fmt.Sprintf("cpu2,host=server01 value=100 %d", mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + fmt.Sprintf("cpu2 value=200 %d", mustParseTime(time.RFC3339Nano, "2012-02-28T01:03:38.703820946Z").UnixNano()), + + fmt.Sprintf("cpu3,company=acme01 value=100 %d", mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + fmt.Sprintf("cpu3 value=200 %d", mustParseTime(time.RFC3339Nano, "2012-02-28T01:03:38.703820946Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "tag without field should return error", + command: `SELECT host FROM db0.rp0.cpu`, + exp: `{"results":[{"error":"statement must have at least one field in select clause"}]}`, + }, + &Query{ + name: "field with tag should succeed", + command: `SELECT host, value FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","host","value"],"values":[["%s","server01",100],["%s","server02",50]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "field with tag and GROUP BY should succeed", + command: `SELECT host, value FROM db0.rp0.cpu GROUP BY host`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",100]]},{"name":"cpu","tags":{"host":"server02"},"columns":["time","value"],"values":[["%s",50]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "field with two tags should succeed", + command: `SELECT host, value, core FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","host","value","core"],"values":[["%s","server01",100,4],["%s","server02",50,2]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "field with two tags and GROUP BY should succeed", + command: `SELECT host, value, core FROM db0.rp0.cpu GROUP BY host`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value","core"],"values":[["%s",100,4]]},{"name":"cpu","tags":{"host":"server02"},"columns":["time","value","core"],"values":[["%s",50,2]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "select * with tags should succeed", + command: `SELECT * FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","core","host","value"],"values":[["%s",4,"server01",100],["%s",2,"server02",50]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "select * with tags with GROUP BY * should succeed", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","core","value"],"values":[["%s",4,100]]},{"name":"cpu","tags":{"host":"server02"},"columns":["time","core","value"],"values":[["%s",2,50]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "group by tag", + command: `SELECT value FROM db0.rp0.cpu GROUP by host`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",100]]},{"name":"cpu","tags":{"host":"server02"},"columns":["time","value"],"values":[["%s",50]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "single field (EQ tag value1)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host = 'server01'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + &Query{ + name: "single field (2 EQ tags)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host = 'server01' AND region = 'us-west'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + &Query{ + name: "single field (OR different tags)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host = 'server03' OR region = 'us-west'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2012-02-28T01:03:38.703820946Z",300],["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + &Query{ + name: "single field (OR with non-existent tag value)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host = 'server01' OR host = 'server66'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + &Query{ + name: "single field (OR with all tag values)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host = 'server01' OR host = 'server02' OR host = 'server03'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2010-02-28T01:03:37.703820946Z",200],["2012-02-28T01:03:38.703820946Z",300],["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + &Query{ + name: "single field (1 EQ and 1 NEQ tag)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host = 'server01' AND region != 'us-west'`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "single field (EQ tag value2)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host = 'server02'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2010-02-28T01:03:37.703820946Z",200]]}]}]}`, + }, + &Query{ + name: "single field (NEQ tag value1)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host != 'server01'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2010-02-28T01:03:37.703820946Z",200],["2012-02-28T01:03:38.703820946Z",300]]}]}]}`, + }, + &Query{ + name: "single field (NEQ tag value1 AND NEQ tag value2)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host != 'server01' AND host != 'server02'`, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2012-02-28T01:03:38.703820946Z",300]]}]}]}`, + }, + &Query{ + name: "single field (NEQ tag value1 OR NEQ tag value2)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host != 'server01' OR host != 'server02'`, // Yes, this is always true, but that's the point. + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","value"],"values":[["2010-02-28T01:03:37.703820946Z",200],["2012-02-28T01:03:38.703820946Z",300],["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + &Query{ + name: "single field (NEQ tag value1 AND NEQ tag value2 AND NEQ tag value3)", + command: `SELECT value FROM db0.rp0.cpu1 WHERE host != 'server01' AND host != 'server02' AND host != 'server03'`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "single field (NEQ tag value1, point without any tags)", + command: `SELECT value FROM db0.rp0.cpu2 WHERE host != 'server01'`, + exp: `{"results":[{"series":[{"name":"cpu2","columns":["time","value"],"values":[["2012-02-28T01:03:38.703820946Z",200]]}]}]}`, + }, + &Query{ + name: "single field (NEQ tag value1, point without any tags)", + command: `SELECT value FROM db0.rp0.cpu3 WHERE company !~ /acme01/`, + exp: `{"results":[{"series":[{"name":"cpu3","columns":["time","value"],"values":[["2012-02-28T01:03:38.703820946Z",200]]}]}]}`, + }, + &Query{ + name: "single field (regex tag match)", + command: `SELECT value FROM db0.rp0.cpu3 WHERE company =~ /acme01/`, + exp: `{"results":[{"series":[{"name":"cpu3","columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + &Query{ + name: "single field (regex tag match)", + command: `SELECT value FROM db0.rp0.cpu3 WHERE company !~ /acme[23]/`, + exp: `{"results":[{"series":[{"name":"cpu3","columns":["time","value"],"values":[["2012-02-28T01:03:38.703820946Z",200],["2015-02-28T01:03:36.703820946Z",100]]}]}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server correctly queries with an alias. +func TestServer_Query_Alias(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf("cpu value=1i,steps=3i %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf("cpu value=2i,steps=4i %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:01:00Z").UnixNano()), + } + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "baseline query - SELECT * FROM db0.rp0.cpu", + command: `SELECT * FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","steps","value"],"values":[["2000-01-01T00:00:00Z",3,1],["2000-01-01T00:01:00Z",4,2]]}]}]}`, + }, + &Query{ + name: "basic query with alias - SELECT steps, value as v FROM db0.rp0.cpu", + command: `SELECT steps, value as v FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","steps","v"],"values":[["2000-01-01T00:00:00Z",3,1],["2000-01-01T00:01:00Z",4,2]]}]}]}`, + }, + &Query{ + name: "double aggregate sum - SELECT sum(value), sum(steps) FROM db0.rp0.cpu", + command: `SELECT sum(value), sum(steps) FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","sum","sum"],"values":[["1970-01-01T00:00:00Z",3,7]]}]}]}`, + }, + &Query{ + name: "double aggregate sum reverse order - SELECT sum(steps), sum(value) FROM db0.rp0.cpu", + command: `SELECT sum(steps), sum(value) FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","sum","sum"],"values":[["1970-01-01T00:00:00Z",7,3]]}]}]}`, + }, + &Query{ + name: "double aggregate sum with alias - SELECT sum(value) as sumv, sum(steps) as sums FROM db0.rp0.cpu", + command: `SELECT sum(value) as sumv, sum(steps) as sums FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","sumv","sums"],"values":[["1970-01-01T00:00:00Z",3,7]]}]}]}`, + }, + &Query{ + name: "double aggregate with same value - SELECT sum(value), mean(value) FROM db0.rp0.cpu", + command: `SELECT sum(value), mean(value) FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","sum","mean"],"values":[["1970-01-01T00:00:00Z",3,1.5]]}]}]}`, + }, + &Query{ + name: "double aggregate with same value and same alias - SELECT mean(value) as mv, max(value) as mv FROM db0.rp0.cpu", + command: `SELECT mean(value) as mv, max(value) as mv FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","mv","mv"],"values":[["1970-01-01T00:00:00Z",1.5,2]]}]}]}`, + }, + &Query{ + name: "double aggregate with non-existent field - SELECT mean(value), max(foo) FROM db0.rp0.cpu", + command: `SELECT mean(value), max(foo) FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","mean","max"],"values":[["1970-01-01T00:00:00Z",1.5,null]]}]}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server will succeed and error for common scenarios. +func TestServer_Query_Common(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = fmt.Sprintf("cpu,host=server01 value=1 %s", strconv.FormatInt(now.UnixNano(), 10)) + + test.addQueries([]*Query{ + &Query{ + name: "selecting a from a non-existent database should error", + command: `SELECT value FROM db1.rp0.cpu`, + exp: `{"results":[{"error":"database not found: db1"}]}`, + }, + &Query{ + name: "selecting a from a non-existent retention policy should error", + command: `SELECT value FROM db0.rp1.cpu`, + exp: `{"results":[{"error":"retention policy not found"}]}`, + }, + &Query{ + name: "selecting a valid measurement and field should succeed", + command: `SELECT value FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["%s",1]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "explicitly selecting time and a valid measurement and field should succeed", + command: `SELECT time,value FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["%s",1]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "selecting a measurement that doesn't exist should result in empty set", + command: `SELECT value FROM db0.rp0.idontexist`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "selecting a field that doesn't exist should result in empty set", + command: `SELECT idontexist FROM db0.rp0.cpu`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "selecting wildcard without specifying a database should error", + command: `SELECT * FROM cpu`, + exp: `{"results":[{"error":"database name required"}]}`, + }, + &Query{ + name: "selecting explicit field without specifying a database should error", + command: `SELECT value FROM cpu`, + exp: `{"results":[{"error":"database name required"}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can query two points. +func TestServer_Query_SelectTwoPoints(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = fmt.Sprintf("cpu value=100 %s\ncpu value=200 %s", strconv.FormatInt(now.UnixNano(), 10), strconv.FormatInt(now.Add(1).UnixNano(), 10)) + + test.addQueries( + &Query{ + name: "selecting two points should result in two points", + command: `SELECT * FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["%s",100],["%s",200]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + &Query{ + name: "selecting two points with GROUP BY * should result in two points", + command: `SELECT * FROM db0.rp0.cpu GROUP BY *`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["%s",100],["%s",200]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }, + ) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can query two negative points. +func TestServer_Query_SelectTwoNegativePoints(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + + test := NewTest("db0", "rp0") + test.write = fmt.Sprintf("cpu value=-100 %s\ncpu value=-200 %s", strconv.FormatInt(now.UnixNano(), 10), strconv.FormatInt(now.Add(1).UnixNano(), 10)) + + test.addQueries(&Query{ + name: "selecting two negative points should succeed", + command: `SELECT * FROM db0.rp0.cpu`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[["%s",-100],["%s",-200]]}]}]}`, now.Format(time.RFC3339Nano), now.Add(1).Format(time.RFC3339Nano)), + }) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can query with relative time. +func TestServer_Query_SelectRelativeTime(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + now := now() + yesterday := yesterday() + + test := NewTest("db0", "rp0") + test.write = fmt.Sprintf("cpu,host=server01 value=100 %s\ncpu,host=server01 value=200 %s", strconv.FormatInt(yesterday.UnixNano(), 10), strconv.FormatInt(now.UnixNano(), 10)) + + test.addQueries([]*Query{ + &Query{ + name: "single point with time pre-calculated for past time queries yesterday", + command: `SELECT * FROM db0.rp0.cpu where time >= '` + yesterday.Add(-1*time.Minute).Format(time.RFC3339Nano) + `' GROUP BY *`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",100],["%s",200]]}]}]}`, yesterday.Format(time.RFC3339Nano), now.Format(time.RFC3339Nano)), + }, + &Query{ + name: "single point with time pre-calculated for relative time queries now", + command: `SELECT * FROM db0.rp0.cpu where time >= now() - 1m GROUP BY *`, + exp: fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["%s",200]]}]}]}`, now.Format(time.RFC3339Nano)), + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Ensure the server can handle various simple calculus queries. +func TestServer_Query_SelectRawCalculus(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + test := NewTest("db0", "rp0") + test.write = fmt.Sprintf("cpu value=210 1278010021000000000\ncpu value=10 1278010022000000000") + + test.addQueries([]*Query{ + &Query{ + name: "calculate single derivate", + command: `SELECT derivative(value) from db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","derivative"],"values":[["2010-07-01T18:47:02Z",-200]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// mergeMany ensures that when merging many series together and some of them have a different number +// of points than others in a group by interval the results are correct +func TestServer_Query_MergeMany(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + // set infinite retention policy as we are inserting data in the past and don't want retention policy enforcement to make this test racy + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + + test := NewTest("db0", "rp0") + + writes := []string{} + for i := 1; i < 11; i++ { + for j := 1; j < 5+i%3; j++ { + data := fmt.Sprintf(`cpu,host=server_%d value=22 %d`, i, time.Unix(int64(j), int64(0)).UTC().UnixNano()) + writes = append(writes, data) + } + } + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "GROUP by time", + command: `SELECT count(value) FROM db0.rp0.cpu WHERE time >= '1970-01-01T00:00:01Z' AND time <= '1970-01-01T00:00:06Z' GROUP BY time(1s)`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","count"],"values":[["1970-01-01T00:00:01Z",10],["1970-01-01T00:00:02Z",10],["1970-01-01T00:00:03Z",10],["1970-01-01T00:00:04Z",10],["1970-01-01T00:00:05Z",7],["1970-01-01T00:00:06Z",3]]}]}]}`, + }, + &Query{ + skip: true, + name: "GROUP by tag - FIXME issue #2875", + command: `SELECT count(value) FROM db0.rp0.cpu where time >= '2000-01-01T00:00:00Z' and time <= '2000-01-01T02:00:00Z' group by host`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","count"],"values":[["2000-01-01T00:00:00Z",1]]},{"name":"cpu","tags":{"host":"server02"},"columns":["time","count"],"values":[["2000-01-01T00:00:00Z",1]]},{"name":"cpu","tags":{"host":"server03"},"columns":["time","count"],"values":[["2000-01-01T00:00:00Z",1]]}]}]}`, + }, + &Query{ + name: "GROUP by field", + command: `SELECT count(value) FROM db0.rp0.cpu group by value`, + exp: `{"results":[{"error":"can not use field in GROUP BY clause: value"}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_SLimitAndSOffset(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + // set infinite retention policy as we are inserting data in the past and don't want retention policy enforcement to make this test racy + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + + test := NewTest("db0", "rp0") + + writes := []string{} + for i := 1; i < 10; i++ { + data := fmt.Sprintf(`cpu,region=us-east,host=server-%d value=%d %d`, i, i, time.Unix(int64(i), int64(0)).UnixNano()) + writes = append(writes, data) + } + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "SLIMIT 2 SOFFSET 1", + command: `SELECT count(value) FROM db0.rp0.cpu GROUP BY * SLIMIT 2 SOFFSET 1`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"server-2","region":"us-east"},"columns":["time","count"],"values":[["1970-01-01T00:00:00Z",1]]},{"name":"cpu","tags":{"host":"server-3","region":"us-east"},"columns":["time","count"],"values":[["1970-01-01T00:00:00Z",1]]}]}]}`, + }, + &Query{ + name: "SLIMIT 2 SOFFSET 3", + command: `SELECT count(value) FROM db0.rp0.cpu GROUP BY * SLIMIT 2 SOFFSET 3`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"server-4","region":"us-east"},"columns":["time","count"],"values":[["1970-01-01T00:00:00Z",1]]},{"name":"cpu","tags":{"host":"server-5","region":"us-east"},"columns":["time","count"],"values":[["1970-01-01T00:00:00Z",1]]}]}]}`, + }, + &Query{ + name: "SLIMIT 3 SOFFSET 8", + command: `SELECT count(value) FROM db0.rp0.cpu GROUP BY * SLIMIT 3 SOFFSET 8`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"server-9","region":"us-east"},"columns":["time","count"],"values":[["1970-01-01T00:00:00Z",1]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_Regex(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu1,host=server01 value=10 %d`, mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + fmt.Sprintf(`cpu2,host=server01 value=20 %d`, mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + fmt.Sprintf(`cpu3,host=server01 value=30 %d`, mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "default db and rp", + command: `SELECT * FROM /cpu[13]/`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu1","columns":["time","host","value"],"values":[["2015-02-28T01:03:36.703820946Z","server01",10]]},{"name":"cpu3","columns":["time","host","value"],"values":[["2015-02-28T01:03:36.703820946Z","server01",30]]}]}]}`, + }, + &Query{ + name: "default db and rp with GROUP BY *", + command: `SELECT * FROM /cpu[13]/ GROUP BY *`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu1","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",10]]},{"name":"cpu3","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",30]]}]}]}`, + }, + &Query{ + name: "specifying db and rp", + command: `SELECT * FROM db0.rp0./cpu[13]/ GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu1","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",10]]},{"name":"cpu3","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",30]]}]}]}`, + }, + &Query{ + name: "default db and specified rp", + command: `SELECT * FROM rp0./cpu[13]/ GROUP BY *`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu1","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",10]]},{"name":"cpu3","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",30]]}]}]}`, + }, + &Query{ + name: "specified db and default rp", + command: `SELECT * FROM db0../cpu[13]/ GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu1","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",10]]},{"name":"cpu3","tags":{"host":"server01"},"columns":["time","value"],"values":[["2015-02-28T01:03:36.703820946Z",30]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_Aggregates(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`int value=45 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + + fmt.Sprintf(`intmax value=%s %d`, maxInt64(), mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`intmax value=%s %d`, maxInt64(), mustParseTime(time.RFC3339Nano, "2000-01-01T01:00:00Z").UnixNano()), + + fmt.Sprintf(`intmany,host=server01 value=2.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`intmany,host=server02 value=4.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`intmany,host=server03 value=4.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:20Z").UnixNano()), + fmt.Sprintf(`intmany,host=server04 value=4.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:30Z").UnixNano()), + fmt.Sprintf(`intmany,host=server05 value=5.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:40Z").UnixNano()), + fmt.Sprintf(`intmany,host=server06 value=5.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:50Z").UnixNano()), + fmt.Sprintf(`intmany,host=server07 value=7.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:01:00Z").UnixNano()), + fmt.Sprintf(`intmany,host=server08 value=9.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:01:10Z").UnixNano()), + + fmt.Sprintf(`intoverlap,region=us-east value=20 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`intoverlap,region=us-east value=30 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`intoverlap,region=us-west value=100 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`intoverlap,region=us-east otherVal=20 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:03Z").UnixNano()), + + fmt.Sprintf(`floatsingle value=45.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + + fmt.Sprintf(`floatmax value=%s %d`, maxFloat64(), mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`floatmax value=%s %d`, maxFloat64(), mustParseTime(time.RFC3339Nano, "2000-01-01T01:00:00Z").UnixNano()), + + fmt.Sprintf(`floatmany,host=server01 value=2.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`floatmany,host=server02 value=4.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`floatmany,host=server03 value=4.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:20Z").UnixNano()), + fmt.Sprintf(`floatmany,host=server04 value=4.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:30Z").UnixNano()), + fmt.Sprintf(`floatmany,host=server05 value=5.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:40Z").UnixNano()), + fmt.Sprintf(`floatmany,host=server06 value=5.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:50Z").UnixNano()), + fmt.Sprintf(`floatmany,host=server07 value=7.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:01:00Z").UnixNano()), + fmt.Sprintf(`floatmany,host=server08 value=9.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:01:10Z").UnixNano()), + + fmt.Sprintf(`floatoverlap,region=us-east value=20.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`floatoverlap,region=us-east value=30.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`floatoverlap,region=us-west value=100.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`floatoverlap,region=us-east otherVal=20.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:03Z").UnixNano()), + + fmt.Sprintf(`load,region=us-east,host=serverA value=20.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`load,region=us-east,host=serverB value=30.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`load,region=us-west,host=serverC value=100.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + + fmt.Sprintf(`cpu,region=uk,host=serverZ,service=redis value=20.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:03Z").UnixNano()), + fmt.Sprintf(`cpu,region=uk,host=serverZ,service=mysql value=30.