Refactor to use AggregatingOutput

Metrics that are unserializable will be logged at debug level, but the
rest of the batch will be sent.  Unserializable metrics can occur during
normal operation such as if you remove all fields from a metric or the
metric cannot fit within the line size limit.

.circleci/config.yml

@@ -0,0 +1,106 @@
+---
+defaults:
+  defaults: &defaults
+    working_directory: '/go/src/github.com/influxdata/telegraf'
+  go-1_8: &go-1_8
+    docker:
+      - image: 'circleci/golang:1.8.7'
+  go-1_9: &go-1_9
+    docker:
+      - image: 'circleci/golang:1.9.5'
+  go-1_10: &go-1_10
+    docker:
+      - image: 'circleci/golang:1.10.1'
+
+version: 2
+jobs:
+  deps:
+    <<: [ *defaults, *go-1_10 ]
+    steps:
+      - checkout
+      - run: 'make deps'
+      - persist_to_workspace:
+          root: '/go/src'
+          paths:
+            - '*'
+  test-go-1.8:
+    <<: [ *defaults, *go-1_8 ]
+    steps:
+      - attach_workspace:
+          at: '/go/src'
+      - run: 'make test-ci'
+  test-go-1.9:
+    <<: [ *defaults, *go-1_9 ]
+    steps:
+      - attach_workspace:
+          at: '/go/src'
+      - run: 'make test-ci'
+  test-go-1.10:
+    <<: [ *defaults, *go-1_10 ]
+    steps:
+      - attach_workspace:
+          at: '/go/src'
+      - run: 'make test-ci'
+      - run: 'GOARCH=386 make test-ci'
+  release:
+    <<: [ *defaults, *go-1_10 ]
+    steps:
+      - attach_workspace:
+          at: '/go/src'
+      - run: './scripts/release.sh'
+      - store_artifacts:
+          path: './artifacts'
+          destination: '.'
+  nightly:
+    <<: [ *defaults, *go-1_10 ]
+    steps:
+      - attach_workspace:
+          at: '/go/src'
+      - run: './scripts/release.sh'
+      - store_artifacts:
+          path: './artifacts'
+          destination: '.'
+
+workflows:
+  version: 2
+  build_and_release:
+    jobs:
+      - 'deps'
+      - 'test-go-1.8':
+          requires:
+            - 'deps'
+      - 'test-go-1.9':
+          requires:
+            - 'deps'
+      - 'test-go-1.10':
+          requires:
+            - 'deps'
+      - 'release':
+          requires:
+            - 'test-go-1.8'
+            - 'test-go-1.9'
+            - 'test-go-1.10'
+  nightly:
+    jobs:
+      - 'deps'
+      - 'test-go-1.8':
+          requires:
+            - 'deps'
+      - 'test-go-1.9':
+          requires:
+            - 'deps'
+      - 'test-go-1.10':
+          requires:
+            - 'deps'
+      - 'nightly':
+          requires:
+            - 'test-go-1.8'
+            - 'test-go-1.9'
+            - 'test-go-1.10'
+    triggers:
+      - schedule:
+          cron: "0 7 * * *"
+          filters:
+            branches:
+              only:
+                - master

.github/ISSUE_TEMPLATE.md

@@ -1,7 +1,7 @@
 ## Directions
 
 GitHub Issues are reserved for actionable bug reports and feature requests.
-General questions should be sent to the [InfluxDB mailing list](https://groups.google.com/forum/#!forum/influxdb).
+General questions should be asked at the [InfluxData Community](https://community.influxdata.com) site.
 
 Before opening an issue, search for similar bug reports or feature requests on GitHub Issues.
 If no similar issue can be found, fill out either the "Bug Report" or the "Feature Request" section below.

.github/PULL_REQUEST_TEMPLATE.md

@@ -1,5 +1,5 @@
 ### Required for all PRs:
 
-- [ ] CHANGELOG.md updated (we recommend not updating this until the PR has been approved by a maintainer)
-- [ ] Sign [CLA](https://influxdata.com/community/cla/) (if not already signed)
-- [ ] README.md updated (if adding a new plugin)
+- [ ] Signed [CLA](https://influxdata.com/community/cla/).
+- [ ] Associated README.md updated.
+- [ ] Has appropriate unit tests.