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:03Z").UnixNano()), + + fmt.Sprintf(`stringdata value="first" %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:03Z").UnixNano()), + fmt.Sprintf(`stringdata value="last" %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:04Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + // int64 + &Query{ + name: "stddev with just one point - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT STDDEV(value) FROM int`, + exp: `{"results":[{"series":[{"name":"int","columns":["time","stddev"],"values":[["1970-01-01T00:00:00Z",null]]}]}]}`, + }, + &Query{ + name: "large mean and stddev - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEAN(value), STDDEV(value) FROM intmax`, + exp: `{"results":[{"series":[{"name":"intmax","columns":["time","mean","stddev"],"values":[["1970-01-01T00:00:00Z",` + maxInt64() + `,0]]}]}]}`, + }, + &Query{ + name: "mean and stddev - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEAN(value), STDDEV(value) FROM intmany WHERE time >= '2000-01-01' AND time < '2000-01-01T00:02:00Z' GROUP BY time(10m)`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","mean","stddev"],"values":[["2000-01-01T00:00:00Z",5,2.138089935299395]]}]}]}`, + }, + &Query{ + name: "first - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT FIRST(value) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","first"],"values":[["1970-01-01T00:00:00Z",2]]}]}]}`, + }, + &Query{ + name: "last - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT LAST(value) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","last"],"values":[["1970-01-01T00:00:00Z",9]]}]}]}`, + }, + &Query{ + name: "spread - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SPREAD(value) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","spread"],"values":[["1970-01-01T00:00:00Z",7]]}]}]}`, + }, + &Query{ + name: "median - even count - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEDIAN(value) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","median"],"values":[["1970-01-01T00:00:00Z",4.5]]}]}]}`, + }, + &Query{ + name: "median - odd count - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEDIAN(value) FROM intmany where time < '2000-01-01T00:01:10Z'`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","median"],"values":[["1970-01-01T00:00:00Z",4]]}]}]}`, + }, + &Query{ + name: "distinct as call - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT(value) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","distinct"],"values":[["1970-01-01T00:00:00Z",[2,4,5,7,9]]]}]}]}`, + }, + &Query{ + name: "distinct alt syntax - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT value FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","distinct"],"values":[["1970-01-01T00:00:00Z",[2,4,5,7,9]]]}]}]}`, + }, + &Query{ + name: "distinct select tag - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT(host) FROM intmany`, + exp: `{"results":[{"error":"statement must have at least one field in select clause"}]}`, + }, + &Query{ + name: "distinct alt select tag - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT host FROM intmany`, + exp: `{"results":[{"error":"statement must have at least one field in select clause"}]}`, + }, + &Query{ + name: "count distinct - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT value) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",5]]}]}]}`, + }, + &Query{ + name: "count distinct as call - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT(value)) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",5]]}]}]}`, + }, + &Query{ + name: "count distinct select tag - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT host) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",0]]}]}]}`, + }, + &Query{ + name: "count distinct as call select tag - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT host) FROM intmany`, + exp: `{"results":[{"series":[{"name":"intmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",0]]}]}]}`, + }, + &Query{ + name: "aggregation with no interval - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT count(value) FROM intoverlap WHERE time = '2000-01-01 00:00:00'`, + exp: `{"results":[{"series":[{"name":"intoverlap","columns":["time","count"],"values":[["2000-01-01T00:00:00Z",2]]}]}]}`, + }, + &Query{ + name: "sum - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SUM(value) FROM intoverlap WHERE time >= '2000-01-01 00:00:05' AND time <= '2000-01-01T00:00:10Z' GROUP BY time(10s), region`, + exp: `{"results":[{"series":[{"name":"intoverlap","tags":{"region":"us-east"},"columns":["time","sum"],"values":[["2000-01-01T00:00:00Z",null],["2000-01-01T00:00:10Z",30]]}]}]}`, + }, + &Query{ + name: "aggregation with a null field value - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SUM(value) FROM intoverlap GROUP BY region`, + exp: `{"results":[{"series":[{"name":"intoverlap","tags":{"region":"us-east"},"columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",50]]},{"name":"intoverlap","tags":{"region":"us-west"},"columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",100]]}]}]}`, + }, + &Query{ + name: "multiple aggregations - int", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SUM(value), MEAN(value) FROM intoverlap GROUP BY region`, + exp: `{"results":[{"series":[{"name":"intoverlap","tags":{"region":"us-east"},"columns":["time","sum","mean"],"values":[["1970-01-01T00:00:00Z",50,25]]},{"name":"intoverlap","tags":{"region":"us-west"},"columns":["time","sum","mean"],"values":[["1970-01-01T00:00:00Z",100,100]]}]}]}`, + }, + &Query{ + skip: true, + name: "multiple aggregations with division - int FIXME issue #2879", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT sum(value), mean(value), sum(value) / mean(value) as div FROM intoverlap GROUP BY region`, + exp: `{"results":[{"series":[{"name":"intoverlap","tags":{"region":"us-east"},"columns":["time","sum","mean","div"],"values":[["1970-01-01T00:00:00Z",50,25,2]]},{"name":"intoverlap","tags":{"region":"us-west"},"columns":["time","div"],"values":[["1970-01-01T00:00:00Z",100,100,1]]}]}]}`, + }, + + // float64 + &Query{ + name: "stddev with just one point - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT STDDEV(value) FROM floatsingle`, + exp: `{"results":[{"series":[{"name":"floatsingle","columns":["time","stddev"],"values":[["1970-01-01T00:00:00Z",null]]}]}]}`, + }, + &Query{ + name: "large mean and stddev - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEAN(value), STDDEV(value) FROM floatmax`, + exp: `{"results":[{"series":[{"name":"floatmax","columns":["time","mean","stddev"],"values":[["1970-01-01T00:00:00Z",` + maxFloat64() + `,0]]}]}]}`, + }, + &Query{ + name: "mean and stddev - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEAN(value), STDDEV(value) FROM floatmany WHERE time >= '2000-01-01' AND time < '2000-01-01T00:02:00Z' GROUP BY time(10m)`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","mean","stddev"],"values":[["2000-01-01T00:00:00Z",5,2.138089935299395]]}]}]}`, + }, + &Query{ + name: "first - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT FIRST(value) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","first"],"values":[["1970-01-01T00:00:00Z",2]]}]}]}`, + }, + &Query{ + name: "last - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT LAST(value) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","last"],"values":[["1970-01-01T00:00:00Z",9]]}]}]}`, + }, + &Query{ + name: "spread - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SPREAD(value) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","spread"],"values":[["1970-01-01T00:00:00Z",7]]}]}]}`, + }, + &Query{ + name: "median - even count - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEDIAN(value) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","median"],"values":[["1970-01-01T00:00:00Z",4.5]]}]}]}`, + }, + &Query{ + name: "median - odd count - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEDIAN(value) FROM floatmany where time < '2000-01-01T00:01:10Z'`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","median"],"values":[["1970-01-01T00:00:00Z",4]]}]}]}`, + }, + &Query{ + name: "distinct as call - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT(value) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","distinct"],"values":[["1970-01-01T00:00:00Z",[2,4,5,7,9]]]}]}]}`, + }, + &Query{ + name: "distinct alt syntax - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT value FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","distinct"],"values":[["1970-01-01T00:00:00Z",[2,4,5,7,9]]]}]}]}`, + }, + &Query{ + name: "distinct select tag - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT(host) FROM floatmany`, + exp: `{"results":[{"error":"statement must have at least one field in select clause"}]}`, + }, + &Query{ + name: "distinct alt select tag - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT DISTINCT host FROM floatmany`, + exp: `{"results":[{"error":"statement must have at least one field in select clause"}]}`, + }, + &Query{ + name: "count distinct - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT value) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",5]]}]}]}`, + }, + &Query{ + name: "count distinct as call - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT(value)) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",5]]}]}]}`, + }, + &Query{ + name: "count distinct select tag - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT host) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",0]]}]}]}`, + }, + &Query{ + name: "count distinct as call select tag - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(DISTINCT host) FROM floatmany`, + exp: `{"results":[{"series":[{"name":"floatmany","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",0]]}]}]}`, + }, + &Query{ + name: "aggregation with no interval - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT count(value) FROM floatoverlap WHERE time = '2000-01-01 00:00:00'`, + exp: `{"results":[{"series":[{"name":"floatoverlap","columns":["time","count"],"values":[["2000-01-01T00:00:00Z",2]]}]}]}`, + }, + &Query{ + name: "sum - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SUM(value) FROM floatoverlap WHERE time >= '2000-01-01 00:00:05' AND time <= '2000-01-01T00:00:10Z' GROUP BY time(10s), region`, + exp: `{"results":[{"series":[{"name":"floatoverlap","tags":{"region":"us-east"},"columns":["time","sum"],"values":[["2000-01-01T00:00:00Z",null],["2000-01-01T00:00:10Z",30]]}]}]}`, + }, + &Query{ + name: "aggregation with a null field value - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SUM(value) FROM floatoverlap GROUP BY region`, + exp: `{"results":[{"series":[{"name":"floatoverlap","tags":{"region":"us-east"},"columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",50]]},{"name":"floatoverlap","tags":{"region":"us-west"},"columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",100]]}]}]}`, + }, + &Query{ + name: "multiple aggregations - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT SUM(value), MEAN(value) FROM floatoverlap GROUP BY region`, + exp: `{"results":[{"series":[{"name":"floatoverlap","tags":{"region":"us-east"},"columns":["time","sum","mean"],"values":[["1970-01-01T00:00:00Z",50,25]]},{"name":"floatoverlap","tags":{"region":"us-west"},"columns":["time","sum","mean"],"values":[["1970-01-01T00:00:00Z",100,100]]}]}]}`, + }, + &Query{ + name: "multiple aggregations with division - float", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT sum(value) / mean(value) as div FROM floatoverlap GROUP BY region`, + exp: `{"results":[{"series":[{"name":"floatoverlap","tags":{"region":"us-east"},"columns":["time","div"],"values":[["1970-01-01T00:00:00Z",2]]},{"name":"floatoverlap","tags":{"region":"us-west"},"columns":["time","div"],"values":[["1970-01-01T00:00:00Z",1]]}]}]}`, + }, + + // strings + &Query{ + name: "STDDEV on string data - string", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT STDDEV(value) FROM stringdata`, + exp: `{"results":[{"series":[{"name":"stringdata","columns":["time","stddev"],"values":[["1970-01-01T00:00:00Z",null]]}]}]}`, + }, + &Query{ + name: "MEAN on string data - string", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEAN(value) FROM stringdata`, + exp: `{"results":[{"series":[{"name":"stringdata","columns":["time","mean"],"values":[["1970-01-01T00:00:00Z",0]]}]}]}`, + }, + &Query{ + name: "MEDIAN on string data - string", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT MEDIAN(value) FROM stringdata`, + exp: `{"results":[{"series":[{"name":"stringdata","columns":["time","median"],"values":[["1970-01-01T00:00:00Z",null]]}]}]}`, + }, + &Query{ + name: "COUNT on string data - string", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT COUNT(value) FROM stringdata`, + exp: `{"results":[{"series":[{"name":"stringdata","columns":["time","count"],"values":[["1970-01-01T00:00:00Z",2]]}]}]}`, + }, + &Query{ + name: "FIRST on string data - string", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT FIRST(value) FROM stringdata`, + exp: `{"results":[{"series":[{"name":"stringdata","columns":["time","first"],"values":[["1970-01-01T00:00:00Z","first"]]}]}]}`, + }, + &Query{ + name: "LAST on string data - string", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT LAST(value) FROM stringdata`, + exp: `{"results":[{"series":[{"name":"stringdata","columns":["time","last"],"values":[["1970-01-01T00:00:00Z","last"]]}]}]}`, + }, + + // general queries + &Query{ + name: "group by multiple dimensions", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT sum(value) FROM load GROUP BY region, host`, + exp: `{"results":[{"series":[{"name":"load","tags":{"host":"serverA","region":"us-east"},"columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",20]]},{"name":"load","tags":{"host":"serverB","region":"us-east"},"columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",30]]},{"name":"load","tags":{"host":"serverC","region":"us-west"},"columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",100]]}]}]}`, + }, + &Query{ + name: "aggregation with WHERE and AND", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT sum(value) FROM cpu WHERE region='uk' AND host='serverZ'`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","sum"],"values":[["1970-01-01T00:00:00Z",50]]}]}]}`, + }, + + // Mathematics + &Query{ + name: "group by multiple dimensions", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT sum(value)*2 FROM load`, + exp: `{"results":[{"series":[{"name":"load","columns":["time",""],"values":[["1970-01-01T00:00:00Z",300]]}]}]}`, + }, + &Query{ + name: "group by multiple dimensions", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT sum(value)/2 FROM load`, + exp: `{"results":[{"series":[{"name":"load","columns":["time",""],"values":[["1970-01-01T00:00:00Z",75]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Write_Precision(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []struct { + write string + params url.Values + }{ + { + write: fmt.Sprintf("cpu_n0_precision value=1 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T12:34:56.789012345Z").UnixNano()), + }, + { + write: fmt.Sprintf("cpu_n1_precision value=1.1 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T12:34:56.789012345Z").UnixNano()), + params: url.Values{"precision": []string{"n"}}, + }, + { + write: fmt.Sprintf("cpu_u_precision value=100 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T12:34:56.789012345Z").Truncate(time.Microsecond).UnixNano()/int64(time.Microsecond)), + params: url.Values{"precision": []string{"u"}}, + }, + { + write: fmt.Sprintf("cpu_ms_precision value=200 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T12:34:56.789012345Z").Truncate(time.Millisecond).UnixNano()/int64(time.Millisecond)), + params: url.Values{"precision": []string{"ms"}}, + }, + { + write: fmt.Sprintf("cpu_s_precision value=300 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T12:34:56.789012345Z").Truncate(time.Second).UnixNano()/int64(time.Second)), + params: url.Values{"precision": []string{"s"}}, + }, + { + write: fmt.Sprintf("cpu_m_precision value=400 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T12:34:56.789012345Z").Truncate(time.Minute).UnixNano()/int64(time.Minute)), + params: url.Values{"precision": []string{"m"}}, + }, + { + write: fmt.Sprintf("cpu_h_precision value=500 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T12:34:56.789012345Z").Truncate(time.Hour).UnixNano()/int64(time.Hour)), + params: url.Values{"precision": []string{"h"}}, + }, + } + + test := NewTest("db0", "rp0") + + test.addQueries([]*Query{ + &Query{ + name: "point with nanosecond precision time - no precision specified on write", + command: `SELECT * FROM cpu_n0_precision`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu_n0_precision","columns":["time","value"],"values":[["2000-01-01T12:34:56.789012345Z",1]]}]}]}`, + }, + &Query{ + name: "point with nanosecond precision time", + command: `SELECT * FROM cpu_n1_precision`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu_n1_precision","columns":["time","value"],"values":[["2000-01-01T12:34:56.789012345Z",1.1]]}]}]}`, + }, + &Query{ + name: "point with microsecond precision time", + command: `SELECT * FROM cpu_u_precision`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu_u_precision","columns":["time","value"],"values":[["2000-01-01T12:34:56.789012Z",100]]}]}]}`, + }, + &Query{ + name: "point with millisecond precision time", + command: `SELECT * FROM cpu_ms_precision`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu_ms_precision","columns":["time","value"],"values":[["2000-01-01T12:34:56.789Z",200]]}]}]}`, + }, + &Query{ + name: "point with second precision time", + command: `SELECT * FROM cpu_s_precision`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu_s_precision","columns":["time","value"],"values":[["2000-01-01T12:34:56Z",300]]}]}]}`, + }, + &Query{ + name: "point with minute precision time", + command: `SELECT * FROM cpu_m_precision`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu_m_precision","columns":["time","value"],"values":[["2000-01-01T12:34:00Z",400]]}]}]}`, + }, + &Query{ + name: "point with hour precision time", + command: `SELECT * FROM cpu_h_precision`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"cpu_h_precision","columns":["time","value"],"values":[["2000-01-01T12:00:00Z",500]]}]}]}`, + }, + }...) + + // we are doing writes that require parameter changes, so we are fighting the test harness a little to make this happen properly + for _, w := range writes { + test.write = w.write + test.params = w.params + test.initialized = false + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_Wildcards(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`wildcard,region=us-east value=10 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`wildcard,region=us-east valx=20 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`wildcard,region=us-east value=30,valx=40 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:20Z").UnixNano()), + + fmt.Sprintf(`wgroup,region=us-east value=10.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`wgroup,region=us-east value=20.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`wgroup,region=us-west value=30.0 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:20Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "wildcard", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT * FROM wildcard`, + exp: `{"results":[{"series":[{"name":"wildcard","columns":["time","region","value","valx"],"values":[["2000-01-01T00:00:00Z","us-east",10,null],["2000-01-01T00:00:10Z","us-east",null,20],["2000-01-01T00:00:20Z","us-east",30,40]]}]}]}`, + }, + &Query{ + name: "wildcard with group by", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT * FROM wildcard GROUP BY *`, + exp: `{"results":[{"series":[{"name":"wildcard","tags":{"region":"us-east"},"columns":["time","value","valx"],"values":[["2000-01-01T00:00:00Z",10,null],["2000-01-01T00:00:10Z",null,20],["2000-01-01T00:00:20Z",30,40]]}]}]}`, + }, + &Query{ + name: "GROUP BY queries", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT mean(value) FROM wgroup GROUP BY *`, + exp: `{"results":[{"series":[{"name":"wgroup","tags":{"region":"us-east"},"columns":["time","mean"],"values":[["1970-01-01T00:00:00Z",15]]},{"name":"wgroup","tags":{"region":"us-west"},"columns":["time","mean"],"values":[["1970-01-01T00:00:00Z",30]]}]}]}`, + }, + &Query{ + name: "GROUP BY queries with time", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT mean(value) FROM wgroup WHERE time >= '2000-01-01T00:00:00Z' AND time < '2000-01-01T00:01:00Z' GROUP BY *,TIME(1m)`, + exp: `{"results":[{"series":[{"name":"wgroup","tags":{"region":"us-east"},"columns":["time","mean"],"values":[["2000-01-01T00:00:00Z",15]]},{"name":"wgroup","tags":{"region":"us-west"},"columns":["time","mean"],"values":[["2000-01-01T00:00:00Z",30]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_WildcardExpansion(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`wildcard,region=us-east,host=A value=10,cpu=80 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`wildcard,region=us-east,host=B value=20,cpu=90 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`wildcard,region=us-west,host=B value=30,cpu=70 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:20Z").UnixNano()), + fmt.Sprintf(`wildcard,region=us-east,host=A value=40,cpu=60 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:30Z").UnixNano()), + + fmt.Sprintf(`dupnames,region=us-east,day=1 value=10,day=3i %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`dupnames,region=us-east,day=2 value=20,day=2i %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:10Z").UnixNano()), + fmt.Sprintf(`dupnames,region=us-west,day=3 value=30,day=1i %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:20Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "wildcard", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT * FROM wildcard`, + exp: `{"results":[{"series":[{"name":"wildcard","columns":["time","cpu","host","region","value"],"values":[["2000-01-01T00:00:00Z",80,"A","us-east",10],["2000-01-01T00:00:10Z",90,"B","us-east",20],["2000-01-01T00:00:20Z",70,"B","us-west",30],["2000-01-01T00:00:30Z",60,"A","us-east",40]]}]}]}`, + }, + &Query{ + name: "no wildcard in select", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT cpu, host, region, value FROM wildcard`, + exp: `{"results":[{"series":[{"name":"wildcard","columns":["time","cpu","host","region","value"],"values":[["2000-01-01T00:00:00Z",80,"A","us-east",10],["2000-01-01T00:00:10Z",90,"B","us-east",20],["2000-01-01T00:00:20Z",70,"B","us-west",30],["2000-01-01T00:00:30Z",60,"A","us-east",40]]}]}]}`, + }, + &Query{ + name: "no wildcard in select, preserve column order", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT host, cpu, region, value FROM wildcard`, + exp: `{"results":[{"series":[{"name":"wildcard","columns":["time","host","cpu","region","value"],"values":[["2000-01-01T00:00:00Z","A",80,"us-east",10],["2000-01-01T00:00:10Z","B",90,"us-east",20],["2000-01-01T00:00:20Z","B",70,"us-west",30],["2000-01-01T00:00:30Z","A",60,"us-east",40]]}]}]}`, + }, + + &Query{ + name: "only tags, no fields", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT host, region FROM wildcard`, + exp: `{"results":[{"error":"statement must have at least one field in select clause"}]}`, + }, + + &Query{ + name: "no wildcard with alias", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT cpu as c, host as h, region, value FROM wildcard`, + exp: `{"results":[{"series":[{"name":"wildcard","columns":["time","c","h","region","value"],"values":[["2000-01-01T00:00:00Z",80,"A","us-east",10],["2000-01-01T00:00:10Z",90,"B","us-east",20],["2000-01-01T00:00:20Z",70,"B","us-west",30],["2000-01-01T00:00:30Z",60,"A","us-east",40]]}]}]}`, + }, + &Query{ + name: "duplicate tag and field name, always favor field over tag", + command: `SELECT * FROM dupnames`, + params: url.Values{"db": []string{"db0"}}, + exp: `{"results":[{"series":[{"name":"dupnames","columns":["time","day","region","value"],"values":[["2000-01-01T00:00:00Z",3,"us-east",10],["2000-01-01T00:00:10Z",2,"us-east",20],["2000-01-01T00:00:20Z",1,"us-west",30]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_AcrossShardsAndFields(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu load=100 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`cpu load=200 %d`, mustParseTime(time.RFC3339Nano, "2010-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`cpu core=4 %d`, mustParseTime(time.RFC3339Nano, "2015-01-01T00:00:00Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "two results for cpu", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT load FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","load"],"values":[["2000-01-01T00:00:00Z",100],["2010-01-01T00:00:00Z",200]]}]}]}`, + }, + &Query{ + name: "two results for cpu, multi-select", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT core,load FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core","load"],"values":[["2000-01-01T00:00:00Z",null,100],["2010-01-01T00:00:00Z",null,200],["2015-01-01T00:00:00Z",4,null]]}]}]}`, + }, + &Query{ + name: "two results for cpu, wildcard select", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT * FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core","load"],"values":[["2000-01-01T00:00:00Z",null,100],["2010-01-01T00:00:00Z",null,200],["2015-01-01T00:00:00Z",4,null]]}]}]}`, + }, + &Query{ + name: "one result for core", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT core FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core"],"values":[["2015-01-01T00:00:00Z",4]]}]}]}`, + }, + &Query{ + name: "empty result set from non-existent field", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT foo FROM cpu`, + exp: `{"results":[{}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_Where_Fields(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu alert_id="alert",tenant_id="tenant",_cust="johnson brothers" %d`, mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + fmt.Sprintf(`cpu alert_id="alert",tenant_id="tenant",_cust="johnson brothers" %d`, mustParseTime(time.RFC3339Nano, "2015-02-28T01:03:36.703820946Z").UnixNano()), + + fmt.Sprintf(`cpu load=100.0,core=4 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:02Z").UnixNano()), + fmt.Sprintf(`cpu load=80.0,core=2 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:01:02Z").UnixNano()), + + fmt.Sprintf(`clicks local=true %d`, mustParseTime(time.RFC3339Nano, "2014-11-10T23:00:01Z").UnixNano()), + fmt.Sprintf(`clicks local=false %d`, mustParseTime(time.RFC3339Nano, "2014-11-10T23:00:02Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + // non type specific + &Query{ + name: "missing measurement with group by", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT load from missing group by *`, + exp: `{"results":[{}]}`, + }, + + // string + &Query{ + name: "single string field", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT alert_id FROM cpu WHERE alert_id='alert'`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","alert_id"],"values":[["2015-02-28T01:03:36.703820946Z","alert"]]}]}]}`, + }, + &Query{ + name: "string AND query, all fields in SELECT", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT alert_id,tenant_id,_cust FROM cpu WHERE alert_id='alert' AND tenant_id='tenant'`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","alert_id","tenant_id","_cust"],"values":[["2015-02-28T01:03:36.703820946Z","alert","tenant","johnson brothers"]]}]}]}`, + }, + &Query{ + name: "string AND query, all fields in SELECT, one in parenthesis", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT alert_id,tenant_id FROM cpu WHERE alert_id='alert' AND (tenant_id='tenant')`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","alert_id","tenant_id"],"values":[["2015-02-28T01:03:36.703820946Z","alert","tenant"]]}]}]}`, + }, + &Query{ + name: "string underscored field", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT alert_id FROM cpu WHERE _cust='johnson brothers'`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","alert_id"],"values":[["2015-02-28T01:03:36.703820946Z","alert"]]}]}]}`, + }, + &Query{ + name: "string no match", + params: url.Values{"db": []string{"db0"}}, + command: `SELECT alert_id FROM cpu WHERE _cust='acme'`, + exp: `{"results":[{}]}`, + }, + + // float64 + &Query{ + name: "float64 GT no match", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load > 100`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "float64 GTE match one", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load >= 100`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","load"],"values":[["2009-11-10T23:00:02Z",100]]}]}]}`, + }, + &Query{ + name: "float64 EQ match upper bound", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load = 100`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","load"],"values":[["2009-11-10T23:00:02Z",100]]}]}]}`, + }, + &Query{ + name: "float64 LTE match two", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load <= 100`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","load"],"values":[["2009-11-10T23:00:02Z",100],["2009-11-10T23:01:02Z",80]]}]}]}`, + }, + &Query{ + name: "float64 GT match one", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load > 99`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","load"],"values":[["2009-11-10T23:00:02Z",100]]}]}]}`, + }, + &Query{ + name: "float64 EQ no match", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load = 99`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "float64 LT match one", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load < 99`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","load"],"values":[["2009-11-10T23:01:02Z",80]]}]}]}`, + }, + &Query{ + name: "float64 LT no match", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load < 80`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "float64 NE match one", + params: url.Values{"db": []string{"db0"}}, + command: `select load from cpu where load != 100`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","load"],"values":[["2009-11-10T23:01:02Z",80]]}]}]}`, + }, + + // int64 + &Query{ + name: "int64 GT no match", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core > 4`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "int64 GTE match one", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core >= 4`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core"],"values":[["2009-11-10T23:00:02Z",4]]}]}]}`, + }, + &Query{ + name: "int64 EQ match upper bound", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core = 4`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core"],"values":[["2009-11-10T23:00:02Z",4]]}]}]}`, + }, + &Query{ + name: "int64 LTE match two ", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core <= 4`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core"],"values":[["2009-11-10T23:00:02Z",4],["2009-11-10T23:01:02Z",2]]}]}]}`, + }, + &Query{ + name: "int64 GT match one", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core > 3`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core"],"values":[["2009-11-10T23:00:02Z",4]]}]}]}`, + }, + &Query{ + name: "int64 EQ no match", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core = 3`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "int64 LT match one", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core < 3`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core"],"values":[["2009-11-10T23:01:02Z",2]]}]}]}`, + }, + &Query{ + name: "int64 LT no match", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core < 2`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: "int64 NE match one", + params: url.Values{"db": []string{"db0"}}, + command: `select core from cpu where core != 4`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","core"],"values":[["2009-11-10T23:01:02Z",2]]}]}]}`, + }, + + // bool + &Query{ + name: "bool EQ match true", + params: url.Values{"db": []string{"db0"}}, + command: `select local from clicks where local = true`, + exp: `{"results":[{"series":[{"name":"clicks","columns":["time","local"],"values":[["2014-11-10T23:00:01Z",true]]}]}]}`, + }, + &Query{ + name: "bool EQ match false", + params: url.Values{"db": []string{"db0"}}, + command: `select local from clicks where local = false`, + exp: `{"results":[{"series":[{"name":"clicks","columns":["time","local"],"values":[["2014-11-10T23:00:02Z",false]]}]}]}`, + }, + + &Query{ + name: "bool NE match one", + params: url.Values{"db": []string{"db0"}}, + command: `select local from clicks where local != true`, + exp: `{"results":[{"series":[{"name":"clicks","columns":["time","local"],"values":[["2014-11-10T23:00:02Z",false]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_Where_With_Tags(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`where_events,tennant=paul foo="bar" %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:02Z").UnixNano()), + fmt.Sprintf(`where_events,tennant=paul foo="baz" %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:03Z").UnixNano()), + fmt.Sprintf(`where_events,tennant=paul foo="bat" %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:04Z").UnixNano()), + fmt.Sprintf(`where_events,tennant=todd foo="bar" %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:05Z").UnixNano()), + fmt.Sprintf(`where_events,tennant=david foo="bap" %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:06Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "tag field and time", + params: url.Values{"db": []string{"db0"}}, + command: `select foo from where_events where (tennant = 'paul' OR tennant = 'david') AND time > 1s AND (foo = 'bar' OR foo = 'baz' OR foo = 'bap')`, + exp: `{"results":[{"series":[{"name":"where_events","columns":["time","foo"],"values":[["2009-11-10T23:00:02Z","bar"],["2009-11-10T23:00:03Z","baz"],["2009-11-10T23:00:06Z","bap"]]}]}]}`, + }, + &Query{ + name: "where on tag that should be double quoted but isn't", + params: url.Values{"db": []string{"db0"}}, + command: `show series where data-center = 'foo'`, + exp: `{"results":[{"error":"invalid expression: data - center = 'foo'"}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_LimitAndOffset(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`limited,tennant=paul foo=2 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:02Z").UnixNano()), + fmt.Sprintf(`limited,tennant=paul foo=3 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:03Z").UnixNano()), + fmt.Sprintf(`limited,tennant=paul foo=4 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:04Z").UnixNano()), + fmt.Sprintf(`limited,tennant=todd foo=5 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:05Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "limit on points", + params: url.Values{"db": []string{"db0"}}, + command: `select foo from "limited" LIMIT 2`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","foo"],"values":[["2009-11-10T23:00:02Z",2],["2009-11-10T23:00:03Z",3]]}]}]}`, + }, + &Query{ + name: "limit higher than the number of data points", + params: url.Values{"db": []string{"db0"}}, + command: `select foo from "limited" LIMIT 20`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","foo"],"values":[["2009-11-10T23:00:02Z",2],["2009-11-10T23:00:03Z",3],["2009-11-10T23:00:04Z",4],["2009-11-10T23:00:05Z",5]]}]}]}`, + }, + &Query{ + name: "limit and offset", + params: url.Values{"db": []string{"db0"}}, + command: `select foo from "limited" LIMIT 2 OFFSET 1`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","foo"],"values":[["2009-11-10T23:00:03Z",3],["2009-11-10T23:00:04Z",4]]}]}]}`, + }, + &Query{ + name: "limit + offset equal to total number of points", + params: url.Values{"db": []string{"db0"}}, + command: `select foo from "limited" LIMIT 3 OFFSET 3`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","foo"],"values":[["2009-11-10T23:00:05Z",5]]}]}]}`, + }, + &Query{ + name: "limit - offset higher than number of points", + command: `select foo from "limited" LIMIT 2 OFFSET 20`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","foo"]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "limit on points with group by time", + command: `select mean(foo) from "limited" WHERE time >= '2009-11-10T23:00:02Z' AND time < '2009-11-10T23:00:06Z' GROUP BY TIME(1s) LIMIT 2`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","mean"],"values":[["2009-11-10T23:00:02Z",2],["2009-11-10T23:00:03Z",3]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "limit higher than the number of data points with group by time", + command: `select mean(foo) from "limited" WHERE time >= '2009-11-10T23:00:02Z' AND time < '2009-11-10T23:00:06Z' GROUP BY TIME(1s) LIMIT 20`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","mean"],"values":[["2009-11-10T23:00:02Z",2],["2009-11-10T23:00:03Z",3],["2009-11-10T23:00:04Z",4],["2009-11-10T23:00:05Z",5]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "limit and offset with group by time", + command: `select mean(foo) from "limited" WHERE time >= '2009-11-10T23:00:02Z' AND time < '2009-11-10T23:00:06Z' GROUP BY TIME(1s) LIMIT 2 OFFSET 1`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","mean"],"values":[["2009-11-10T23:00:03Z",3],["2009-11-10T23:00:04Z",4]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "limit + offset equal to the number of points with group by time", + command: `select mean(foo) from "limited" WHERE time >= '2009-11-10T23:00:02Z' AND time < '2009-11-10T23:00:06Z' GROUP BY TIME(1s) LIMIT 3 OFFSET 3`, + exp: `{"results":[{"series":[{"name":"limited","columns":["time","mean"],"values":[["2009-11-10T23:00:05Z",5]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "limit - offset higher than number of points with group by time", + command: `select mean(foo) from "limited" WHERE time >= '2009-11-10T23:00:02Z' AND time < '2009-11-10T23:00:06Z' GROUP BY TIME(1s) LIMIT 2 OFFSET 20`, + exp: `{"results":[{}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "limit higher than the number of data points should error", + command: `select mean(foo) from "limited" where time > '2000-01-01T00:00:00Z' group by time(1s), * fill(0) limit 2147483647`, + exp: `{"results":[{"error":"too many points in the group by interval. maybe you forgot to specify a where time clause?"}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "limit1 higher than MaxGroupBy but the number of data points is less than MaxGroupBy", + command: `select mean(foo) from "limited" where time >= '2009-11-10T23:00:02Z' and time < '2009-11-10T23:00:03Z' group by time(1s), * fill(0) limit 2147483647`, + exp: `{"results":[{"series":[{"name":"limited","tags":{"tennant":"paul"},"columns":["time","mean"],"values":[["2009-11-10T23:00:02Z",2]]},{"name":"limited","tags":{"tennant":"todd"},"columns":["time","mean"],"values":[["2009-11-10T23:00:02Z",0]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_Fill(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`fills val=3 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:02Z").UnixNano()), + fmt.Sprintf(`fills val=5 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:03Z").UnixNano()), + fmt.Sprintf(`fills val=4 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:06Z").UnixNano()), + fmt.Sprintf(`fills val=10 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:16Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "fill with value", + command: `select mean(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' group by time(5s) FILL(1)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","mean"],"values":[["2009-11-10T23:00:00Z",4],["2009-11-10T23:00:05Z",4],["2009-11-10T23:00:10Z",1],["2009-11-10T23:00:15Z",10]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill with value, WHERE all values match condition", + command: `select mean(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' and val < 50 group by time(5s) FILL(1)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","mean"],"values":[["2009-11-10T23:00:00Z",4],["2009-11-10T23:00:05Z",4],["2009-11-10T23:00:10Z",1],["2009-11-10T23:00:15Z",10]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill with value, WHERE no values match condition", + command: `select mean(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' and val > 50 group by time(5s) FILL(1)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","mean"],"values":[["2009-11-10T23:00:00Z",1],["2009-11-10T23:00:05Z",1],["2009-11-10T23:00:10Z",1],["2009-11-10T23:00:15Z",1]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill with previous", + command: `select mean(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' group by time(5s) FILL(previous)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","mean"],"values":[["2009-11-10T23:00:00Z",4],["2009-11-10T23:00:05Z",4],["2009-11-10T23:00:10Z",4],["2009-11-10T23:00:15Z",10]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill with none, i.e. clear out nulls", + command: `select mean(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' group by time(5s) FILL(none)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","mean"],"values":[["2009-11-10T23:00:00Z",4],["2009-11-10T23:00:05Z",4],["2009-11-10T23:00:15Z",10]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill defaults to null", + command: `select mean(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' group by time(5s)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","mean"],"values":[["2009-11-10T23:00:00Z",4],["2009-11-10T23:00:05Z",4],["2009-11-10T23:00:10Z",null],["2009-11-10T23:00:15Z",10]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill with count aggregate defaults to null", + command: `select count(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' group by time(5s)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","count"],"values":[["2009-11-10T23:00:00Z",2],["2009-11-10T23:00:05Z",1],["2009-11-10T23:00:10Z",null],["2009-11-10T23:00:15Z",1]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill with count aggregate defaults to null, no values match", + command: `select count(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' and val > 100 group by time(5s)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","count"],"values":[["2009-11-10T23:00:00Z",null],["2009-11-10T23:00:05Z",null],["2009-11-10T23:00:10Z",null],["2009-11-10T23:00:15Z",null]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "fill with count aggregate specific value", + command: `select count(val) from fills where time >= '2009-11-10T23:00:00Z' and time < '2009-11-10T23:00:20Z' group by time(5s) fill(1234)`, + exp: `{"results":[{"series":[{"name":"fills","columns":["time","count"],"values":[["2009-11-10T23:00:00Z",2],["2009-11-10T23:00:05Z",1],["2009-11-10T23:00:10Z",1234],["2009-11-10T23:00:15Z",1]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_Chunk(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := make([]string, 10001) // 10,000 is the default chunking size, even when no chunking requested. + expectedValues := make([]string, len(writes)) + for i := 0; i < len(writes); i++ { + writes[i] = fmt.Sprintf(`cpu value=%d %d`, i, time.Unix(0, int64(i)).UnixNano()) + expectedValues[i] = fmt.Sprintf(`["%s",%d]`, time.Unix(0, int64(i)).UTC().Format(time.RFC3339Nano), i) + } + expected := fmt.Sprintf(`{"results":[{"series":[{"name":"cpu","columns":["time","value"],"values":[%s]}]}]}`, strings.Join(expectedValues, ",")) + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "SELECT all values, no chunking", + command: `SELECT value FROM cpu`, + exp: expected, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } + +} + +func TestServer_Query_DropAndRecreateMeasurement(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=serverA,region=uswest val=23.2 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf(`memory,host=serverB,region=uswest val=33.2 %d`, mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:01Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "Drop Measurement, series tags preserved tests", + command: `SHOW MEASUREMENTS`, + exp: `{"results":[{"series":[{"name":"measurements","columns":["name"],"values":[["cpu"],["memory"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "show series", + command: `SHOW SERIES`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=serverA,region=uswest","serverA","uswest"]]},{"name":"memory","columns":["_key","host","region"],"values":[["memory,host=serverB,region=uswest","serverB","uswest"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "ensure we can query for memory with both tags", + command: `SELECT * FROM memory where region='uswest' and host='serverB' GROUP BY *`, + exp: `{"results":[{"series":[{"name":"memory","tags":{"host":"serverB","region":"uswest"},"columns":["time","val"],"values":[["2000-01-01T00:00:01Z",33.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "drop measurement cpu", + command: `DROP MEASUREMENT cpu`, + exp: `{"results":[{}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "verify measurements", + command: `SHOW MEASUREMENTS`, + exp: `{"results":[{"series":[{"name":"measurements","columns":["name"],"values":[["memory"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "verify series", + command: `SHOW SERIES`, + exp: `{"results":[{"series":[{"name":"memory","columns":["_key","host","region"],"values":[["memory,host=serverB,region=uswest","serverB","uswest"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "verify cpu measurement is gone", + command: `SELECT * FROM cpu`, + exp: `{"results":[{}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "verify selecting from a tag 'host' still works", + command: `SELECT * FROM memory where host='serverB' GROUP BY *`, + exp: `{"results":[{"series":[{"name":"memory","tags":{"host":"serverB","region":"uswest"},"columns":["time","val"],"values":[["2000-01-01T00:00:01Z",33.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "verify selecting from a tag 'region' still works", + command: `SELECT * FROM memory where region='uswest' GROUP BY *`, + exp: `{"results":[{"series":[{"name":"memory","tags":{"host":"serverB","region":"uswest"},"columns":["time","val"],"values":[["2000-01-01T00:00:01Z",33.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "verify selecting from a tag 'host' and 'region' still works", + command: `SELECT * FROM memory where region='uswest' and host='serverB' GROUP BY *`, + exp: `{"results":[{"series":[{"name":"memory","tags":{"host":"serverB","region":"uswest"},"columns":["time","val"],"values":[["2000-01-01T00:00:01Z",33.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "Drop non-existant measurement", + command: `DROP MEASUREMENT doesntexist`, + exp: `{"results":[{"error":"measurement not found: doesntexist"}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + // Test that re-inserting the measurement works fine. + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } + + test = NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "verify measurements after recreation", + command: `SHOW MEASUREMENTS`, + exp: `{"results":[{"series":[{"name":"measurements","columns":["name"],"values":[["cpu"],["memory"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "verify cpu measurement has been re-inserted", + command: `SELECT * FROM cpu GROUP BY *`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"serverA","region":"uswest"},"columns":["time","val"],"values":[["2000-01-01T00:00:00Z",23.2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_ShowSeries(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=server01 value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:01Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=uswest value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:02Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:03Z").UnixNano()), + fmt.Sprintf(`cpu,host=server02,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:04Z").UnixNano()), + fmt.Sprintf(`gpu,host=server02,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:05Z").UnixNano()), + fmt.Sprintf(`gpu,host=server03,region=caeast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:06Z").UnixNano()), + fmt.Sprintf(`disk,host=server03,region=caeast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:07Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: `show series`, + command: "SHOW SERIES", + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=server01","server01",""],["cpu,host=server01,region=uswest","server01","uswest"],["cpu,host=server01,region=useast","server01","useast"],["cpu,host=server02,region=useast","server02","useast"]]},{"name":"disk","columns":["_key","host","region"],"values":[["disk,host=server03,region=caeast","server03","caeast"]]},{"name":"gpu","columns":["_key","host","region"],"values":[["gpu,host=server02,region=useast","server02","useast"],["gpu,host=server03,region=caeast","server03","caeast"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show series from measurement`, + command: "SHOW SERIES FROM cpu", + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=server01","server01",""],["cpu,host=server01,region=uswest","server01","uswest"],["cpu,host=server01,region=useast","server01","useast"],["cpu,host=server02,region=useast","server02","useast"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show series from regular expression`, + command: "SHOW SERIES FROM /[cg]pu/", + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=server01","server01",""],["cpu,host=server01,region=uswest","server01","uswest"],["cpu,host=server01,region=useast","server01","useast"],["cpu,host=server02,region=useast","server02","useast"]]},{"name":"gpu","columns":["_key","host","region"],"values":[["gpu,host=server02,region=useast","server02","useast"],["gpu,host=server03,region=caeast","server03","caeast"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show series with where tag`, + command: "SHOW SERIES WHERE region = 'uswest'", + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=server01,region=uswest","server01","uswest"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show series where tag matches regular expression`, + command: "SHOW SERIES WHERE region =~ /ca.*/", + exp: `{"results":[{"series":[{"name":"disk","columns":["_key","host","region"],"values":[["disk,host=server03,region=caeast","server03","caeast"]]},{"name":"gpu","columns":["_key","host","region"],"values":[["gpu,host=server03,region=caeast","server03","caeast"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show series`, + command: "SHOW SERIES WHERE host !~ /server0[12]/", + exp: `{"results":[{"series":[{"name":"disk","columns":["_key","host","region"],"values":[["disk,host=server03,region=caeast","server03","caeast"]]},{"name":"gpu","columns":["_key","host","region"],"values":[["gpu,host=server03,region=caeast","server03","caeast"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show series with from and where`, + command: "SHOW SERIES FROM cpu WHERE region = 'useast'", + exp: `{"results":[{"series":[{"name":"cpu","columns":["_key","host","region"],"values":[["cpu,host=server01,region=useast","server01","useast"],["cpu,host=server02,region=useast","server02","useast"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_ShowMeasurements(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=server01 value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=uswest value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server02,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`gpu,host=server02,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`gpu,host=server02,region=caeast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`other,host=server03,region=caeast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: `show measurements with limit 2`, + command: "SHOW MEASUREMENTS LIMIT 2", + exp: `{"results":[{"series":[{"name":"measurements","columns":["name"],"values":[["cpu"],["gpu"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show measurements where tag matches regular expression`, + command: "SHOW MEASUREMENTS WHERE region =~ /ca.*/", + exp: `{"results":[{"series":[{"name":"measurements","columns":["name"],"values":[["gpu"],["other"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show measurements where tag does not match a regular expression`, + command: "SHOW MEASUREMENTS WHERE region !~ /ca.*/", + exp: `{"results":[{"series":[{"name":"measurements","columns":["name"],"values":[["cpu"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_ShowTagKeys(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=server01 value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=uswest value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server02,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`gpu,host=server02,region=useast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`gpu,host=server03,region=caeast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`disk,host=server03,region=caeast value=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: `show tag keys`, + command: "SHOW TAG KEYS", + exp: `{"results":[{"series":[{"name":"cpu","columns":["tagKey"],"values":[["host"],["region"]]},{"name":"disk","columns":["tagKey"],"values":[["host"],["region"]]},{"name":"gpu","columns":["tagKey"],"values":[["host"],["region"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "show tag keys from", + command: "SHOW TAG KEYS FROM cpu", + exp: `{"results":[{"series":[{"name":"cpu","columns":["tagKey"],"values":[["host"],["region"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "show tag keys from regex", + command: "SHOW TAG KEYS FROM /[cg]pu/", + exp: `{"results":[{"series":[{"name":"cpu","columns":["tagKey"],"values":[["host"],["region"]]},{"name":"gpu","columns":["tagKey"],"values":[["host"],["region"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "show tag keys measurement not found", + command: "SHOW TAG KEYS FROM bad", + exp: `{"results":[{"error":"measurement not found: bad"}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: "show tag values with key", + command: "SHOW TAG VALUES WITH KEY = host", + exp: `{"results":[{"series":[{"name":"hostTagValues","columns":["host"],"values":[["server01"],["server02"],["server03"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show tag values with key and where`, + command: `SHOW TAG VALUES FROM cpu WITH KEY = host WHERE region = 'uswest'`, + exp: `{"results":[{"series":[{"name":"hostTagValues","columns":["host"],"values":[["server01"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show tag values with key and where matches regular expression`, + command: `SHOW TAG VALUES WITH KEY = host WHERE region =~ /ca.*/`, + exp: `{"results":[{"series":[{"name":"hostTagValues","columns":["host"],"values":[["server03"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show tag values with key and where does not matche regular expression`, + command: `SHOW TAG VALUES WITH KEY = region WHERE host !~ /server0[12]/`, + exp: `{"results":[{"series":[{"name":"regionTagValues","columns":["region"],"values":[["caeast"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show tag values with key in and where does not matche regular expression`, + command: `SHOW TAG VALUES FROM cpu WITH KEY IN (host, region) WHERE region = 'uswest'`, + exp: `{"results":[{"series":[{"name":"hostTagValues","columns":["host"],"values":[["server01"]]},{"name":"regionTagValues","columns":["region"],"values":[["uswest"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show tag values with key and measurement matches regular expression`, + command: `SHOW TAG VALUES FROM /[cg]pu/ WITH KEY = host`, + exp: `{"results":[{"series":[{"name":"hostTagValues","columns":["host"],"values":[["server01"],["server02"],["server03"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_ShowFieldKeys(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + writes := []string{ + fmt.Sprintf(`cpu,host=server01 field1=100 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=uswest field1=200,field2=300,field3=400 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server01,region=useast field1=200,field2=300,field3=400 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`cpu,host=server02,region=useast field1=200,field2=300,field3=400 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`gpu,host=server01,region=useast field4=200,field5=300 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`gpu,host=server03,region=caeast field6=200,field7=300 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + fmt.Sprintf(`disk,host=server03,region=caeast field8=200,field9=300 %d`, mustParseTime(time.RFC3339Nano, "2009-11-10T23:00:00Z").UnixNano()), + } + + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: `show field keys`, + command: `SHOW FIELD KEYS`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["fieldKey"],"values":[["field1"],["field2"],["field3"]]},{"name":"disk","columns":["fieldKey"],"values":[["field8"],["field9"]]},{"name":"gpu","columns":["fieldKey"],"values":[["field4"],["field5"],["field6"],["field7"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show field keys from measurement`, + command: `SHOW FIELD KEYS FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["fieldKey"],"values":[["field1"],["field2"],["field3"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + &Query{ + name: `show field keys measurement with regex`, + command: `SHOW FIELD KEYS FROM /[cg]pu/`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["fieldKey"],"values":[["field1"],["field2"],["field3"]]},{"name":"gpu","columns":["fieldKey"],"values":[["field4"],["field5"],["field6"],["field7"]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +func TestServer_Query_CreateContinuousQuery(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + test := NewTest("db0", "rp0") + + test.addQueries([]*Query{ + &Query{ + name: "create continuous query", + command: `CREATE CONTINUOUS QUERY "my.query" ON db0 BEGIN SELECT count(value) INTO measure1 FROM myseries GROUP BY time(10m) END`, + exp: `{"results":[{}]}`, + }, + &Query{ + name: `show continuous queries`, + command: `SHOW CONTINUOUS QUERIES`, + exp: `{"results":[{"series":[{"name":"db0","columns":["name","query"],"values":[["my.query","CREATE CONTINUOUS QUERY \"my.query\" ON db0 BEGIN SELECT count(value) INTO \"db0\".\"rp0\".measure1 FROM \"db0\".\"rp0\".myseries GROUP BY time(10m) END"]]}]}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} + +// Tests that a known CQ query with concurrent writes does not deadlock the server +func TestServer_ContinuousQuery_Deadlock(t *testing.T) { + + // Skip until #3517 & #3522 are merged + t.Skip("Skipping CQ deadlock test") + if testing.Short() { + t.Skip("skipping CQ deadlock test") + } + t.Parallel() + s := OpenServer(NewConfig(), "") + defer func() { + s.Close() + // Nil the server so our deadlock detector goroutine can determine if we completed writes + // without timing out + s.Server = nil + }() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + test := NewTest("db0", "rp0") + + test.addQueries([]*Query{ + &Query{ + name: "create continuous query", + command: `CREATE CONTINUOUS QUERY "my.query" ON db0 BEGIN SELECT sum(visits) as visits INTO test_1m FROM myseries GROUP BY time(1m), host END`, + exp: `{"results":[{}]}`, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } + + // Deadlock detector. If the deadlock is fixed, this test should complete all the writes in ~2.5s seconds (with artifical delays + // added). After 10 seconds, if the server has not been closed then we hit the deadlock bug. + iterations := 0 + go func(s *Server) { + <-time.After(10 * time.Second) + + // If the server is not nil then the test is still running and stuck. We panic to avoid + // having the whole test suite hang indefinitely. + if s.Server != nil { + panic("possible deadlock. writes did not complete in time") + } + }(s) + + for { + + // After the second write, if the deadlock exists, we'll get a write timeout and + // all subsequent writes will timeout + if iterations > 5 { + break + } + writes := []string{} + for i := 0; i < 1000; i++ { + writes = append(writes, fmt.Sprintf(`myseries,host=host-%d visits=1i`, i)) + } + write := strings.Join(writes, "\n") + + if _, err := s.Write(test.db, test.rp, write, test.params); err != nil { + t.Fatal(err) + } + iterations += 1 + time.Sleep(500 * time.Millisecond) + } +} + +func TestServer_Query_EvilIdentifiers(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 0)); err != nil { + t.Fatal(err) + } + if err := s.MetaStore.SetDefaultRetentionPolicy("db0", "rp0"); err != nil { + t.Fatal(err) + } + + test := NewTest("db0", "rp0") + test.write = fmt.Sprintf("cpu select=1,in-bytes=2 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()) + + test.addQueries([]*Query{ + &Query{ + name: `query evil identifiers`, + command: `SELECT "select", "in-bytes" FROM cpu`, + exp: `{"results":[{"series":[{"name":"cpu","columns":["time","select","in-bytes"],"values":[["2000-01-01T00:00:00Z",1,2]]}]}]}`, + params: url.Values{"db": []string{"db0"}}, + }, + }...) + + for i, query := range test.queries { + if i == 0 { + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + } + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_test.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_test.md new file mode 100644 index 000000000..2b6883de7 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/cmd/influxd/run/server_test.md @@ -0,0 +1,150 @@ +# Server Integration Tests + +Currently, the file `server_test.go` has integration tests for single node scenarios. +At some point we'll need to add cluster tests, and may add them in a different file, or +rename `server_test.go` to `server_single_node_test.go` or something like that. + +## What is in a test? + +Each test is broken apart effectively into the following areas: + +- Write sample data +- Use cases for table driven test, that include a command (typically a query) and an expected result. + +When each test runs it does the following: + +- init: determines if there are any writes and if so, writes them to the in-memory database +- queries: iterate through each query, executing the command, and comparing the results to the expected result. + +## Idempotent - Allows for parallel tests + +Each test should be `idempotent`, meaining that its data will not be affected by other tests, or use cases within the table tests themselves. +This allows for parallel testing, keeping the test suite total execution time very low. + +### Basic sample test + +```go +// Ensure the server can have a database with multiple measurements. +func TestServer_Query_Multiple_Measurements(t *testing.T) { + t.Parallel() + s := OpenServer(NewConfig(), "") + defer s.Close() + + if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) + } + + // Make sure we do writes for measurements that will span across shards + writes := []string{ + fmt.Sprintf("cpu,host=server01 value=100,core=4 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf("cpu1,host=server02 value=50,core=2 %d", mustParseTime(time.RFC3339Nano, "2015-01-01T00:00:00Z").UnixNano()), + } + test := NewTest("db0", "rp0") + test.write = strings.Join(writes, "\n") + + test.addQueries([]*Query{ + &Query{ + name: "measurement in one shard but not another shouldn't panic server", + command: `SELECT host,value FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["2000-01-01T00:00:00Z",100]]}]}]}`, + }, + }...) + + if err := test.init(s); err != nil { + t.Fatalf("test init failed: %s", err) + } + + for _, query := range test.queries { + if query.skip { + t.Logf("SKIP:: %s", query.name) + continue + } + if err := query.Execute(s); err != nil { + t.Error(query.Error(err)) + } else if !query.success() { + t.Error(query.failureMessage()) + } + } +} +``` + +Let's break this down: + +In this test, we first tell it to run in parallel with the `t.Parallel()` call. + +We then open a new server with: + +```go +s := OpenServer(NewConfig(), "") +defer s.Close() +``` + +If needed, we create a database and default retention policy. This is usually needed +when inserting and querying data. This is not needed if you are testing commands like `CREATE DATABASE`, `SHOW DIAGNOSTICS`, etc. + +```go +if err := s.CreateDatabaseAndRetentionPolicy("db0", newRetentionPolicyInfo("rp0", 1, 1*time.Hour)); err != nil { + t.Fatal(err) +} +``` + +Next, set up the write data you need: + +```go +writes := []string{ + fmt.Sprintf("cpu,host=server01 value=100,core=4 %d", mustParseTime(time.RFC3339Nano, "2000-01-01T00:00:00Z").UnixNano()), + fmt.Sprintf("cpu1,host=server02 value=50,core=2 %d", mustParseTime(time.RFC3339Nano, "2015-01-01T00:00:00Z").UnixNano()), +} +``` +Create a new test with the database and retention policy: + +```go +test := NewTest("db0", "rp0") +``` + +Send in the writes: +```go +test.write = strings.Join(writes, "\n") +``` + +Add some queries (the second one is mocked out to show how to add more than one): + +```go +test.addQueries([]*Query{ + &Query{ + name: "measurement in one shard but not another shouldn't panic server", + command: `SELECT host,value FROM db0.rp0.cpu`, + exp: `{"results":[{"series":[{"name":"cpu","tags":{"host":"server01"},"columns":["time","value"],"values":[["2000-01-01T00:00:00Z",100]]}]}]}`, + }, + &Query{ + name: "another test here...", + command: `Some query command`, + exp: `the expected results`, + }, +}...) +``` + +The rest of the code is boilerplate execution code. It is purposefully not refactored out to a helper +to make sure the test failure reports the proper lines for debugging purposes. + +#### Running the tests + +To run the tests: + +```sh +go test ./cmd/influxd/run -parallel 500 -timeout 10s +``` + +#### Running a specific test + +```sh +go test ./cmd/influxd/run -parallel 500 -timeout 10s -run TestServer_Query_Fill +``` + +#### Verbose feedback + +By default, all logs are silenced when testing. If you pass in the `-v` flag, the test suite becomes verbose, and enables all logging in the system + +```sh +go test ./cmd/influxd/run -parallel 500 -timeout 10s -run TestServer_Query_Fill -v +``` diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/diagnostics.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/diagnostics.go new file mode 100644 index 000000000..c1e02565b --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/diagnostics.go @@ -0,0 +1,143 @@ +package influxdb + +import ( + "os" + "runtime" + "time" + + "github.com/influxdb/influxdb/influxql" +) + +// GoDiagnostics captures basic information about the runtime. +type GoDiagnostics struct { + GoMaxProcs int + NumGoroutine int + Version string +} + +// NewGoDiagnostics returns a GoDiagnostics object. +func NewGoDiagnostics() *GoDiagnostics { + return &GoDiagnostics{ + GoMaxProcs: runtime.GOMAXPROCS(0), + NumGoroutine: runtime.NumGoroutine(), + Version: runtime.Version(), + } +} + +// AsRow returns the GoDiagnostic object as an InfluxQL row. +func (g *GoDiagnostics) AsRow(measurement string, tags map[string]string) *influxql.Row { + return &influxql.Row{ + Name: measurement, + Columns: []string{"time", "goMaxProcs", "numGoRoutine", "version"}, + Tags: tags, + Values: [][]interface{}{[]interface{}{time.Now().UTC(), + g.GoMaxProcs, g.NumGoroutine, g.Version}}, + } +} + +// SystemDiagnostics captures basic machine data. +type SystemDiagnostics struct { + Hostname string + PID int + OS string + Arch string + NumCPU int +} + +// NewSystemDiagnostics returns a SystemDiagnostics object. +func NewSystemDiagnostics() *SystemDiagnostics { + hostname, err := os.Hostname() + if err != nil { + hostname = "unknown" + } + + return &SystemDiagnostics{ + Hostname: hostname, + PID: os.Getpid(), + OS: runtime.GOOS, + Arch: runtime.GOARCH, + NumCPU: runtime.NumCPU(), + } +} + +// AsRow returns the GoDiagnostic object as an InfluxQL row. +func (s *SystemDiagnostics) AsRow(measurement string, tags map[string]string) *influxql.Row { + return &influxql.Row{ + Name: measurement, + Columns: []string{"time", "hostname", "pid", "os", "arch", "numCPU"}, + Tags: tags, + Values: [][]interface{}{[]interface{}{time.Now().UTC(), + s.Hostname, s.PID, s.OS, s.Arch, s.NumCPU}}, + } +} + +// MemoryDiagnostics captures Go memory stats. +type MemoryDiagnostics struct { + Alloc int64 + TotalAlloc int64 + Sys int64 + Lookups int64 + Mallocs int64 + Frees int64 + HeapAlloc int64 + HeapSys int64 + HeapIdle int64 + HeapInUse int64 + HeapReleased int64 + HeapObjects int64 + PauseTotalNs int64 + NumGC int64 +} + +// NewMemoryDiagnostics returns a MemoryDiagnostics object. +func NewMemoryDiagnostics() *MemoryDiagnostics { + var m runtime.MemStats + runtime.ReadMemStats(&m) + + return &MemoryDiagnostics{ + Alloc: int64(m.Alloc), + TotalAlloc: int64(m.TotalAlloc), + Sys: int64(m.Sys), + Lookups: int64(m.Lookups), + Mallocs: int64(m.Mallocs), + Frees: int64(m.Frees), + HeapAlloc: int64(m.HeapAlloc), + HeapSys: int64(m.HeapSys), + HeapIdle: int64(m.HeapIdle), + HeapInUse: int64(m.HeapInuse), + HeapReleased: int64(m.HeapReleased), + HeapObjects: int64(m.HeapObjects), + PauseTotalNs: int64(m.PauseTotalNs), + NumGC: int64(m.NumGC), + } +} + +// AsRow returns the MemoryDiagnostics object as an InfluxQL row. +func (m *MemoryDiagnostics) AsRow(measurement string, tags map[string]string) *influxql.Row { + return &influxql.Row{ + Name: measurement, + Columns: []string{"time", "alloc", "totalAlloc", "sys", "lookups", "mallocs", "frees", "heapAlloc", + "heapSys", "heapIdle", "heapInUse", "heapReleased", "heapObjects", "pauseTotalNs", "numGG"}, + Tags: tags, + Values: [][]interface{}{[]interface{}{time.Now().UTC(), + m.Alloc, m.TotalAlloc, m.Sys, m.Lookups, m.Mallocs, m.Frees, m.HeapAlloc, + m.HeapSys, m.HeapIdle, m.HeapInUse, m.HeapReleased, m.HeapObjects, m.PauseTotalNs, m.NumGC}}, + } +} + +// BuildDiagnostics capture basic build version information. +type BuildDiagnostics struct { + Version string + CommitHash string +} + +// AsRow returns the BuildDiagnostics object as an InfluxQL row. +func (b *BuildDiagnostics) AsRow(measurement string, tags map[string]string) *influxql.Row { + return &influxql.Row{ + Name: measurement, + Columns: []string{"time", "version", "commitHash"}, + Tags: tags, + Values: [][]interface{}{[]interface{}{time.Now().UTC(), + b.Version, b.CommitHash}}, + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/errors.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/errors.go new file mode 100644 index 000000000..c18f2c449 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/errors.go @@ -0,0 +1,78 @@ +package influxdb + +import ( + "encoding/json" + "errors" + "fmt" + "runtime" + "strings" +) + +var ( + // ErrFieldsRequired is returned when a point does not any fields. + ErrFieldsRequired = errors.New("fields required") + + // ErrFieldTypeConflict is returned when a new field already exists with a different type. + ErrFieldTypeConflict = errors.New("field type conflict") +) + +func ErrDatabaseNotFound(name string) error { return fmt.Errorf("database not found: %s", name) } + +func ErrMeasurementNotFound(name string) error { return fmt.Errorf("measurement not found: %s", name) } + +func Errorf(format string, a ...interface{}) (err error) { + if _, file, line, ok := runtime.Caller(2); ok { + a = append(a, file, line) + err = fmt.Errorf(format+" (%s:%d)", a...) + } else { + err = fmt.Errorf(format, a...) + } + return +} + +// IsClientError indicates whether an error is a known client error. +func IsClientError(err error) bool { + if err == nil { + return false + } + + if err == ErrFieldsRequired { + return true + } + if err == ErrFieldTypeConflict { + return true + } + + if strings.Contains(err.Error(), ErrFieldTypeConflict.Error()) { + return true + } + + return false +} + +// mustMarshal encodes a value to JSON. +// This will panic if an error occurs. This should only be used internally when +// an invalid marshal will cause corruption and a panic is appropriate. +func mustMarshalJSON(v interface{}) []byte { + b, err := json.Marshal(v) + if err != nil { + panic("marshal: " + err.Error()) + } + return b +} + +// mustUnmarshalJSON decodes a value from JSON. +// This will panic if an error occurs. This should only be used internally when +// an invalid unmarshal will cause corruption and a panic is appropriate. +func mustUnmarshalJSON(b []byte, v interface{}) { + if err := json.Unmarshal(b, v); err != nil { + panic("unmarshal: " + err.Error()) + } +} + +// assert will panic with a given formatted message if the given condition is false. +func assert(condition bool, msg string, v ...interface{}) { + if !condition { + panic(fmt.Sprintf("assert failed: "+msg, v...)) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/.rvmrc b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/.rvmrc new file mode 100644 index 000000000..a9c1a9ca3 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/.rvmrc @@ -0,0 +1 @@ +rvm use ruby-2.1.0@burn-in --create diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/Gemfile b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/Gemfile new file mode 100644 index 000000000..b1816e8b6 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/Gemfile @@ -0,0 +1,4 @@ +source 'https://rubygems.org' + +gem "colorize" +gem "influxdb" diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/Gemfile.lock b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/Gemfile.lock new file mode 100644 index 000000000..9e721c3a7 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/Gemfile.lock @@ -0,0 +1,14 @@ +GEM + remote: https://rubygems.org/ + specs: + colorize (0.6.0) + influxdb (0.0.16) + json + json (1.8.1) + +PLATFORMS + ruby + +DEPENDENCIES + colorize + influxdb diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/burn-in.rb b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/burn-in.rb new file mode 100644 index 000000000..1d44bc2c0 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/burn-in.rb @@ -0,0 +1,79 @@ +require "influxdb" +require "colorize" +require "benchmark" + +require_relative "log" +require_relative "random_gaussian" + +BATCH_SIZE = 10_000 + +Log.info "Starting burn-in suite" +master = InfluxDB::Client.new +master.delete_database("burn-in") rescue nil +master.create_database("burn-in") +master.create_database_user("burn-in", "user", "pass") + +master.database = "burn-in" +# master.query "select * from test1 into test2;" +# master.query "select count(value) from test1 group by time(1m) into test2;" + +influxdb = InfluxDB::Client.new "burn-in", username: "user", password: "pass" + +Log.success "Connected to server #{influxdb.host}:#{influxdb.port}" + +Log.log "Creating RandomGaussian(500, 25)" +gaussian = RandomGaussian.new(500, 25) +point_count = 0 + +while true + Log.log "Generating 10,000 points.." + points = [] + BATCH_SIZE.times do |n| + points << {value: gaussian.rand.to_i.abs} + end + point_count += points.length + + Log.info "Sending points to server.." + begin + st = Time.now + foo = influxdb.write_point("test1", points) + et = Time.now + Log.log foo.inspect + Log.log "#{et-st} seconds elapsed" + Log.success "Write successful." + rescue => e + Log.failure "Write failed:" + Log.log e + end + sleep 0.5 + + Log.info "Checking regular points" + st = Time.now + response = influxdb.query("select count(value) from test1;") + et = Time.now + + Log.log "#{et-st} seconds elapsed" + + response_count = response["test1"].first["count"] + if point_count == response_count + Log.success "Point counts match: #{point_count} == #{response_count}" + else + Log.failure "Point counts don't match: #{point_count} != #{response_count}" + end + + # Log.info "Checking continuous query points for test2" + # st = Time.now + # response = influxdb.query("select count(value) from test2;") + # et = Time.now + + # Log.log "#{et-st} seconds elapsed" + + # response_count = response["test2"].first["count"] + # if point_count == response_count + # Log.success "Point counts match: #{point_count} == #{response_count}" + # else + # Log.failure "Point counts don't match: #{point_count} != #{response_count}" + # end +end + + diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/log.rb b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/log.rb new file mode 100644 index 000000000..0f70d7633 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/log.rb @@ -0,0 +1,23 @@ +module Log + def self.info(msg) + print Time.now.strftime("%r") + " | " + puts msg.to_s.colorize(:yellow) + end + + def self.success(msg) + print Time.now.strftime("%r") + " | " + puts msg.to_s.colorize(:green) + end + + def self.failure(msg) + print Time.now.strftime("%r") + " | " + puts msg.to_s.colorize(:red) + end + + def self.log(msg) + print Time.now.strftime("%r") + " | " + puts msg.to_s + end +end + + diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/random_gaussian.rb b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/random_gaussian.rb new file mode 100644 index 000000000..51d6c3c04 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/random_gaussian.rb @@ -0,0 +1,31 @@ +class RandomGaussian + def initialize(mean, stddev, rand_helper = lambda { Kernel.rand }) + @rand_helper = rand_helper + @mean = mean + @stddev = stddev + @valid = false + @next = 0 + end + + def rand + if @valid then + @valid = false + return @next + else + @valid = true + x, y = self.class.gaussian(@mean, @stddev, @rand_helper) + @next = y + return x + end + end + + private + def self.gaussian(mean, stddev, rand) + theta = 2 * Math::PI * rand.call + rho = Math.sqrt(-2 * Math.log(1 - rand.call)) + scale = stddev * rho + x = mean + scale * Math.cos(theta) + y = mean + scale * Math.sin(theta) + return x, y + end +end diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/random_points.rb b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/random_points.rb new file mode 100644 index 000000000..93bc8314f --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/burn-in/random_points.rb @@ -0,0 +1,29 @@ +require "influxdb" + +ONE_WEEK_IN_SECONDS = 7*24*60*60 +NUM_POINTS = 10_000 +BATCHES = 100 + +master = InfluxDB::Client.new +master.delete_database("ctx") rescue nil +master.create_database("ctx") + +influxdb = InfluxDB::Client.new "ctx" +influxdb.time_precision = "s" + +names = ["foo", "bar", "baz", "quu", "qux"] + +st = Time.now +BATCHES.times do |m| + points = [] + + puts "Writing #{NUM_POINTS} points, time ##{m}.." + NUM_POINTS.times do |n| + timestamp = Time.now.to_i - rand(ONE_WEEK_IN_SECONDS) + points << {value: names.sample, time: timestamp} + end + + influxdb.write_point("ct1", points) +end +puts st +puts Time.now diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/config.sample.toml b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/config.sample.toml new file mode 100644 index 000000000..9614277ab --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/etc/config.sample.toml @@ -0,0 +1,246 @@ +### Welcome to the InfluxDB configuration file. + +# Once every 24 hours InfluxDB will report anonymous data to m.influxdb.com +# The data includes raft id (random 8 bytes), os, arch, version, and metadata. +# We don't track ip addresses of servers reporting. This is only used +# to track the number of instances running and the versions, which +# is very helpful for us. +# Change this option to true to disable reporting. +reporting-disabled = false + +### +### [meta] +### +### Controls the parameters for the Raft consensus group that stores metadata +### about the InfluxDB cluster. +### + +[meta] + dir = "/var/opt/influxdb/meta" + hostname = "localhost" + bind-address = ":8088" + retention-autocreate = true + election-timeout = "1s" + heartbeat-timeout = "1s" + leader-lease-timeout = "500ms" + commit-timeout = "50ms" + +### +### [data] +### +### Controls where the actual shard data for InfluxDB lives and how it is +### flushed from the WAL. "dir" may need to be changed to a suitable place +### for your system, but the WAL settings are an advanced configuration. The +### defaults should work for most systems. +### + +[data] + dir = "/var/opt/influxdb/data" + + # The following WAL settings are for the b1 storage engine used in 0.9.2. They won't + # apply to any new shards created after upgrading to a version > 0.9.3. + max-wal-size = 104857600 # Maximum size the WAL can reach before a flush. Defaults to 100MB. + wal-flush-interval = "10m" # Maximum time data can sit in WAL before a flush. + wal-partition-flush-delay = "2s" # The delay time between each WAL partition being flushed. + + # These are the WAL settings for the storage engine >= 0.9.3 + wal-dir = "/var/opt/influxdb/wal" + wal-enable-logging = true + + # When a series in the WAL in-memory cache reaches this size in bytes it is marked as ready to + # flush to the index + # wal-ready-series-size = 25600 + + # Flush and compact a partition once this ratio of series are over the ready size + # wal-compaction-threshold = 0.6 + + # Force a flush and compaction if any series in a partition gets above this size in bytes + # wal-max-series-size = 2097152 + + # Force a flush of all series and full compaction if there have been no writes in this + # amount of time. This is useful for ensuring that shards that are cold for writes don't + # keep a bunch of data cached in memory and in the WAL. + # wal-flush-cold-interval = "10m" + + # Force a partition to flush its largest series if it reaches this approximate size in + # bytes. Remember there are 5 partitions so you'll need at least 5x this amount of memory. + # The more memory you have, the bigger this can be. + # wal-partition-size-threshold = 20971520 + +### +### [cluster] +### +### Controls non-Raft cluster behavior, which generally includes how data is +### shared across shards. +### + +[cluster] + shard-writer-timeout = "5s" # The time within which a shard must respond to write. + write-timeout = "5s" # The time within which a write operation must complete on the cluster. + +### +### [retention] +### +### Controls the enforcement of retention policies for evicting old data. +### + +[retention] + enabled = true + check-interval = "10m" + +### +### [admin] +### +### Controls the availability of the built-in, web-based admin interface. If HTTPS is +### enabled for the admin interface, HTTPS must also be enabled on the [http] service. +### + +[admin] + enabled = true + bind-address = ":8083" + https-enabled = false + https-certificate = "/etc/ssl/influxdb.pem" + +### +### [http] +### +### Controls how the HTTP endpoints are configured. These are the primary +### mechanism for getting data into and out of InfluxDB. +### + +[http] + enabled = true + bind-address = ":8086" + auth-enabled = false + log-enabled = true + write-tracing = false + pprof-enabled = false + https-enabled = false + https-certificate = "/etc/ssl/influxdb.pem" + +### +### [[graphite]] +### +### Controls one or many listeners for Graphite data. +### + +[[graphite]] + enabled = false + # bind-address = ":2003" + # protocol = "tcp" + # consistency-level = "one" + # name-separator = "." + + # These next lines control how batching works. You should have this enabled + # otherwise you could get dropped metrics or poor performance. Batching + # will buffer points in memory if you have many coming in. + + # batch-size = 1000 # will flush if this many points get buffered + # batch-timeout = "1s" # will flush at least this often even if we haven't hit buffer limit + + ## "name-schema" configures tag names for parsing the metric name from graphite protocol; + ## separated by `name-separator`. + ## The "measurement" tag is special and the corresponding field will become + ## the name of the metric. + ## e.g. "type.host.measurement.device" will parse "server.localhost.cpu.cpu0" as + ## { + ## measurement: "cpu", + ## tags: { + ## "type": "server", + ## "host": "localhost, + ## "device": "cpu0" + ## } + ## } + # name-schema = "type.host.measurement.device" + + ## If set to true, when the input metric name has more fields than `name-schema` specified, + ## the extra fields will be ignored. + ## Otherwise an error will be logged and the metric rejected. + # ignore-unnamed = true + +### +### [collectd] +### +### Controls the listener for collectd data. +### + +[collectd] + enabled = false + # bind-address = "" + # database = "" + # typesdb = "" + + # These next lines control how batching works. You should have this enabled + # otherwise you could get dropped metrics or poor performance. Batching + # will buffer points in memory if you have many coming in. + + # batch-size = 1000 # will flush if this many points get buffered + # batch-timeout = "1s" # will flush at least this often even if we haven't hit buffer limit + +### +### [opentsdb] +### +### Controls the listener for OpenTSDB data. +### + +[opentsdb] + enabled = false + # bind-address = "" + # database = "" + # retention-policy = "" + +### +### [[udp]] +### +### Controls the listeners for InfluxDB line protocol data via UDP. +### + +[[udp]] + enabled = false + # bind-address = "" + # database = "" + + # These next lines control how batching works. You should have this enabled + # otherwise you could get dropped metrics or poor performance. Batching + # will buffer points in memory if you have many coming in. + + # batch-size = 1000 # will flush if this many points get buffered + # batch-timeout = "1s" # will flush at least this often even if we haven't hit buffer limit + +### +### [monitoring] +### + +[monitoring] + enabled = true + write-interval = "24h" + +### +### [continuous_queries] +### +### Controls how continuous queries are run within InfluxDB. +### + +[continuous_queries] + log-enabled = true + enabled = true + recompute-previous-n = 2 + recompute-no-older-than = "10m" + compute-runs-per-interval = 10 + compute-no-more-than = "2m" + +### +### [hinted-handoff] +### +### Controls the hinted handoff feature, which allows nodes to temporarily +### store queued data when one node of a cluster is down for a short period +### of time. +### + +[hinted-handoff] + enabled = true + dir = "/var/opt/influxdb/hh" + max-size = 1073741824 + max-age = "168h" + retry-rate-limit = 0 + retry-interval = "1s" diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/importer/README.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/importer/README.md new file mode 100644 index 000000000..272fedf6c --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/importer/README.md @@ -0,0 +1,186 @@ +# Import/Export + +## Exporting from 0.8.9 + +Version `0.8.9` of InfluxDB adds support to export your data to a format that can be imported into `0.9.3` and later. + +### Design + +`0.8.9` exports raw data to a flat file that includes two sections, `DDL` and `DML`. You can choose to export them independently (see below). + +The `DDL` section contains the sql commands to create databases and retention policies. the `DML` section is [line protocol](https://github.com/influxdb/influxdb/blob/master/tsdb/README.md) and can be directly posted to the [http endpoint](https://influxdb.com/docs/v0.9/guides/writing_data.html) in `0.9`. Remember that batching is important and we don't recommend batch sizes over 5k. + +You need to specify a database and shard group when you export. + +To list out your shards, use the following http endpoint: + +`/cluster/shard_spaces` + +example: +```sh +http://username:password@localhost:8086/cluster/shard_spaces +``` + +Then, to export a database with then name "metrics" and a shard space with the name "default", issue the following curl command: + +```sh +curl -o export http://username:password@http://localhost:8086/export/metrics/default +``` + +Compression is supported, and will result in a significantly smaller file size. + +Use the following command for compression: +```sh +curl -o export.gz --compressed http://username:password@http://localhost:8086/export/metrics/default +``` + +You can also export just the `DDL` with this option: + +```sh +curl -o export.ddl http://username:password@http://localhost:8086/export/metrics/default?l=ddl +``` + +Or just the `DML` with this option: + +```sh +curl -o export.dml.gz --compressed http://username:password@http://localhost:8086/export/metrics/default?l=dml +``` + +### Assumptions + +- Series name mapping follows these [guidelines](https://influxdb.com/docs/v0.8/advanced_topics/schema_design.html) +- Database name will map directly from `0.8` to `0.9` +- Shard Spaces map to Retention Policies +- Shard Space Duration is ignored, as in `0.9` we determine shard size automatically +- Regex is used to match the correct series names and only exports that data for the database +- Duration becomes the new Retention Policy duration + +- Users are not migrated due to inability to get passwords. Anyone using users will need to manually set these back up in `0.9` + +### Upgrade Recommendations + +It's recommended that you upgrade to `0.9.3` first and have all your writes going there. Then, on the `0.8.X` instances, upgrade to `0.8.9`. + +It is important that when exporting you change your config to allow for the http endpoints not timing out. To do so, make this change in your config: + +```toml +# Configure the http api +[api] +read-timeout = "0s" +``` + +### Exceptions + +If a series can't be exported to tags based on the guidelines mentioned above, +we will insert the entire series name as the measurement name. You can either +allow that to import into the new InfluxDB instance, or you can do your own +data massage on it prior to importing it. + +For example, if you have the following series name: + +``` +metric.disk.c.host.server01.single +``` + +It will export as exactly thta as the measurement name and no tags: + +``` +metric.disk.c.host.server01.single +``` + +### Export Metrics + +When you export, you will now get comments inline in the `DML`: + +`# Found 999 Series for export` + +As well as count totals for each series exported: + +`# Series FOO - Points Exported: 999` + +With a total at the bottom: + +`# Points Exported: 999` + +You can grep the file that was exported at the end to get all the export metrics: + +`cat myexport | grep Exported` + +## Importing + +Version `0.9.3` of InfluxDB adds support to import your data from version `0.8.9`. + +## Caveats + +For the export/import to work, all requisites have to be met. For export, all series names in `0.8` should be in the following format: + +``` +.... +``` +for example: +``` +az.us-west-1.host.serverA.cpu +``` +or any number of tags +``` +building.2.temperature +``` + +Additionally, the fields need to have a consistent type (all float64, int64, etc) for every write in `0.8`. Otherwise they have the potential to fail writes in the import. +See below for more information. + +## Running the import command + + To import via the cli, you can specify the following command: + + ```sh + influx -import -path=metrics-default.gz -compressed + ``` + + If the file is not compressed you can issue it without the `-compressed` flag: + + ```sh + influx -import -path=metrics-default + ``` + + To redirect failed import lines to another file, run this command: + + ```sh + influx -import -path=metrics-default.gz -compressed > failures + ``` + + The import will use the line protocol in batches of 5,000 lines per batch when sending data to the server. + +### Throttiling the import + + If you need to throttle the import so the database has time to ingest, you can use the `-pps` flag. This will limit the points per second that will be sent to the server. + + ```sh + influx -import -path=metrics-default.gz -compressed -pps 50000 > failures + ``` + + Which is stating that you don't want MORE than 50,000 points per second to write to the database. Due to the processing that is taking place however, you will likely never get exactly 50,000 pps, more like 35,000 pps, etc. + +## Understanding the results of the import + +During the import, a status message will write out for every 100,000 points imported and report stats on the progress of the import: + +``` +2015/08/21 14:48:01 Processed 3100000 lines. Time elapsed: 56.740578415s. Points per second (PPS): 54634 +``` + + The batch will give some basic stats when finished: + + ```sh + 2015/07/29 23:15:20 Processed 2 commands + 2015/07/29 23:15:20 Processed 70207923 inserts + 2015/07/29 23:15:20 Failed 29785000 inserts + ``` + + Most inserts fail due to the following types of error: + + ```sh + 2015/07/29 22:18:28 error writing batch: write failed: field type conflict: input field "value" on measurement "metric" is type float64, already exists as type integer + ``` + + This is due to the fact that in `0.8` a field could get created and saved as int or float types for independent writes. In `0.9` the field has to have a consistent type. diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/importer/v8/importer.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/importer/v8/importer.go new file mode 100644 index 000000000..5095868f3 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/importer/v8/importer.go @@ -0,0 +1,236 @@ +package v8 + +import ( + "bufio" + "compress/gzip" + "fmt" + "io" + "log" + "net/url" + "os" + "strings" + "time" + + "github.com/influxdb/influxdb/client" +) + +const batchSize = 5000 + +// Config is the config used to initialize a Importer importer +type Config struct { + Username string + Password string + URL url.URL + Precision string + WriteConsistency string + Path string + Version string + Compressed bool + PPS int +} + +// NewConfig returns an initialized *Config +func NewConfig() *Config { + return &Config{} +} + +// Importer is the importer used for importing 0.8 data +type Importer struct { + client *client.Client + database string + retentionPolicy string + config *Config + batch []string + totalInserts int + failedInserts int + totalCommands int + throttlePointsWritten int + lastWrite time.Time + throttle *time.Ticker +} + +// NewImporter will return an intialized Importer struct +func NewImporter(config *Config) *Importer { + return &Importer{ + config: config, + batch: make([]string, 0, batchSize), + } +} + +// Import processes the specified file in the Config and writes the data to the databases in chunks specified by batchSize +func (i *Importer) Import() error { + // Create a client and try to connect + config := client.NewConfig() + config.URL = i.config.URL + config.Username = i.config.Username + config.Password = i.config.Password + config.UserAgent = fmt.Sprintf("influxDB importer/%s", i.config.Version) + cl, err := client.NewClient(config) + if err != nil { + return fmt.Errorf("could not create client %s", err) + } + i.client = cl + if _, _, e := i.client.Ping(); e != nil { + return fmt.Errorf("failed to connect to %s\n", i.client.Addr()) + } + + // Validate args + if i.config.Path == "" { + return fmt.Errorf("file argument required") + } + + defer func() { + if i.totalInserts > 0 { + log.Printf("Processed %d commands\n", i.totalCommands) + log.Printf("Processed %d inserts\n", i.totalInserts) + log.Printf("Failed %d inserts\n", i.failedInserts) + } + }() + + // Open the file + f, err := os.Open(i.config.Path) + if err != nil { + return err + } + defer f.Close() + + var r io.Reader + + // If gzipped, wrap in a gzip reader + if i.config.Compressed { + gr, err := gzip.NewReader(f) + if err != nil { + return err + } + defer gr.Close() + // Set the reader to the gzip reader + r = gr + } else { + // Standard text file so our reader can just be the file + r = f + } + + // Get our reader + scanner := bufio.NewScanner(r) + + // Process the DDL + i.processDDL(scanner) + + // Set up our throttle channel. Since there is effectively no other activity at this point + // the smaller resolution gets us much closer to the requested PPS + i.throttle = time.NewTicker(time.Microsecond) + defer i.throttle.Stop() + + // Prime the last write + i.lastWrite = time.Now() + + // Process the DML + i.processDML(scanner) + + // Check if we had any errors scanning the file + if err := scanner.Err(); err != nil { + return fmt.Errorf("reading standard input: %s", err) + } + + return nil +} + +func (i *Importer) processDDL(scanner *bufio.Scanner) { + for scanner.Scan() { + line := scanner.Text() + // If we find the DML token, we are done with DDL + if strings.HasPrefix(line, "# DML") { + return + } + if strings.HasPrefix(line, "#") { + continue + } + i.queryExecutor(line) + } +} + +func (i *Importer) processDML(scanner *bufio.Scanner) { + start := time.Now() + for scanner.Scan() { + line := scanner.Text() + if strings.HasPrefix(line, "# CONTEXT-DATABASE:") { + i.database = strings.TrimSpace(strings.Split(line, ":")[1]) + } + if strings.HasPrefix(line, "# CONTEXT-RETENTION-POLICY:") { + i.retentionPolicy = strings.TrimSpace(strings.Split(line, ":")[1]) + } + if strings.HasPrefix(line, "#") { + continue + } + i.batchAccumulator(line, start) + } +} + +func (i *Importer) execute(command string) { + response, err := i.client.Query(client.Query{Command: command, Database: i.database}) + if err != nil { + log.Printf("error: %s\n", err) + return + } + if err := response.Error(); err != nil { + log.Printf("error: %s\n", response.Error()) + } +} + +func (i *Importer) queryExecutor(command string) { + i.totalCommands++ + i.execute(command) +} + +func (i *Importer) batchAccumulator(line string, start time.Time) { + i.batch = append(i.batch, line) + if len(i.batch) == batchSize { + if e := i.batchWrite(); e != nil { + log.Println("error writing batch: ", e) + // Output failed lines to STDOUT so users can capture lines that failed to import + fmt.Println(strings.Join(i.batch, "\n")) + i.failedInserts += len(i.batch) + } else { + i.totalInserts += len(i.batch) + } + i.batch = i.batch[:0] + // Give some status feedback every 100000 lines processed + processed := i.totalInserts + i.failedInserts + if processed%100000 == 0 { + since := time.Since(start) + pps := float64(processed) / since.Seconds() + log.Printf("Processed %d lines. Time elapsed: %s. Points per second (PPS): %d", processed, since.String(), int64(pps)) + } + } +} + +func (i *Importer) batchWrite() error { + // Accumulate the batch size to see how many points we have written this second + i.throttlePointsWritten += len(i.batch) + + // Find out when we last wrote data + since := time.Since(i.lastWrite) + + // Check to see if we've exceeded our points per second for the current timeframe + var currentPPS int + if since.Seconds() > 0 { + currentPPS = int(float64(i.throttlePointsWritten) / since.Seconds()) + } else { + currentPPS = i.throttlePointsWritten + } + + // If our currentPPS is greater than the PPS specified, then we wait and retry + if int(currentPPS) > i.config.PPS && i.config.PPS != 0 { + // Wait for the next tick + <-i.throttle.C + + // Decrement the batch size back out as it is going to get called again + i.throttlePointsWritten -= len(i.batch) + return i.batchWrite() + } + + _, e := i.client.WriteLineProtocol(strings.Join(i.batch, "\n"), i.database, i.retentionPolicy, i.config.Precision, i.config.WriteConsistency) + i.throttlePointsWritten = 0 + i.lastWrite = time.Now() + return e +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/ast.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/ast.go index 6235e405a..31cf82e24 100644 --- a/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/ast.go +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/ast.go @@ -985,6 +985,10 @@ func (s *SelectStatement) hasTimeDimensions(node Node) bool { } func (s *SelectStatement) validate(tr targetRequirement) error { + if err := s.validateFields(); err != nil { + return err + } + if err := s.validateDistinct(); err != nil { return err } @@ -1008,6 +1012,14 @@ func (s *SelectStatement) validate(tr targetRequirement) error { return nil } +func (s *SelectStatement) validateFields() error { + ns := s.NamesInSelect() + if len(ns) == 1 && ns[0] == "time" { + return fmt.Errorf("at least 1 non-time field must be queried") + } + return nil +} + func (s *SelectStatement) validateAggregates(tr targetRequirement) error { // First, if 1 field is an aggregate, then all fields must be an aggregate. This is // a explicit limitation of the current system. diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/parser_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/parser_test.go index 71ccbffc3..b9bb9756d 100644 --- a/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/parser_test.go +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/influxql/parser_test.go @@ -1206,6 +1206,7 @@ func TestParser_ParseStatement(t *testing.T) { // Errors {s: ``, err: `found EOF, expected SELECT, DELETE, SHOW, CREATE, DROP, GRANT, REVOKE, ALTER, SET at line 1, char 1`}, {s: `SELECT`, err: `found EOF, expected identifier, string, number, bool at line 1, char 8`}, + {s: `SELECT time FROM myseries`, err: `at least 1 non-time field must be queried`}, {s: `blah blah`, err: `found blah, expected SELECT, DELETE, SHOW, CREATE, DROP, GRANT, REVOKE, ALTER, SET at line 1, char 1`}, {s: `SELECT field1 X`, err: `found X, expected FROM at line 1, char 15`}, {s: `SELECT field1 FROM "series" WHERE X +;`, err: `found ;, expected identifier, string, number, bool at line 1, char 38`}, diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data.go index 33c829bb8..2b9168901 100644 --- a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data.go +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data.go @@ -212,6 +212,7 @@ func (data *Data) UpdateRetentionPolicy(database, name string, rpu *RetentionPol } if rpu.Duration != nil { rpi.Duration = *rpu.Duration + rpi.ShardGroupDuration = shardGroupDuration(rpi.Duration) } if rpu.ReplicaN != nil { rpi.ReplicaN = *rpu.ReplicaN diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data_test.go index 945d78242..6e0314c76 100644 --- a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data_test.go +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/data_test.go @@ -178,7 +178,7 @@ func TestData_UpdateRetentionPolicy(t *testing.T) { if rpi, _ := data.RetentionPolicy("db0", "rp1"); !reflect.DeepEqual(rpi, &meta.RetentionPolicyInfo{ Name: "rp1", Duration: 10 * time.Hour, - ShardGroupDuration: 604800000000000, + ShardGroupDuration: 3600000000000, ReplicaN: 3, }) { t.Fatalf("unexpected policy: %#v", rpi) diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/state.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/state.go index a442a8007..c9fc02c4d 100644 --- a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/state.go +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/state.go @@ -334,8 +334,6 @@ func (r *localRaft) leader() string { } func (r *localRaft) isLeader() bool { - r.store.mu.RLock() - defer r.store.mu.RUnlock() if r.raft == nil { return false } diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/store_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/store_test.go index 6dca9b57c..ef79306b7 100644 --- a/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/store_test.go +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/meta/store_test.go @@ -361,7 +361,7 @@ func TestStore_UpdateRetentionPolicy(t *testing.T) { } else if !reflect.DeepEqual(rpi, &meta.RetentionPolicyInfo{ Name: "rp1", Duration: 10 * time.Hour, - ShardGroupDuration: 7 * 24 * time.Hour, + ShardGroupDuration: 1 * time.Hour, ReplicaN: 1, }) { t.Fatalf("unexpected policy: %#v", rpi) diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/nightly.sh b/Godeps/_workspace/src/github.com/influxdb/influxdb/nightly.sh new file mode 100644 index 000000000..9052201a2 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/nightly.sh @@ -0,0 +1,14 @@ +#!/bin/bash + +REPO_DIR=`mktemp -d` +echo "Using $REPO_DIR for all work..." + +cd $REPO_DIR +export GOPATH=`pwd` +mkdir -p $GOPATH/src/github.com/influxdb +cd $GOPATH/src/github.com/influxdb +git clone https://github.com/influxdb/influxdb.git + +cd $GOPATH/src/github.com/influxdb/influxdb +NIGHTLY_BUILD=true ./package.sh 0.9.3-nightly-`git log --pretty=format:'%h' -n 1` +rm -rf $REPO_DIR diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/package.sh b/Godeps/_workspace/src/github.com/influxdb/influxdb/package.sh new file mode 100644 index 000000000..c86a8b3eb --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/package.sh @@ -0,0 +1,409 @@ +#!/usr/bin/env bash + +########################################################################### +# Packaging script which creates debian and RPM packages. It optionally +# tags the repo with the given version. +# +# Requirements: GOPATH must be set. 'fpm' must be on the path, and the AWS +# CLI tools must also be installed. +# +# https://github.com/jordansissel/fpm +# http://aws.amazon.com/cli/ +# +# Packaging process: to package a build, simple execute: +# +# package.sh +# +# where is the desired version. If generation of a debian and RPM +# package is successful, the script will offer to tag the repo using the +# supplied version string. +# +# AWS upload: the script will also offer to upload the packages to S3. If +# this option is selected, the credentials should be present in the file +# ~/aws.conf. The contents should be of the form: +# +# [default] +# aws_access_key_id= +# aws_secret_access_key= +# region = us-east-1 +# +# Trim the leading spaces when creating the file. The script will exit if +# S3 upload is requested, but this file does not exist. + +[ -z $DEBUG ] || set -x + +AWS_FILE=~/aws.conf + +INSTALL_ROOT_DIR=/opt/influxdb +INFLUXDB_LOG_DIR=/var/log/influxdb +INFLUXDB_DATA_DIR=/var/opt/influxdb +CONFIG_ROOT_DIR=/etc/opt/influxdb + +SAMPLE_CONFIGURATION=etc/config.sample.toml +INITD_SCRIPT=scripts/init.sh + +TMP_WORK_DIR=`mktemp -d` +POST_INSTALL_PATH=`mktemp` +ARCH=`uname -i` +LICENSE=MIT +URL=influxdb.com +MAINTAINER=support@influxdb.com +VENDOR=Influxdb +DESCRIPTION="Distributed time-series database" + +# Allow path to FPM to be set by environment variables. Some execution contexts +# like cron don't have PATH set correctly to pick it up. +if [ -z "$FPM" ]; then + FPM=`which fpm` +fi + +GO_VERSION="go1.4.2" +GOPATH_INSTALL= +BINS=( + influxd + influx + ) + +########################################################################### +# Helper functions. + +# usage prints simple usage information. +usage() { + echo -e "$0 [] [-h]\n" + cleanup_exit $1 +} + +# cleanup_exit removes all resources created during the process and exits with +# the supplied returned code. +cleanup_exit() { + rm -r $TMP_WORK_DIR + rm $POST_INSTALL_PATH + exit $1 +} + +# current_branch echos the current git branch. +current_branch() { + echo `git rev-parse --abbrev-ref HEAD` +} + +# check_gopath sanity checks the value of the GOPATH env variable, and determines +# the path where build artifacts are installed. GOPATH may be a colon-delimited +# list of directories. +check_gopath() { + [ -z "$GOPATH" ] && echo "GOPATH is not set." && cleanup_exit 1 + GOPATH_INSTALL=`echo $GOPATH | cut -d ':' -f 1` + [ ! -d "$GOPATH_INSTALL" ] && echo "GOPATH_INSTALL is not a directory." && cleanup_exit 1 + echo "GOPATH ($GOPATH) looks sane, using $GOPATH_INSTALL for installation." +} + +check_gvm() { + if [ -n "$GOPATH" ]; then + existing_gopath=$GOPATH + fi + + source $HOME/.gvm/scripts/gvm + which gvm + if [ $? -ne 0 ]; then + echo "gvm not found -- aborting." + cleanup_exit $1 + fi + gvm use $GO_VERSION + if [ $? -ne 0 ]; then + echo "gvm cannot find Go version $GO_VERSION -- aborting." + cleanup_exit $1 + fi + + # Keep any existing GOPATH set. + if [ -n "$existing_gopath" ]; then + GOPATH=$existing_gopath + fi +} + +# check_clean_tree ensures that no source file is locally modified. +check_clean_tree() { + modified=$(git ls-files --modified | wc -l) + if [ $modified -ne 0 ]; then + echo "The source tree is not clean -- aborting." + cleanup_exit 1 + fi + echo "Git tree is clean." +} + +# update_tree ensures the tree is in-sync with the repo. +update_tree() { + git pull origin $TARGET_BRANCH + if [ $? -ne 0 ]; then + echo "Failed to pull latest code -- aborting." + cleanup_exit 1 + fi + git fetch --tags + if [ $? -ne 0 ]; then + echo "Failed to fetch tags -- aborting." + cleanup_exit 1 + fi + echo "Git tree updated successfully." +} + +# check_tag_exists checks if the existing release already exists in the tags. +check_tag_exists () { + version=$1 + git tag | grep -q "^v$version$" + if [ $? -eq 0 ]; then + echo "Proposed version $version already exists as a tag -- aborting." + cleanup_exit 1 + fi +} + +# make_dir_tree creates the directory structure within the packages. +make_dir_tree() { + work_dir=$1 + version=$2 + mkdir -p $work_dir/$INSTALL_ROOT_DIR/versions/$version/scripts + if [ $? -ne 0 ]; then + echo "Failed to create installation directory -- aborting." + cleanup_exit 1 + fi + mkdir -p $work_dir/$CONFIG_ROOT_DIR + if [ $? -ne 0 ]; then + echo "Failed to create configuration directory -- aborting." + cleanup_exit 1 + fi +} + + +# do_build builds the code. The version and commit must be passed in. +do_build() { + for b in ${BINS[*]}; do + rm -f $GOPATH_INSTALL/bin/$b + done + go get -u -f -d ./... + if [ $? -ne 0 ]; then + echo "WARNING: failed to 'go get' packages." + fi + + git checkout $TARGET_BRANCH # go get switches to master, so ensure we're back. + version=$1 + commit=`git rev-parse HEAD` + branch=`current_branch` + if [ $? -ne 0 ]; then + echo "Unable to retrieve current commit -- aborting" + cleanup_exit 1 + fi + + go install -a -ldflags="-X main.version $version -X main.branch $branch -X main.commit $commit" ./... + if [ $? -ne 0 ]; then + echo "Build failed, unable to create package -- aborting" + cleanup_exit 1 + fi + echo "Build completed successfully." +} + +# generate_postinstall_script creates the post-install script for the +# package. It must be passed the version. +generate_postinstall_script() { + version=$1 + cat <$POST_INSTALL_PATH +rm -f $INSTALL_ROOT_DIR/influxd +rm -f $INSTALL_ROOT_DIR/influx +rm -f $INSTALL_ROOT_DIR/init.sh +ln -s $INSTALL_ROOT_DIR/versions/$version/influxd $INSTALL_ROOT_DIR/influxd +ln -s $INSTALL_ROOT_DIR/versions/$version/influx $INSTALL_ROOT_DIR/influx +ln -s $INSTALL_ROOT_DIR/versions/$version/scripts/init.sh $INSTALL_ROOT_DIR/init.sh + +rm -f /etc/init.d/influxdb +ln -sfn $INSTALL_ROOT_DIR/init.sh /etc/init.d/influxdb +chmod +x /etc/init.d/influxdb +if which update-rc.d > /dev/null 2>&1 ; then + update-rc.d -f influxdb remove + update-rc.d influxdb defaults +else + chkconfig --add influxdb +fi + +if ! id influxdb >/dev/null 2>&1; then + useradd --system -U -M influxdb +fi +chown -R -L influxdb:influxdb $INSTALL_ROOT_DIR +chmod -R a+rX $INSTALL_ROOT_DIR + +mkdir -p $INFLUXDB_LOG_DIR +chown -R -L influxdb:influxdb $INFLUXDB_LOG_DIR +mkdir -p $INFLUXDB_DATA_DIR +chown -R -L influxdb:influxdb $INFLUXDB_DATA_DIR +EOF + echo "Post-install script created successfully at $POST_INSTALL_PATH" +} + +########################################################################### +# Start the packaging process. + +if [ $# -ne 1 ]; then + usage 1 +elif [ $1 == "-h" ]; then + usage 0 +else + VERSION=$1 + VERSION_UNDERSCORED=`echo "$VERSION" | tr - _` +fi + +echo -e "\nStarting package process...\n" + +# Ensure the current is correct. +TARGET_BRANCH=`current_branch` +if [ -z "$NIGHTLY_BUILD" ]; then +echo -n "Current branch is $TARGET_BRANCH. Start packaging this branch? [Y/n] " + read response + response=`echo $response | tr 'A-Z' 'a-z'` + if [ "x$response" == "xn" ]; then + echo "Packaging aborted." + cleanup_exit 1 + fi +fi + +check_gvm +check_gopath +if [ -z "$NIGHTLY_BUILD" ]; then + check_clean_tree + update_tree + check_tag_exists $VERSION +fi + +do_build $VERSION +make_dir_tree $TMP_WORK_DIR $VERSION + +########################################################################### +# Copy the assets to the installation directories. + +for b in ${BINS[*]}; do + cp $GOPATH_INSTALL/bin/$b $TMP_WORK_DIR/$INSTALL_ROOT_DIR/versions/$VERSION + if [ $? -ne 0 ]; then + echo "Failed to copy binaries to packaging directory -- aborting." + cleanup_exit 1 + fi +done +echo "${BINS[*]} copied to $TMP_WORK_DIR/$INSTALL_ROOT_DIR/versions/$VERSION" + +cp $INITD_SCRIPT $TMP_WORK_DIR/$INSTALL_ROOT_DIR/versions/$VERSION/scripts +if [ $? -ne 0 ]; then + echo "Failed to copy init.d script to packaging directory -- aborting." + cleanup_exit 1 +fi +echo "$INITD_SCRIPT copied to $TMP_WORK_DIR/$INSTALL_ROOT_DIR/versions/$VERSION/scripts" + +cp $SAMPLE_CONFIGURATION $TMP_WORK_DIR/$CONFIG_ROOT_DIR/influxdb.conf +if [ $? -ne 0 ]; then + echo "Failed to copy $SAMPLE_CONFIGURATION to packaging directory -- aborting." + cleanup_exit 1 +fi + +generate_postinstall_script $VERSION + +########################################################################### +# Create the actual packages. + +if [ -z "$NIGHTLY_BUILD" ]; then + echo -n "Commence creation of $ARCH packages, version $VERSION? [Y/n] " + read response + response=`echo $response | tr 'A-Z' 'a-z'` + if [ "x$response" == "xn" ]; then + echo "Packaging aborted." + cleanup_exit 1 + fi +fi + +if [ $ARCH == "i386" ]; then + rpm_package=influxdb-${VERSION}-1.i686.rpm # RPM packages use 1 for default package release. + debian_package=influxdb_${VERSION}_i686.deb + deb_args="-a i686" + rpm_args="setarch i686" +elif [ $ARCH == "arm" ]; then + rpm_package=influxdb-${VERSION}-1.armel.rpm + debian_package=influxdb_${VERSION}_armel.deb +else + rpm_package=influxdb-${VERSION}-1.x86_64.rpm + debian_package=influxdb_${VERSION}_amd64.deb +fi + +COMMON_FPM_ARGS="-C $TMP_WORK_DIR --vendor $VENDOR --url $URL --license $LICENSE --maintainer $MAINTAINER --after-install $POST_INSTALL_PATH --name influxdb --version $VERSION --config-files $CONFIG_ROOT_DIR ." +$rpm_args $FPM -s dir -t rpm --description "$DESCRIPTION" $COMMON_FPM_ARGS +if [ $? -ne 0 ]; then + echo "Failed to create RPM package -- aborting." + cleanup_exit 1 +fi +echo "RPM package created successfully." + +$FPM -s dir -t deb $deb_args --description "$DESCRIPTION" $COMMON_FPM_ARGS +if [ $? -ne 0 ]; then + echo "Failed to create Debian package -- aborting." + cleanup_exit 1 +fi +echo "Debian package created successfully." + +$FPM -s dir -t tar --prefix influxdb_${VERSION}_${ARCH} -p influxdb_${VERSION}_${ARCH}.tar.gz --description "$DESCRIPTION" $COMMON_FPM_ARGS +if [ $? -ne 0 ]; then + echo "Failed to create Tar package -- aborting." + cleanup_exit 1 +fi +echo "Tar package created successfully." + +########################################################################### +# Offer to tag the repo. + +if [ -z "$NIGHTLY_BUILD" ]; then + echo -n "Tag source tree with v$VERSION and push to repo? [y/N] " + read response + response=`echo $response | tr 'A-Z' 'a-z'` + if [ "x$response" == "xy" ]; then + echo "Creating tag v$VERSION and pushing to repo" + git tag v$VERSION + if [ $? -ne 0 ]; then + echo "Failed to create tag v$VERSION -- aborting" + cleanup_exit 1 + fi + git push origin v$VERSION + if [ $? -ne 0 ]; then + echo "Failed to push tag v$VERSION to repo -- aborting" + cleanup_exit 1 + fi + else + echo "Not creating tag v$VERSION." + fi +fi + +########################################################################### +# Offer to publish the packages. + +if [ -z "$NIGHTLY_BUILD" ]; then + echo -n "Publish packages to S3? [y/N] " + read response + response=`echo $response | tr 'A-Z' 'a-z'` +fi + +if [ "x$response" == "xy" -o -n "$NIGHTLY_BUILD" ]; then + echo "Publishing packages to S3." + if [ ! -e "$AWS_FILE" ]; then + echo "$AWS_FILE does not exist -- aborting." + cleanup_exit 1 + fi + + for filepath in `ls *.{deb,rpm,gz}`; do + filename=`basename $filepath` + if [ -n "$NIGHTLY_BUILD" ]; then + filename=`echo $filename | sed s/$VERSION/nightly/` + filename=`echo $filename | sed s/$VERSION_UNDERSCORED/nightly/` + fi + AWS_CONFIG_FILE=$AWS_FILE aws s3 cp $filepath s3://influxdb/$filename --acl public-read --region us-east-1 + if [ $? -ne 0 ]; then + echo "Upload failed -- aborting". + cleanup_exit 1 + fi + done +else + echo "Not publishing packages to S3." +fi + +########################################################################### +# All done. + +echo -e "\nPackaging process complete." +cleanup_exit 0 diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/scripts/influxdb.service b/Godeps/_workspace/src/github.com/influxdb/influxdb/scripts/influxdb.service new file mode 100644 index 000000000..c164ed3b2 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/scripts/influxdb.service @@ -0,0 +1,16 @@ +# If you modify this, please also make sure to edit init.sh + +[Unit] +Description=InfluxDB is an open-source, distributed, time series database +After=network.target + +[Service] +User=influxdb +Group=influxdb +LimitNOFILE=65536 +EnvironmentFile=-/etc/default/influxdb +ExecStart=/opt/influxdb/influxd -config /etc/opt/influxdb/influxdb.conf $INFLUXD_OPTS +Restart=on-failure + +[Install] +WantedBy=multi-user.target diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/scripts/init.sh b/Godeps/_workspace/src/github.com/influxdb/influxdb/scripts/init.sh new file mode 100644 index 000000000..7e97b0def --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/scripts/init.sh @@ -0,0 +1,205 @@ +#! /usr/bin/env bash + +### BEGIN INIT INFO +# Provides: influxd +# Required-Start: $all +# Required-Stop: $remote_fs $syslog +# Default-Start: 2 3 4 5 +# Default-Stop: 0 1 6 +# Short-Description: Start influxd at boot time +### END INIT INFO + +# If you modify this, please make sure to also edit influxdb.service +# this init script supports three different variations: +# 1. New lsb that define start-stop-daemon +# 2. Old lsb that don't have start-stop-daemon but define, log, pidofproc and killproc +# 3. Centos installations without lsb-core installed +# +# In the third case we have to define our own functions which are very dumb +# and expect the args to be positioned correctly. + +# Command-line options that can be set in /etc/default/influxdb. These will override +# any config file values. Example: "-join http://1.2.3.4:8086" +DEFAULT=/etc/default/influxdb + +# Daemon options +INFLUXD_OPTS= + +# Process name ( For display ) +NAME=influxdb + +# User and group +USER=influxdb +GROUP=influxdb + +# Daemon name, where is the actual executable +# If the daemon is not there, then exit. +DAEMON=/opt/influxdb/influxd +[ -x $DAEMON ] || exit 5 + +# Configuration file +CONFIG=/etc/opt/influxdb/influxdb.conf + +# PID file for the daemon +PIDFILE=/var/run/influxdb/influxd.pid +PIDDIR=`dirname $PIDFILE` +if [ ! -d "$PIDDIR" ]; then + mkdir -p $PIDDIR + chown $GROUP:$USER $PIDDIR +fi + +# Max open files +OPEN_FILE_LIMIT=65536 + +if [ -r /lib/lsb/init-functions ]; then + source /lib/lsb/init-functions +fi + +# Logging +if [ -z "$STDOUT" ]; then + STDOUT=/dev/null +fi + +if [ ! -f "$STDOUT" ]; then + mkdir -p $(dirname $STDOUT) +fi + +if [ -z "$STDERR" ]; then + STDERR=/var/log/influxdb/influxd.log +fi + +if [ ! -f "$STDERR" ]; then + mkdir -p $(dirname $STDERR) +fi + +# Overwrite init script variables with /etc/default/influxdb values +if [ -r $DEFAULT ]; then + source $DEFAULT +fi + +function pidofproc() { + if [ $# -ne 3 ]; then + echo "Expected three arguments, e.g. $0 -p pidfile daemon-name" + fi + + PID=`pgrep -f $3` + local PIDFILE=`cat $2` + + if [ "x$PIDFILE" == "x" ]; then + return 1 + fi + + if [ "x$PID" != "x" -a "$PIDFILE" == "$PID" ]; then + return 0 + fi + + return 1 +} + +function killproc() { + if [ $# -ne 3 ]; then + echo "Expected three arguments, e.g. $0 -p pidfile signal" + fi + + PID=`cat $2` + + kill -s $3 $PID +} + +function log_failure_msg() { + echo "$@" "[ FAILED ]" +} + +function log_success_msg() { + echo "$@" "[ OK ]" +} + +case $1 in + start) + # Check if config file exist + if [ ! -r $CONFIG ]; then + log_failure_msg "config file doesn't exists" + exit 4 + fi + + # Checked the PID file exists and check the actual status of process + if [ -e $PIDFILE ]; then + pidofproc -p $PIDFILE $DAEMON > /dev/null 2>&1 && STATUS="0" || STATUS="$?" + # If the status is SUCCESS then don't need to start again. + if [ "x$STATUS" = "x0" ]; then + log_failure_msg "$NAME process is running" + exit 0 # Exit + fi + # if PID file does not exist, check if writable + else + su -c "touch $PIDFILE" $USER > /dev/null 2>&1 + if [ $? -ne 0 ]; then + log_failure_msg "$PIDFILE not writable, check permissions" + exit 5 + fi + fi + + # Bump the file limits, before launching the daemon. These will carry over to + # launched processes. + ulimit -n $OPEN_FILE_LIMIT + if [ $? -ne 0 ]; then + log_failure_msg "set open file limit to $OPEN_FILE_LIMIT" + exit 1 + fi + + log_success_msg "Starting the process" "$NAME" + if which start-stop-daemon > /dev/null 2>&1; then + start-stop-daemon --chuid $GROUP:$USER --start --quiet --pidfile $PIDFILE --exec $DAEMON -- -pidfile $PIDFILE -config $CONFIG $INFLUXD_OPTS >>$STDOUT 2>>$STDERR & + else + nohup $DAEMON -pidfile $PIDFILE -config $CONFIG $INFLUXD_OPTS >>$STDOUT 2>>$STDERR & + fi + log_success_msg "$NAME process was started" + ;; + + stop) + # Stop the daemon. + if [ -e $PIDFILE ]; then + pidofproc -p $PIDFILE $DAEMON > /dev/null 2>&1 && STATUS="0" || STATUS="$?" + if [ "$STATUS" = 0 ]; then + if killproc -p $PIDFILE SIGTERM && /bin/rm -rf $PIDFILE; then + log_success_msg "$NAME process was stopped" + else + log_failure_msg "$NAME failed to stop service" + fi + fi + else + log_failure_msg "$NAME process is not running" + fi + ;; + + restart) + # Restart the daemon. + $0 stop && sleep 2 && $0 start + ;; + + status) + # Check the status of the process. + if [ -e $PIDFILE ]; then + if pidofproc -p $PIDFILE $DAEMON > /dev/null; then + log_success_msg "$NAME Process is running" + exit 0 + else + log_failure_msg "$NAME Process is not running" + exit 1 + fi + else + log_failure_msg "$NAME Process is not running" + exit 3 + fi + ;; + + version) + $DAEMON version + ;; + + *) + # For invalid arguments, print the usage message. + echo "Usage: $0 {start|stop|restart|status|version}" + exit 2 + ;; +esac diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/config.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/config.go new file mode 100644 index 000000000..860dff864 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/config.go @@ -0,0 +1,21 @@ +package admin + +const ( + // DefaultBindAddress is the default bind address for the HTTP server. + DefaultBindAddress = ":8083" +) + +type Config struct { + Enabled bool `toml:"enabled"` + BindAddress string `toml:"bind-address"` + HttpsEnabled bool `toml:"https-enabled"` + HttpsCertificate string `toml:"https-certificate"` +} + +func NewConfig() Config { + return Config{ + BindAddress: DefaultBindAddress, + HttpsEnabled: false, + HttpsCertificate: "/etc/ssl/influxdb.pem", + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/config_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/config_test.go new file mode 100644 index 000000000..1b0422505 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/config_test.go @@ -0,0 +1,32 @@ +package admin_test + +import ( + "testing" + + "github.com/BurntSushi/toml" + "github.com/influxdb/influxdb/services/admin" +) + +func TestConfig_Parse(t *testing.T) { + // Parse configuration. + var c admin.Config + if _, err := toml.Decode(` +enabled = true +bind-address = ":8083" +https-enabled = true +https-certificate = "/dev/null" +`, &c); err != nil { + t.Fatal(err) + } + + // Validate configuration. + if c.Enabled != true { + t.Fatalf("unexpected enabled: %v", c.Enabled) + } else if c.BindAddress != ":8083" { + t.Fatalf("unexpected bind address: %s", c.BindAddress) + } else if c.HttpsEnabled != true { + t.Fatalf("unexpected https enabled: %v", c.HttpsEnabled) + } else if c.HttpsCertificate != "/dev/null" { + t.Fatalf("unexpected https certificate: %v", c.HttpsCertificate) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/service.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/service.go new file mode 100644 index 000000000..2618bdb6b --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/service.go @@ -0,0 +1,111 @@ +package admin + +import ( + "crypto/tls" + "fmt" + "log" + "net" + "net/http" + "os" + "strings" + + // Register static assets via statik. + _ "github.com/influxdb/influxdb/statik" + "github.com/rakyll/statik/fs" +) + +// Service manages the listener for an admin endpoint. +type Service struct { + listener net.Listener + addr string + https bool + cert string + err chan error + + logger *log.Logger +} + +// NewService returns a new instance of Service. +func NewService(c Config) *Service { + return &Service{ + addr: c.BindAddress, + https: c.HttpsEnabled, + cert: c.HttpsCertificate, + err: make(chan error), + logger: log.New(os.Stderr, "[admin] ", log.LstdFlags), + } +} + +// Open starts the service +func (s *Service) Open() error { + s.logger.Printf("Starting admin service") + + // Open listener. + if s.https { + cert, err := tls.LoadX509KeyPair(s.cert, s.cert) + if err != nil { + return err + } + + listener, err := tls.Listen("tcp", s.addr, &tls.Config{ + Certificates: []tls.Certificate{cert}, + }) + if err != nil { + return err + } + + s.logger.Println("Listening on HTTPS:", listener.Addr().String()) + s.listener = listener + } else { + listener, err := net.Listen("tcp", s.addr) + if err != nil { + return err + } + + s.logger.Println("Listening on HTTP:", listener.Addr().String()) + s.listener = listener + } + + // Begin listening for requests in a separate goroutine. + go s.serve() + return nil +} + +// Close closes the underlying listener. +func (s *Service) Close() error { + if s.listener != nil { + return s.listener.Close() + } + return nil +} + +// SetLogger sets the internal logger to the logger passed in. +func (s *Service) SetLogger(l *log.Logger) { + s.logger = l +} + +// Err returns a channel for fatal errors that occur on the listener. +func (s *Service) Err() <-chan error { return s.err } + +// Addr returns the listener's address. Returns nil if listener is closed. +func (s *Service) Addr() net.Addr { + if s.listener != nil { + return s.listener.Addr() + } + return nil +} + +// serve serves the handler from the listener. +func (s *Service) serve() { + // Instantiate file system from embedded admin. + statikFS, err := fs.New() + if err != nil { + panic(err) + } + + // Run file system handler on listener. + err = http.Serve(s.listener, http.FileServer(statikFS)) + if err != nil && !strings.Contains(err.Error(), "closed") { + s.err <- fmt.Errorf("listener error: addr=%s, err=%s", s.Addr(), err) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/service_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/service_test.go new file mode 100644 index 000000000..497b12ea3 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/admin/service_test.go @@ -0,0 +1,33 @@ +package admin_test + +import ( + "io/ioutil" + "net/http" + "testing" + + "github.com/influxdb/influxdb/services/admin" +) + +// Ensure service can serve the root index page of the admin. +func TestService_Index(t *testing.T) { + // Start service on random port. + s := admin.NewService(admin.Config{BindAddress: "127.0.0.1:0"}) + if err := s.Open(); err != nil { + t.Fatal(err) + } + defer s.Close() + + // Request root index page. + resp, err := http.Get("http://" + s.Addr().String()) + if err != nil { + t.Fatal(err) + } + defer resp.Body.Close() + + // Validate status code and body. + if resp.StatusCode != http.StatusOK { + t.Fatalf("unexpected status: %d", resp.StatusCode) + } else if _, err := ioutil.ReadAll(resp.Body); err != nil { + t.Fatalf("unable to read body: %s", err) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/collectd_test.conf b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/collectd_test.conf new file mode 100644 index 000000000..97cc4cc08 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/collectd_test.conf @@ -0,0 +1,209 @@ +absolute value:ABSOLUTE:0:U +apache_bytes value:DERIVE:0:U +apache_connections value:GAUGE:0:65535 +apache_idle_workers value:GAUGE:0:65535 +apache_requests value:DERIVE:0:U +apache_scoreboard value:GAUGE:0:65535 +ath_nodes value:GAUGE:0:65535 +ath_stat value:DERIVE:0:U +backends value:GAUGE:0:65535 +bitrate value:GAUGE:0:4294967295 +bytes value:GAUGE:0:U +cache_eviction value:DERIVE:0:U +cache_operation value:DERIVE:0:U +cache_ratio value:GAUGE:0:100 +cache_result value:DERIVE:0:U +cache_size value:GAUGE:0:U +charge value:GAUGE:0:U +compression_ratio value:GAUGE:0:2 +compression uncompressed:DERIVE:0:U, compressed:DERIVE:0:U +connections value:DERIVE:0:U +conntrack value:GAUGE:0:4294967295 +contextswitch value:DERIVE:0:U +counter value:COUNTER:U:U +cpufreq value:GAUGE:0:U +cpu value:DERIVE:0:U +current_connections value:GAUGE:0:U +current_sessions value:GAUGE:0:U +current value:GAUGE:U:U +delay value:GAUGE:-1000000:1000000 +derive value:DERIVE:0:U +df_complex value:GAUGE:0:U +df_inodes value:GAUGE:0:U +df used:GAUGE:0:1125899906842623, free:GAUGE:0:1125899906842623 +disk_latency read:GAUGE:0:U, write:GAUGE:0:U +disk_merged read:DERIVE:0:U, write:DERIVE:0:U +disk_octets read:DERIVE:0:U, write:DERIVE:0:U +disk_ops_complex value:DERIVE:0:U +disk_ops read:DERIVE:0:U, write:DERIVE:0:U +disk_time read:DERIVE:0:U, write:DERIVE:0:U +dns_answer value:DERIVE:0:U +dns_notify value:DERIVE:0:U +dns_octets queries:DERIVE:0:U, responses:DERIVE:0:U +dns_opcode value:DERIVE:0:U +dns_qtype_cached value:GAUGE:0:4294967295 +dns_qtype value:DERIVE:0:U +dns_query value:DERIVE:0:U +dns_question value:DERIVE:0:U +dns_rcode value:DERIVE:0:U +dns_reject value:DERIVE:0:U +dns_request value:DERIVE:0:U +dns_resolver value:DERIVE:0:U +dns_response value:DERIVE:0:U +dns_transfer value:DERIVE:0:U +dns_update value:DERIVE:0:U +dns_zops value:DERIVE:0:U +duration seconds:GAUGE:0:U +email_check value:GAUGE:0:U +email_count value:GAUGE:0:U +email_size value:GAUGE:0:U +entropy value:GAUGE:0:4294967295 +fanspeed value:GAUGE:0:U +file_size value:GAUGE:0:U +files value:GAUGE:0:U +flow value:GAUGE:0:U +fork_rate value:DERIVE:0:U +frequency_offset value:GAUGE:-1000000:1000000 +frequency value:GAUGE:0:U +fscache_stat value:DERIVE:0:U +gauge value:GAUGE:U:U +hash_collisions value:DERIVE:0:U +http_request_methods value:DERIVE:0:U +http_requests value:DERIVE:0:U +http_response_codes value:DERIVE:0:U +humidity value:GAUGE:0:100 +if_collisions value:DERIVE:0:U +if_dropped rx:DERIVE:0:U, tx:DERIVE:0:U +if_errors rx:DERIVE:0:U, tx:DERIVE:0:U +if_multicast value:DERIVE:0:U +if_octets rx:DERIVE:0:U, tx:DERIVE:0:U +if_packets rx:DERIVE:0:U, tx:DERIVE:0:U +if_rx_errors value:DERIVE:0:U +if_rx_octets value:DERIVE:0:U +if_tx_errors value:DERIVE:0:U +if_tx_octets value:DERIVE:0:U +invocations value:DERIVE:0:U +io_octets rx:DERIVE:0:U, tx:DERIVE:0:U +io_packets rx:DERIVE:0:U, tx:DERIVE:0:U +ipt_bytes value:DERIVE:0:U +ipt_packets value:DERIVE:0:U +irq value:DERIVE:0:U +latency value:GAUGE:0:U +links value:GAUGE:0:U +load shortterm:GAUGE:0:5000, midterm:GAUGE:0:5000, longterm:GAUGE:0:5000 +md_disks value:GAUGE:0:U +memcached_command value:DERIVE:0:U +memcached_connections value:GAUGE:0:U +memcached_items value:GAUGE:0:U +memcached_octets rx:DERIVE:0:U, tx:DERIVE:0:U +memcached_ops value:DERIVE:0:U +memory value:GAUGE:0:281474976710656 +multimeter value:GAUGE:U:U +mutex_operations value:DERIVE:0:U +mysql_commands value:DERIVE:0:U +mysql_handler value:DERIVE:0:U +mysql_locks value:DERIVE:0:U +mysql_log_position value:DERIVE:0:U +mysql_octets rx:DERIVE:0:U, tx:DERIVE:0:U +nfs_procedure value:DERIVE:0:U +nginx_connections value:GAUGE:0:U +nginx_requests value:DERIVE:0:U +node_octets rx:DERIVE:0:U, tx:DERIVE:0:U +node_rssi value:GAUGE:0:255 +node_stat value:DERIVE:0:U +node_tx_rate value:GAUGE:0:127 +objects value:GAUGE:0:U +operations value:DERIVE:0:U +percent value:GAUGE:0:100.1 +percent_bytes value:GAUGE:0:100.1 +percent_inodes value:GAUGE:0:100.1 +pf_counters value:DERIVE:0:U +pf_limits value:DERIVE:0:U +pf_source value:DERIVE:0:U +pf_states value:GAUGE:0:U +pf_state value:DERIVE:0:U +pg_blks value:DERIVE:0:U +pg_db_size value:GAUGE:0:U +pg_n_tup_c value:DERIVE:0:U +pg_n_tup_g value:GAUGE:0:U +pg_numbackends value:GAUGE:0:U +pg_scan value:DERIVE:0:U +pg_xact value:DERIVE:0:U +ping_droprate value:GAUGE:0:100 +ping_stddev value:GAUGE:0:65535 +ping value:GAUGE:0:65535 +players value:GAUGE:0:1000000 +power value:GAUGE:0:U +protocol_counter value:DERIVE:0:U +ps_code value:GAUGE:0:9223372036854775807 +ps_count processes:GAUGE:0:1000000, threads:GAUGE:0:1000000 +ps_cputime user:DERIVE:0:U, syst:DERIVE:0:U +ps_data value:GAUGE:0:9223372036854775807 +ps_disk_octets read:DERIVE:0:U, write:DERIVE:0:U +ps_disk_ops read:DERIVE:0:U, write:DERIVE:0:U +ps_pagefaults minflt:DERIVE:0:U, majflt:DERIVE:0:U +ps_rss value:GAUGE:0:9223372036854775807 +ps_stacksize value:GAUGE:0:9223372036854775807 +ps_state value:GAUGE:0:65535 +ps_vm value:GAUGE:0:9223372036854775807 +queue_length value:GAUGE:0:U +records value:GAUGE:0:U +requests value:GAUGE:0:U +response_time value:GAUGE:0:U +response_code value:GAUGE:0:U +route_etx value:GAUGE:0:U +route_metric value:GAUGE:0:U +routes value:GAUGE:0:U +serial_octets rx:DERIVE:0:U, tx:DERIVE:0:U +signal_noise value:GAUGE:U:0 +signal_power value:GAUGE:U:0 +signal_quality value:GAUGE:0:U +snr value:GAUGE:0:U +spam_check value:GAUGE:0:U +spam_score value:GAUGE:U:U +spl value:GAUGE:U:U +swap_io value:DERIVE:0:U +swap value:GAUGE:0:1099511627776 +tcp_connections value:GAUGE:0:4294967295 +temperature value:GAUGE:U:U +threads value:GAUGE:0:U +time_dispersion value:GAUGE:-1000000:1000000 +timeleft value:GAUGE:0:U +time_offset value:GAUGE:-1000000:1000000 +total_bytes value:DERIVE:0:U +total_connections value:DERIVE:0:U +total_objects value:DERIVE:0:U +total_operations value:DERIVE:0:U +total_requests value:DERIVE:0:U +total_sessions value:DERIVE:0:U +total_threads value:DERIVE:0:U +total_time_in_ms value:DERIVE:0:U +total_values value:DERIVE:0:U +uptime value:GAUGE:0:4294967295 +users value:GAUGE:0:65535 +vcl value:GAUGE:0:65535 +vcpu value:GAUGE:0:U +virt_cpu_total value:DERIVE:0:U +virt_vcpu value:DERIVE:0:U +vmpage_action value:DERIVE:0:U +vmpage_faults minflt:DERIVE:0:U, majflt:DERIVE:0:U +vmpage_io in:DERIVE:0:U, out:DERIVE:0:U +vmpage_number value:GAUGE:0:4294967295 +volatile_changes value:GAUGE:0:U +voltage_threshold value:GAUGE:U:U, threshold:GAUGE:U:U +voltage value:GAUGE:U:U +vs_memory value:GAUGE:0:9223372036854775807 +vs_processes value:GAUGE:0:65535 +vs_threads value:GAUGE:0:65535 + +# +# Legacy types +# (required for the v5 upgrade target) +# +arc_counts demand_data:COUNTER:0:U, demand_metadata:COUNTER:0:U, prefetch_data:COUNTER:0:U, prefetch_metadata:COUNTER:0:U +arc_l2_bytes read:COUNTER:0:U, write:COUNTER:0:U +arc_l2_size value:GAUGE:0:U +arc_ratio value:GAUGE:0:U +arc_size current:GAUGE:0:U, target:GAUGE:0:U, minlimit:GAUGE:0:U, maxlimit:GAUGE:0:U +mysql_qcache hits:COUNTER:0:U, inserts:COUNTER:0:U, not_cached:COUNTER:0:U, lowmem_prunes:COUNTER:0:U, queries_in_cache:GAUGE:0:U +mysql_threads running:GAUGE:0:U, connected:GAUGE:0:U, cached:GAUGE:0:U, created:COUNTER:0:U diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/config.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/config.go new file mode 100644 index 000000000..3129c3378 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/config.go @@ -0,0 +1,44 @@ +package collectd + +import ( + "time" + + "github.com/influxdb/influxdb/toml" +) + +const ( + DefaultBindAddress = ":25826" + + DefaultDatabase = "collectd" + + DefaultRetentionPolicy = "" + + DefaultBatchSize = 5000 + + DefaultBatchDuration = toml.Duration(10 * time.Second) + + DefaultTypesDB = "/usr/share/collectd/types.db" +) + +// Config represents a configuration for the collectd service. +type Config struct { + Enabled bool `toml:"enabled"` + BindAddress string `toml:"bind-address"` + Database string `toml:"database"` + RetentionPolicy string `toml:"retention-policy"` + BatchSize int `toml:"batch-size"` + BatchDuration toml.Duration `toml:"batch-timeout"` + TypesDB string `toml:"typesdb"` +} + +// NewConfig returns a new instance of Config with defaults. +func NewConfig() Config { + return Config{ + BindAddress: DefaultBindAddress, + Database: DefaultDatabase, + RetentionPolicy: DefaultRetentionPolicy, + BatchSize: DefaultBatchSize, + BatchDuration: DefaultBatchDuration, + TypesDB: DefaultTypesDB, + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/config_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/config_test.go new file mode 100644 index 000000000..c419dcfa9 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/config_test.go @@ -0,0 +1,32 @@ +package collectd_test + +import ( + "testing" + + "github.com/BurntSushi/toml" + "github.com/influxdb/influxdb/services/collectd" +) + +func TestConfig_Parse(t *testing.T) { + // Parse configuration. + var c collectd.Config + if _, err := toml.Decode(` +enabled = true +bind-address = ":9000" +database = "xxx" +typesdb = "yyy" +`, &c); err != nil { + t.Fatal(err) + } + + // Validate configuration. + if c.Enabled != true { + t.Fatalf("unexpected enabled: %v", c.Enabled) + } else if c.BindAddress != ":9000" { + t.Fatalf("unexpected bind address: %s", c.BindAddress) + } else if c.Database != "xxx" { + t.Fatalf("unexpected database: %s", c.Database) + } else if c.TypesDB != "yyy" { + t.Fatalf("unexpected types db: %s", c.TypesDB) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/service.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/service.go new file mode 100644 index 000000000..b0fee0bc4 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/service.go @@ -0,0 +1,278 @@ +package collectd + +import ( + "fmt" + "log" + "net" + "os" + "sync" + "time" + + "github.com/influxdb/influxdb/cluster" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/tsdb" + "github.com/kimor79/gollectd" +) + +const leaderWaitTimeout = 30 * time.Second + +// pointsWriter is an internal interface to make testing easier. +type pointsWriter interface { + WritePoints(p *cluster.WritePointsRequest) error +} + +// metaStore is an internal interface to make testing easier. +type metaStore interface { + WaitForLeader(d time.Duration) error + CreateDatabaseIfNotExists(name string) (*meta.DatabaseInfo, error) +} + +// Service represents a UDP server which receives metrics in collectd's binary +// protocol and stores them in InfluxDB. +type Service struct { + Config *Config + MetaStore metaStore + PointsWriter pointsWriter + Logger *log.Logger + + wg sync.WaitGroup + err chan error + stop chan struct{} + ln *net.UDPConn + batcher *tsdb.PointBatcher + typesdb gollectd.Types + addr net.Addr +} + +// NewService returns a new instance of the collectd service. +func NewService(c Config) *Service { + s := &Service{ + Config: &c, + Logger: log.New(os.Stderr, "[collectd] ", log.LstdFlags), + err: make(chan error), + } + + return s +} + +// Open starts the service. +func (s *Service) Open() error { + s.Logger.Printf("Starting collectd service") + + if s.Config.BindAddress == "" { + return fmt.Errorf("bind address is blank") + } else if s.Config.Database == "" { + return fmt.Errorf("database name is blank") + } else if s.PointsWriter == nil { + return fmt.Errorf("PointsWriter is nil") + } + + if err := s.MetaStore.WaitForLeader(leaderWaitTimeout); err != nil { + s.Logger.Printf("Failed to detect a cluster leader: %s", err.Error()) + return err + } + + if _, err := s.MetaStore.CreateDatabaseIfNotExists(s.Config.Database); err != nil { + s.Logger.Printf("Failed to ensure target database %s exists: %s", s.Config.Database, err.Error()) + return err + } + + if s.typesdb == nil { + // Open collectd types. + typesdb, err := gollectd.TypesDBFile(s.Config.TypesDB) + if err != nil { + return fmt.Errorf("Open(): %s", err) + } + s.typesdb = typesdb + } + + // Resolve our address. + addr, err := net.ResolveUDPAddr("udp", s.Config.BindAddress) + if err != nil { + return fmt.Errorf("unable to resolve UDP address: %s", err) + } + s.addr = addr + + // Start listening + ln, err := net.ListenUDP("udp", addr) + if err != nil { + return fmt.Errorf("unable to listen on UDP: %s", err) + } + s.ln = ln + + s.Logger.Println("Listening on UDP: ", ln.LocalAddr().String()) + + // Start the points batcher. + s.batcher = tsdb.NewPointBatcher(s.Config.BatchSize, time.Duration(s.Config.BatchDuration)) + s.batcher.Start() + + // Create channel and wait group for signalling goroutines to stop. + s.stop = make(chan struct{}) + s.wg.Add(2) + + // Start goroutines that process collectd packets. + go s.serve() + go s.writePoints() + + return nil +} + +// Close stops the service. +func (s *Service) Close() error { + // Close the connection, and wait for the goroutine to exit. + if s.stop != nil { + close(s.stop) + } + if s.ln != nil { + s.ln.Close() + } + if s.batcher != nil { + s.batcher.Stop() + } + s.wg.Wait() + + // Release all remaining resources. + s.stop = nil + s.ln = nil + s.batcher = nil + s.Logger.Println("collectd UDP closed") + return nil +} + +// SetLogger sets the internal logger to the logger passed in. +func (s *Service) SetLogger(l *log.Logger) { + s.Logger = l +} + +// SetTypes sets collectd types db. +func (s *Service) SetTypes(types string) (err error) { + s.typesdb, err = gollectd.TypesDB([]byte(types)) + return +} + +// Err returns a channel for fatal errors that occur on go routines. +func (s *Service) Err() chan error { return s.err } + +// Addr returns the listener's address. Returns nil if listener is closed. +func (s *Service) Addr() net.Addr { + return s.ln.LocalAddr() +} + +func (s *Service) serve() { + defer s.wg.Done() + + // From https://collectd.org/wiki/index.php/Binary_protocol + // 1024 bytes (payload only, not including UDP / IP headers) + // In versions 4.0 through 4.7, the receive buffer has a fixed size + // of 1024 bytes. When longer packets are received, the trailing data + // is simply ignored. Since version 4.8, the buffer size can be + // configured. Version 5.0 will increase the default buffer size to + // 1452 bytes (the maximum payload size when using UDP/IPv6 over + // Ethernet). + buffer := make([]byte, 1452) + + for { + select { + case <-s.stop: + // We closed the connection, time to go. + return + default: + // Keep processing. + } + + n, _, err := s.ln.ReadFromUDP(buffer) + if err != nil { + s.Logger.Printf("collectd ReadFromUDP error: %s", err) + continue + } + if n > 0 { + s.handleMessage(buffer[:n]) + } + } +} + +func (s *Service) handleMessage(buffer []byte) { + packets, err := gollectd.Packets(buffer, s.typesdb) + if err != nil { + s.Logger.Printf("Collectd parse error: %s", err) + return + } + for _, packet := range *packets { + points := Unmarshal(&packet) + for _, p := range points { + s.batcher.In() <- p + } + } +} + +func (s *Service) writePoints() { + defer s.wg.Done() + + for { + select { + case <-s.stop: + return + case batch := <-s.batcher.Out(): + req := &cluster.WritePointsRequest{ + Database: s.Config.Database, + RetentionPolicy: s.Config.RetentionPolicy, + ConsistencyLevel: cluster.ConsistencyLevelAny, + Points: batch, + } + if err := s.PointsWriter.WritePoints(req); err != nil { + s.Logger.Printf("failed to write batch: %s", err) + continue + } + } + } +} + +// Unmarshal translates a collectd packet into InfluxDB data points. +func Unmarshal(packet *gollectd.Packet) []tsdb.Point { + // Prefer high resolution timestamp. + var timestamp time.Time + if packet.TimeHR > 0 { + // TimeHR is "near" nanosecond measurement, but not exactly nanasecond time + // Since we store time in microseconds, we round here (mostly so tests will work easier) + sec := packet.TimeHR >> 30 + // Shifting, masking, and dividing by 1 billion to get nanoseconds. + nsec := ((packet.TimeHR & 0x3FFFFFFF) << 30) / 1000 / 1000 / 1000 + timestamp = time.Unix(int64(sec), int64(nsec)).UTC().Round(time.Microsecond) + } else { + // If we don't have high resolution time, fall back to basic unix time + timestamp = time.Unix(int64(packet.Time), 0).UTC() + } + + var points []tsdb.Point + for i := range packet.Values { + name := fmt.Sprintf("%s_%s", packet.Plugin, packet.Values[i].Name) + tags := make(map[string]string) + fields := make(map[string]interface{}) + + fields["value"] = packet.Values[i].Value + + if packet.Hostname != "" { + tags["host"] = packet.Hostname + } + if packet.PluginInstance != "" { + tags["instance"] = packet.PluginInstance + } + if packet.Type != "" { + tags["type"] = packet.Type + } + if packet.TypeInstance != "" { + tags["type_instance"] = packet.TypeInstance + } + p := tsdb.NewPoint(name, tags, fields, timestamp) + + points = append(points, p) + } + return points +} + +// assert will panic with a given formatted message if the given condition is false. +func assert(condition bool, msg string, v ...interface{}) { + if !condition { + panic(fmt.Sprintf("assert failed: "+msg, v...)) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/service_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/service_test.go new file mode 100644 index 000000000..a24c9d29f --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/service_test.go @@ -0,0 +1,501 @@ +package collectd + +import ( + "encoding/hex" + "errors" + "io/ioutil" + "log" + "net" + "testing" + "time" + + "github.com/influxdb/influxdb/cluster" + "github.com/influxdb/influxdb/meta" + "github.com/influxdb/influxdb/toml" + "github.com/influxdb/influxdb/tsdb" +) + +// Test that the service checks / creates the target database on startup. +func TestService_CreatesDatabase(t *testing.T) { + t.Parallel() + + s := newTestService(1, time.Second) + + createDatabaseCalled := false + + ms := &testMetaStore{} + ms.CreateDatabaseIfNotExistsFn = func(name string) (*meta.DatabaseInfo, error) { + if name != s.Config.Database { + t.Errorf("\n\texp = %s\n\tgot = %s\n", s.Config.Database, name) + } + createDatabaseCalled = true + return nil, nil + } + s.Service.MetaStore = ms + + s.Open() + s.Close() + + if !createDatabaseCalled { + t.Errorf("CreateDatabaseIfNotExists should have been called when the service opened.") + } +} + +// Test that the collectd service correctly batches points by BatchSize. +func TestService_BatchSize(t *testing.T) { + t.Parallel() + + totalPoints := len(expPoints) + + // Batch sizes that totalTestPoints divide evenly by. + batchSizes := []int{1, 2, 13} + + for _, batchSize := range batchSizes { + func() { + s := newTestService(batchSize, time.Second) + + pointCh := make(chan tsdb.Point) + s.MetaStore.CreateDatabaseIfNotExistsFn = func(name string) (*meta.DatabaseInfo, error) { return nil, nil } + s.PointsWriter.WritePointsFn = func(req *cluster.WritePointsRequest) error { + if len(req.Points) != batchSize { + t.Errorf("\n\texp = %d\n\tgot = %d\n", batchSize, len(req.Points)) + } + + for _, p := range req.Points { + pointCh <- p + } + return nil + } + + if err := s.Open(); err != nil { + t.Fatal(err) + } + defer func() { t.Log("closing service"); s.Close() }() + + // Get the address & port the service is listening on for collectd data. + addr := s.Addr() + conn, err := net.Dial("udp", addr.String()) + if err != nil { + t.Fatal(err) + } + + // Send the test data to the service. + if n, err := conn.Write(testData); err != nil { + t.Fatal(err) + } else if n != len(testData) { + t.Fatalf("only sent %d of %d bytes", n, len(testData)) + } + + points := []tsdb.Point{} + Loop: + for { + select { + case p := <-pointCh: + points = append(points, p) + if len(points) == totalPoints { + break Loop + } + case <-time.After(time.Second): + t.Logf("exp %d points, got %d", totalPoints, len(points)) + t.Fatal("timed out waiting for points from collectd service") + } + } + + if len(points) != totalPoints { + t.Fatalf("exp %d points, got %d", totalPoints, len(points)) + } + + for i, exp := range expPoints { + got := points[i].String() + if got != exp { + t.Fatalf("\n\texp = %s\n\tgot = %s\n", exp, got) + } + } + }() + } +} + +// Test that the collectd service correctly batches points using BatchDuration. +func TestService_BatchDuration(t *testing.T) { + t.Parallel() + + totalPoints := len(expPoints) + + s := newTestService(5000, 250*time.Millisecond) + + pointCh := make(chan tsdb.Point, 1000) + s.MetaStore.CreateDatabaseIfNotExistsFn = func(name string) (*meta.DatabaseInfo, error) { return nil, nil } + s.PointsWriter.WritePointsFn = func(req *cluster.WritePointsRequest) error { + for _, p := range req.Points { + pointCh <- p + } + return nil + } + + if err := s.Open(); err != nil { + t.Fatal(err) + } + defer func() { t.Log("closing service"); s.Close() }() + + // Get the address & port the service is listening on for collectd data. + addr := s.Addr() + conn, err := net.Dial("udp", addr.String()) + if err != nil { + t.Fatal(err) + } + + // Send the test data to the service. + if n, err := conn.Write(testData); err != nil { + t.Fatal(err) + } else if n != len(testData) { + t.Fatalf("only sent %d of %d bytes", n, len(testData)) + } + + points := []tsdb.Point{} +Loop: + for { + select { + case p := <-pointCh: + points = append(points, p) + if len(points) == totalPoints { + break Loop + } + case <-time.After(time.Second): + t.Logf("exp %d points, got %d", totalPoints, len(points)) + t.Fatal("timed out waiting for points from collectd service") + } + } + + if len(points) != totalPoints { + t.Fatalf("exp %d points, got %d", totalPoints, len(points)) + } + + for i, exp := range expPoints { + got := points[i].String() + if got != exp { + t.Fatalf("\n\texp = %s\n\tgot = %s\n", exp, got) + } + } +} + +type testService struct { + *Service + MetaStore testMetaStore + PointsWriter testPointsWriter +} + +func newTestService(batchSize int, batchDuration time.Duration) *testService { + s := &testService{ + Service: NewService(Config{ + BindAddress: "127.0.0.1:0", + Database: "collectd_test", + BatchSize: batchSize, + BatchDuration: toml.Duration(batchDuration), + }), + } + s.Service.PointsWriter = &s.PointsWriter + s.Service.MetaStore = &s.MetaStore + + // Set the collectd types using test string. + if err := s.SetTypes(typesDBText); err != nil { + panic(err) + } + + if !testing.Verbose() { + s.Logger = log.New(ioutil.Discard, "", log.LstdFlags) + } + + return s +} + +type testPointsWriter struct { + WritePointsFn func(*cluster.WritePointsRequest) error +} + +func (w *testPointsWriter) WritePoints(p *cluster.WritePointsRequest) error { + return w.WritePointsFn(p) +} + +type testMetaStore struct { + CreateDatabaseIfNotExistsFn func(name string) (*meta.DatabaseInfo, error) + //DatabaseFn func(name string) (*meta.DatabaseInfo, error) +} + +func (ms *testMetaStore) CreateDatabaseIfNotExists(name string) (*meta.DatabaseInfo, error) { + return ms.CreateDatabaseIfNotExistsFn(name) +} + +func (ms *testMetaStore) WaitForLeader(d time.Duration) error { + return nil +} + +func wait(c chan struct{}, d time.Duration) (err error) { + select { + case <-c: + case <-time.After(d): + err = errors.New("timed out") + } + return +} + +func waitInt(c chan int, d time.Duration) (i int, err error) { + select { + case i = <-c: + case <-time.After(d): + err = errors.New("timed out") + } + return +} + +func check(err error) { + if err != nil { + panic(err) + } +} + +// Raw data sent by collectd, captured using Wireshark. +var testData = func() []byte { + b, err := hex.DecodeString("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") + check(err) + return b +}() + +var expPoints = []string{ + "entropy_value,host=pf1-62-210-94-173,type=entropy value=288.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=1,type=cpu,type_instance=idle value=10908770.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=1,type=cpu,type_instance=wait value=0.0 1414080767000000000", + "df_used,host=pf1-62-210-94-173,type=df,type_instance=live-cow value=378576896.0 1414080767000000000", + "df_free,host=pf1-62-210-94-173,type=df,type_instance=live-cow value=50287988736.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=1,type=cpu,type_instance=interrupt value=254.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=1,type=cpu,type_instance=softirq value=0.0 1414080767000000000", + "df_used,host=pf1-62-210-94-173,type=df,type_instance=live value=0.0 1414080767000000000", + "df_free,host=pf1-62-210-94-173,type=df,type_instance=live value=50666565632.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=1,type=cpu,type_instance=steal value=0.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=2,type=cpu,type_instance=user value=24374.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=2,type=cpu,type_instance=nice value=2776.0 1414080767000000000", + "interface_rx,host=pf1-62-210-94-173,type=if_octets,type_instance=dummy0 value=0.0 1414080767000000000", + "interface_tx,host=pf1-62-210-94-173,type=if_octets,type_instance=dummy0 value=1050.0 1414080767000000000", + "df_used,host=pf1-62-210-94-173,type=df,type_instance=tmp value=73728.0 1414080767000000000", + "df_free,host=pf1-62-210-94-173,type=df,type_instance=tmp value=50666491904.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=2,type=cpu,type_instance=system value=17875.0 1414080767000000000", + "interface_rx,host=pf1-62-210-94-173,type=if_packets,type_instance=dummy0 value=0.0 1414080767000000000", + "interface_tx,host=pf1-62-210-94-173,type=if_packets,type_instance=dummy0 value=15.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=2,type=cpu,type_instance=idle value=10904704.0 1414080767000000000", + "df_used,host=pf1-62-210-94-173,type=df,type_instance=run-lock value=0.0 1414080767000000000", + "df_free,host=pf1-62-210-94-173,type=df,type_instance=run-lock value=5242880.0 1414080767000000000", + "interface_rx,host=pf1-62-210-94-173,type=if_errors,type_instance=dummy0 value=0.0 1414080767000000000", + "interface_tx,host=pf1-62-210-94-173,type=if_errors,type_instance=dummy0 value=0.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=2,type=cpu,type_instance=wait value=0.0 1414080767000000000", + "cpu_value,host=pf1-62-210-94-173,instance=2,type=cpu,type_instance=interrupt value=306.0 1414080767000000000", +} + +// Taken from /usr/share/collectd/types.db on a Ubuntu system +var typesDBText = ` +absolute value:ABSOLUTE:0:U +apache_bytes value:DERIVE:0:U +apache_connections value:GAUGE:0:65535 +apache_idle_workers value:GAUGE:0:65535 +apache_requests value:DERIVE:0:U +apache_scoreboard value:GAUGE:0:65535 +ath_nodes value:GAUGE:0:65535 +ath_stat value:DERIVE:0:U +backends value:GAUGE:0:65535 +bitrate value:GAUGE:0:4294967295 +bytes value:GAUGE:0:U +cache_eviction value:DERIVE:0:U +cache_operation value:DERIVE:0:U +cache_ratio value:GAUGE:0:100 +cache_result value:DERIVE:0:U +cache_size value:GAUGE:0:4294967295 +charge value:GAUGE:0:U +compression_ratio value:GAUGE:0:2 +compression uncompressed:DERIVE:0:U, compressed:DERIVE:0:U +connections value:DERIVE:0:U +conntrack value:GAUGE:0:4294967295 +contextswitch value:DERIVE:0:U +counter value:COUNTER:U:U +cpufreq value:GAUGE:0:U +cpu value:DERIVE:0:U +current_connections value:GAUGE:0:U +current_sessions value:GAUGE:0:U +current value:GAUGE:U:U +delay value:GAUGE:-1000000:1000000 +derive value:DERIVE:0:U +df_complex value:GAUGE:0:U +df_inodes value:GAUGE:0:U +df used:GAUGE:0:1125899906842623, free:GAUGE:0:1125899906842623 +disk_latency read:GAUGE:0:U, write:GAUGE:0:U +disk_merged read:DERIVE:0:U, write:DERIVE:0:U +disk_octets read:DERIVE:0:U, write:DERIVE:0:U +disk_ops_complex value:DERIVE:0:U +disk_ops read:DERIVE:0:U, write:DERIVE:0:U +disk_time read:DERIVE:0:U, write:DERIVE:0:U +dns_answer value:DERIVE:0:U +dns_notify value:DERIVE:0:U +dns_octets queries:DERIVE:0:U, responses:DERIVE:0:U +dns_opcode value:DERIVE:0:U +dns_qtype_cached value:GAUGE:0:4294967295 +dns_qtype value:DERIVE:0:U +dns_query value:DERIVE:0:U +dns_question value:DERIVE:0:U +dns_rcode value:DERIVE:0:U +dns_reject value:DERIVE:0:U +dns_request value:DERIVE:0:U +dns_resolver value:DERIVE:0:U +dns_response value:DERIVE:0:U +dns_transfer value:DERIVE:0:U +dns_update value:DERIVE:0:U +dns_zops value:DERIVE:0:U +duration seconds:GAUGE:0:U +email_check value:GAUGE:0:U +email_count value:GAUGE:0:U +email_size value:GAUGE:0:U +entropy value:GAUGE:0:4294967295 +fanspeed value:GAUGE:0:U +file_size value:GAUGE:0:U +files value:GAUGE:0:U +fork_rate value:DERIVE:0:U +frequency_offset value:GAUGE:-1000000:1000000 +frequency value:GAUGE:0:U +fscache_stat value:DERIVE:0:U +gauge value:GAUGE:U:U +hash_collisions value:DERIVE:0:U +http_request_methods value:DERIVE:0:U +http_requests value:DERIVE:0:U +http_response_codes value:DERIVE:0:U +humidity value:GAUGE:0:100 +if_collisions value:DERIVE:0:U +if_dropped rx:DERIVE:0:U, tx:DERIVE:0:U +if_errors rx:DERIVE:0:U, tx:DERIVE:0:U +if_multicast value:DERIVE:0:U +if_octets rx:DERIVE:0:U, tx:DERIVE:0:U +if_packets rx:DERIVE:0:U, tx:DERIVE:0:U +if_rx_errors value:DERIVE:0:U +if_rx_octets value:DERIVE:0:U +if_tx_errors value:DERIVE:0:U +if_tx_octets value:DERIVE:0:U +invocations value:DERIVE:0:U +io_octets rx:DERIVE:0:U, tx:DERIVE:0:U +io_packets rx:DERIVE:0:U, tx:DERIVE:0:U +ipt_bytes value:DERIVE:0:U +ipt_packets value:DERIVE:0:U +irq value:DERIVE:0:U +latency value:GAUGE:0:U +links value:GAUGE:0:U +load shortterm:GAUGE:0:5000, midterm:GAUGE:0:5000, longterm:GAUGE:0:5000 +md_disks value:GAUGE:0:U +memcached_command value:DERIVE:0:U +memcached_connections value:GAUGE:0:U +memcached_items value:GAUGE:0:U +memcached_octets rx:DERIVE:0:U, tx:DERIVE:0:U +memcached_ops value:DERIVE:0:U +memory value:GAUGE:0:281474976710656 +multimeter value:GAUGE:U:U +mutex_operations value:DERIVE:0:U +mysql_commands value:DERIVE:0:U +mysql_handler value:DERIVE:0:U +mysql_locks value:DERIVE:0:U +mysql_log_position value:DERIVE:0:U +mysql_octets rx:DERIVE:0:U, tx:DERIVE:0:U +nfs_procedure value:DERIVE:0:U +nginx_connections value:GAUGE:0:U +nginx_requests value:DERIVE:0:U +node_octets rx:DERIVE:0:U, tx:DERIVE:0:U +node_rssi value:GAUGE:0:255 +node_stat value:DERIVE:0:U +node_tx_rate value:GAUGE:0:127 +objects value:GAUGE:0:U +operations value:DERIVE:0:U +percent value:GAUGE:0:100.1 +percent_bytes value:GAUGE:0:100.1 +percent_inodes value:GAUGE:0:100.1 +pf_counters value:DERIVE:0:U +pf_limits value:DERIVE:0:U +pf_source value:DERIVE:0:U +pf_states value:GAUGE:0:U +pf_state value:DERIVE:0:U +pg_blks value:DERIVE:0:U +pg_db_size value:GAUGE:0:U +pg_n_tup_c value:DERIVE:0:U +pg_n_tup_g value:GAUGE:0:U +pg_numbackends value:GAUGE:0:U +pg_scan value:DERIVE:0:U +pg_xact value:DERIVE:0:U +ping_droprate value:GAUGE:0:100 +ping_stddev value:GAUGE:0:65535 +ping value:GAUGE:0:65535 +players value:GAUGE:0:1000000 +power value:GAUGE:0:U +protocol_counter value:DERIVE:0:U +ps_code value:GAUGE:0:9223372036854775807 +ps_count processes:GAUGE:0:1000000, threads:GAUGE:0:1000000 +ps_cputime user:DERIVE:0:U, syst:DERIVE:0:U +ps_data value:GAUGE:0:9223372036854775807 +ps_disk_octets read:DERIVE:0:U, write:DERIVE:0:U +ps_disk_ops read:DERIVE:0:U, write:DERIVE:0:U +ps_pagefaults minflt:DERIVE:0:U, majflt:DERIVE:0:U +ps_rss value:GAUGE:0:9223372036854775807 +ps_stacksize value:GAUGE:0:9223372036854775807 +ps_state value:GAUGE:0:65535 +ps_vm value:GAUGE:0:9223372036854775807 +queue_length value:GAUGE:0:U +records value:GAUGE:0:U +requests value:GAUGE:0:U +response_time value:GAUGE:0:U +response_code value:GAUGE:0:U +route_etx value:GAUGE:0:U +route_metric value:GAUGE:0:U +routes value:GAUGE:0:U +serial_octets rx:DERIVE:0:U, tx:DERIVE:0:U +signal_noise value:GAUGE:U:0 +signal_power value:GAUGE:U:0 +signal_quality value:GAUGE:0:U +snr value:GAUGE:0:U +spam_check value:GAUGE:0:U +spam_score value:GAUGE:U:U +spl value:GAUGE:U:U +swap_io value:DERIVE:0:U +swap value:GAUGE:0:1099511627776 +tcp_connections value:GAUGE:0:4294967295 +temperature value:GAUGE:U:U +threads value:GAUGE:0:U +time_dispersion value:GAUGE:-1000000:1000000 +timeleft value:GAUGE:0:U +time_offset value:GAUGE:-1000000:1000000 +total_bytes value:DERIVE:0:U +total_connections value:DERIVE:0:U +total_objects value:DERIVE:0:U +total_operations value:DERIVE:0:U +total_requests value:DERIVE:0:U +total_sessions value:DERIVE:0:U +total_threads value:DERIVE:0:U +total_time_in_ms value:DERIVE:0:U +total_values value:DERIVE:0:U +uptime value:GAUGE:0:4294967295 +users value:GAUGE:0:65535 +vcl value:GAUGE:0:65535 +vcpu value:GAUGE:0:U +virt_cpu_total value:DERIVE:0:U +virt_vcpu value:DERIVE:0:U +vmpage_action value:DERIVE:0:U +vmpage_faults minflt:DERIVE:0:U, majflt:DERIVE:0:U +vmpage_io in:DERIVE:0:U, out:DERIVE:0:U +vmpage_number value:GAUGE:0:4294967295 +volatile_changes value:GAUGE:0:U +voltage_threshold value:GAUGE:U:U, threshold:GAUGE:U:U +voltage value:GAUGE:U:U +vs_memory value:GAUGE:0:9223372036854775807 +vs_processes value:GAUGE:0:65535 +vs_threads value:GAUGE:0:65535 +# +# Legacy types +# (required for the v5 upgrade target) +# +arc_counts demand_data:COUNTER:0:U, demand_metadata:COUNTER:0:U, prefetch_data:COUNTER:0:U, prefetch_metadata:COUNTER:0:U +arc_l2_bytes read:COUNTER:0:U, write:COUNTER:0:U +arc_l2_size value:GAUGE:0:U +arc_ratio value:GAUGE:0:U +arc_size current:GAUGE:0:U, target:GAUGE:0:U, minlimit:GAUGE:0:U, maxlimit:GAUGE:0:U +mysql_qcache hits:COUNTER:0:U, inserts:COUNTER:0:U, not_cached:COUNTER:0:U, lowmem_prunes:COUNTER:0:U, queries_in_cache:GAUGE:0:U +mysql_threads running:GAUGE:0:U, connected:GAUGE:0:U, cached:GAUGE:0:U, created:COUNTER:0:U +` diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/test_client/README.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/test_client/README.md new file mode 100644 index 000000000..90de2b2b6 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/test_client/README.md @@ -0,0 +1,3 @@ +collectD Client +============ +This directory contains code for generating collectd load. diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/test_client/client.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/test_client/client.go new file mode 100644 index 000000000..790f5e871 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/collectd/test_client/client.go @@ -0,0 +1,71 @@ +package main + +import ( + "collectd.org/api" + "collectd.org/network" + + "flag" + "fmt" + "math/rand" + "os" + "strconv" + "time" +) + +var nMeasurments = flag.Int("m", 1, "Number of measurements") +var tagVariance = flag.Int("v", 1, "Number of values per tag. Client is fixed at one tag") +var rate = flag.Int("r", 1, "Number of points per second") +var total = flag.Int("t", -1, "Total number of points to send (default is no limit)") +var host = flag.String("u", "127.0.0.1:25826", "Destination host in the form host:port") + +func main() { + flag.Parse() + + conn, err := network.Dial(*host, network.ClientOptions{}) + if err != nil { + fmt.Println(err) + os.Exit(1) + } + defer conn.Close() + + rateLimiter := make(chan int, *rate) + + go func() { + ticker := time.NewTicker(time.Second) + for { + select { + case <-ticker.C: + for i := 0; i < *rate; i++ { + rateLimiter <- i + } + } + } + }() + + nSent := 0 + for { + if nSent >= *total && *total > 0 { + break + } + <-rateLimiter + + vl := api.ValueList{ + Identifier: api.Identifier{ + Host: "tagvalue" + strconv.Itoa(int(rand.Int31n(int32(*tagVariance)))), + Plugin: "golang" + strconv.Itoa(int(rand.Int31n(int32(*nMeasurments)))), + Type: "gauge", + }, + Time: time.Now(), + Interval: 10 * time.Second, + Values: []api.Value{api.Gauge(42.0)}, + } + if err := conn.Write(vl); err != nil { + fmt.Println(err) + os.Exit(1) + } + conn.Flush() + nSent = nSent + 1 + } + + fmt.Println("Number of points sent:", nSent) +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/config.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/config.go new file mode 100644 index 000000000..1549a816b --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/config.go @@ -0,0 +1,65 @@ +package continuous_querier + +import ( + "time" + + "github.com/influxdb/influxdb/toml" +) + +const ( + DefaultRecomputePreviousN = 2 + + DefaultRecomputeNoOlderThan = 10 * time.Minute + + DefaultComputeRunsPerInterval = 10 + + DefaultComputeNoMoreThan = 2 * time.Minute +) + +// Config represents a configuration for the continuous query service. +type Config struct { + // Enables logging in CQ service to display when CQ's are processed and how many points are wrote. + LogEnabled bool `toml:"log-enabled"` + + // If this flag is set to false, both the brokers and data nodes should ignore any CQ processing. + Enabled bool `toml:"enabled"` + + // when continuous queries are run we'll automatically recompute previous intervals + // in case lagged data came in. Set to zero if you never have lagged data. We do + // it this way because invalidating previously computed intervals would be insanely hard + // and expensive. + RecomputePreviousN int `toml:"recompute-previous-n"` + + // The RecomputePreviousN setting provides guidance for how far back to recompute, the RecomputeNoOlderThan + // setting sets a ceiling on how far back in time it will go. For example, if you have 2 PreviousN + // and have this set to 10m, then we'd only compute the previous two intervals for any + // CQs that have a group by time <= 5m. For all others, we'd only recompute the previous window + RecomputeNoOlderThan toml.Duration `toml:"recompute-no-older-than"` + + // ComputeRunsPerInterval will determine how many times the current and previous N intervals + // will be computed. The group by time will be divided by this and it will get computed this many times: + // group by time seconds / runs per interval + // This will give partial results for current group by intervals and will determine how long it will + // be until lagged data is recomputed. For example, if this number is 10 and the group by time is 10m, it + // will be a minute past the previous 10m bucket of time before lagged data is picked up + ComputeRunsPerInterval int `toml:"compute-runs-per-interval"` + + // ComputeNoMoreThan paired with the RunsPerInterval will determine the ceiling of how many times smaller + // group by times will be computed. For example, if you have RunsPerInterval set to 10 and this setting + // to 1m. Then for a group by time(1m) will actually only get computed once per interval (and once per PreviousN). + // If you have a group by time(5m) then you'll get five computes per interval. Any group by time window larger + // than 10m will get computed 10 times for each interval. + ComputeNoMoreThan toml.Duration `toml:"compute-no-more-than"` +} + +// NewConfig returns a new instance of Config with defaults. +func NewConfig() Config { + return Config{ + LogEnabled: true, + Enabled: true, + RecomputePreviousN: DefaultRecomputePreviousN, + RecomputeNoOlderThan: toml.Duration(DefaultRecomputeNoOlderThan), + ComputeRunsPerInterval: DefaultComputeRunsPerInterval, + ComputeNoMoreThan: toml.Duration(DefaultComputeNoMoreThan), + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/config_test.go b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/config_test.go new file mode 100644 index 000000000..2a0edc4f2 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/config_test.go @@ -0,0 +1,36 @@ +package continuous_querier_test + +import ( + "testing" + "time" + + "github.com/BurntSushi/toml" + "github.com/influxdb/influxdb/services/continuous_querier" +) + +func TestConfig_Parse(t *testing.T) { + // Parse configuration. + var c continuous_querier.Config + if _, err := toml.Decode(` +recompute-previous-n = 1 +recompute-no-older-than = "10s" +compute-runs-per-interval = 2 +compute-no-more-than = "20s" +enabled = true +`, &c); err != nil { + t.Fatal(err) + } + + // Validate configuration. + if c.RecomputePreviousN != 1 { + t.Fatalf("unexpected recompute previous n: %d", c.RecomputePreviousN) + } else if time.Duration(c.RecomputeNoOlderThan) != 10*time.Second { + t.Fatalf("unexpected recompute no older than: %v", c.RecomputeNoOlderThan) + } else if c.ComputeRunsPerInterval != 2 { + t.Fatalf("unexpected compute runs per interval: %d", c.ComputeRunsPerInterval) + } else if time.Duration(c.ComputeNoMoreThan) != 20*time.Second { + t.Fatalf("unexpected compute no more than: %v", c.ComputeNoMoreThan) + } else if c.Enabled != true { + t.Fatalf("unexpected enabled: %v", c.Enabled) + } +} diff --git a/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/continuous_queries.md b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/continuous_queries.md new file mode 100644 index 000000000..bd14b5161 --- /dev/null +++ b/Godeps/_workspace/src/github.com/influxdb/influxdb/services/continuous_querier/continuous_queries.md @@ -0,0 +1,236 @@ +# Continuous Queries + +This document lays out continuous queries and a proposed architecture for how they'll work within an InfluxDB cluster. + +## Definition of Continuous Queries + +Continuous queries serve two purposes in InfluxDB: + +1. Combining many series into a single series (i.e. removing 1 or more tag dimensions to make queries more efficient) +2. Aggregating and downsampling series + +The purpose of both types of continuous queries is to duplicate or downsample data automatically in the background to make querying thier results fast and efficient. Think of them as another way to create indexes on data. + +Generally, there are continuous queries that create copyies of data into another measurement or tagset and queries that downsample and aggregate data. The only difference between the two types is if the query has a `GROUP BY time` clause. + +Before we get to the continuous query examples, we need to define the `INTO` syntax of queries. + +### INTO + +`INTO` is a method for running a query and having it output into either another measurement name, retention policy, or database. The syntax looks like this: + +```sql +SELECT * +INTO [.] [ON ] +FROM +[WHERE ...] +[GROUP BY ...] +``` + +The syntax states that the retention policy, database, where clause, and group by clause are all optional. If a retention policy isn't specified, the database's default retention policy will be written into. If the database isn't specified, the database the query is running from will be written into. + +By selecting specific fields, `INTO` can merge many series into one that will go into a new either a new measurement, retention policy, or database. For example: + +```sql +SELECT mean(value) as value, region +INTO "1h.cpu_load" +FROM cpu_load +GROUP BY time(1h), region +``` + +That will give 1h summaries of the mean value of the `cpu_load` for each `region`. Specifying `region` in the `GROUP BY` clause is unnecessary since having it in the `SELECT` clause forces it to be grouped by that tag, we've just included it in the example for clarity. + +With `SELECT ... INTO`, fields will be written as fields and tags will be written as tags. + +### Continuous Query Syntax + +The `INTO` queries run once. Continuous queries will turn `INTO` queries into something that run in the background in the cluster. They're kind of like triggers in SQL. + +```sql +CREATE CONTINUOUS QUERY "1h_cpu_load" +ON database_name +BEGIN + SELECT mean(value) as value, region + INTO "1h.cpu_load" + FROM cpu_load + GROUP BY time(1h), region +END +``` + +Or chain them together: + +```sql +CREATE CONTINUOUS QUERY "10m_event_count" +ON database_name +BEGIN + SELECT count(value) + INTO "10m.events" + FROM events + GROUP BY time(10m) +END + +-- this selects from the output of one continuous query and outputs to another series +CREATE CONTINUOUS QUERY "1h_event_count" +ON database_name +BEGIN + SELECT sum(count) as count + INTO "1h.events" + FROM events + GROUP BY time(1h) +END +``` + +Or multiple aggregations from all series in a measurement. This example assumes you have a retention policy named `1h`. + +```sql +CREATE CONTINUOUS QUERY "1h_cpu_load" +ON database_name +BEGIN + SELECT mean(value), percentile(80, value) as percentile_80, percentile(95, value) as percentile_95 + INTO "1h.cpu_load" + FROM cpu_load + GROUP BY time(1h), * +END +``` + +The `GROUP BY *` indicates that we want to group by the tagset of the points written in. The same tags will be written to the output series. The multiple aggregates in the `SELECT` clause (percentile, mean) will be written in as fields to the resulting series. + +Showing what continuous queries we have: + +```sql +LIST CONTINUOUS QUERIES +``` + +Dropping continuous queries: + +```sql +DROP CONTINUOUS QUERY +ON +``` + +### Security + +To create or drop a continuous query, the user must be an admin. + +### Limitations + +In order to prevent cycles and endless copying of data, the following limitation is enforced on continuous queries at create time: + +*The output of a continuous query must go to either a different measurement or to a different retention policy.* + +In theory they'd still be able to create a cycle with multiple continuous queries. We should check for these and disallow. + +## Proposed Architecture + +Continuous queries should be stored in the metastore cluster wide. That is, they amount to database schema that should be stored in every server in a cluster. + +Continuous queries will have to be handled in a different way for two different use cases: those that simply copy data (CQs without a group by time) and those that aggregate and downsample data (those with a group by time). + +### No group by time + +For CQs that have no `GROUP BY time` clause, they should be evaluated at the data node as part of the write. The single write should create any other writes for the CQ and submit those in the same request to the brokers to ensure that all writes succeed (both the original and the new CQ writes) or none do. + +I imagine the process going something like this: + +1. Convert the data point into its compact form `