.gitignore

@@ -1,7 +1,4 @@
-build
-tivan
-.vagrant
+/build
 /telegraf
-.idea
-*~
-*#
+/telegraf.exe
+/telegraf.gz

CHANGELOG.md

@@ -1,4 +1,661 @@
-## v1.2 [unreleased]
+## v1.7 [unreleased]
+
+### Release Notes
+
+- The `cassandra` input plugin has been deprecated in favor of the `jolokia2`
+  input plugin which is much more configurable and more performant.  There is
+  an [example configuration](./plugins/inputs/jolokia2/examples) to help you
+  get started.
+
+### New Inputs
+
+- [fibaro](./plugins/inputs/fibaro/README.md) - Contributed by @dynek
+- [mcrouter](./plugins/inputs/mcrouter/README.md) - Contributed by @cthayer
+- [nvidia_smi](./plugins/inputs/nvidia_smi/README.md) - Contributed by @jackzampolin
+
+### Features
+
+- [#3964](https://github.com/influxdata/telegraf/pull/3964): Add repl_oplog_window_sec metric to mongodb input.
+- [#3819](https://github.com/influxdata/telegraf/pull/3819): Add per-host shard metrics in mongodb input.
+- [#3999](https://github.com/influxdata/telegraf/pull/3999): Skip files with leading `..` in config directory.
+- [#4021](https://github.com/influxdata/telegraf/pull/4021): Add TLS support to socket_writer and socket_listener plugins.
+- [#4025](https://github.com/influxdata/telegraf/pull/4025): Add snmp input option to strip non fixed length index suffixes.
+- [#4035](https://github.com/influxdata/telegraf/pull/4035): Add server version tag to docker input.
+- [#4044](https://github.com/influxdata/telegraf/pull/4044): Add support for LeoFS 1.4 to leofs input.
+- [#4068](https://github.com/influxdata/telegraf/pull/4068): Add parameter to force the interval of gather for sysstat.
+- [#3877](https://github.com/influxdata/telegraf/pull/3877): Support busybox ping in the ping input.
+- [#4077](https://github.com/influxdata/telegraf/pull/4077): Add input plugin for McRouter.
+
+### Bugfixes
+
+- [#4018](https://github.com/influxdata/telegraf/pull/4018): Write to working file outputs if any files are not writeable.
+- [#4036](https://github.com/influxdata/telegraf/pull/4036): Add all win_perf_counters fields for a series in a single metric.
+
+## v1.6.2 [unreleased]
+
+### Bugfixes
+
+- [#4078](https://github.com/influxdata/telegraf/pull/4078): Use same timestamp for fields in system input.
+
+## v1.6.1 [2018-04-23]
+
+### Bugfixes
+
+- [#3835](https://github.com/influxdata/telegraf/issues/3835): Report mem input fields as gauges instead counters.
+- [#4030](https://github.com/influxdata/telegraf/issues/4030): Fix graphite outputs unsigned integers in wrong format.
+- [#4043](https://github.com/influxdata/telegraf/issues/4043): Report available fields if utmp is unreadable.
+- [#4039](https://github.com/influxdata/telegraf/issues/4039): Fix potential "no fields" error writing to outputs.
+- [#4037](https://github.com/influxdata/telegraf/issues/4037): Fix uptime reporting in system input when ran inside docker.
+- [#3750](https://github.com/influxdata/telegraf/issues/3750): Fix mem input "cannot allocate memory" error on FreeBSD based systems.
+- [#4056](https://github.com/influxdata/telegraf/pull/4056): Fix duplicate tags when overriding an existing tag.
+- [#4062](https://github.com/influxdata/telegraf/pull/4062): Add server argument as first argument in unbound input.
+- [#4063](https://github.com/influxdata/telegraf/issues/4063): Fix handling of floats with multiple leading zeroes.
+- [#4064](https://github.com/influxdata/telegraf/issues/4064): Return errors in mongodb SSL/TLS configuration.
+
+## v1.6 [2018-04-16]
+
+### Release Notes
+
+- The `mysql` input plugin has been updated fix a number of type convertion
+  issues.  This may cause a `field type error` when inserting into InfluxDB due
+  the change of types.
+
+  To address this we have introduced a new `metric_version` option to control
+  enabling the new format.  For in depth recommendations on upgrading please
+  reference the [mysql plugin documentation](./plugins/inputs/mysql/README.md#metric-version).
+
+  It is encouraged to migrate to the new model when possible as the old version
+  is deprecated and will be removed in a future version.
+
+- The `postgresql` plugins now defaults to using a persistent connection to the database.
+  In environments where TCP connections are terminated the `max_lifetime`
+  setting should be set less than the collection `interval` to prevent errors.
+
+- The `sqlserver` input plugin has a new query and data model that can be enabled
+  by setting `query_version = 2`.  It is encouraged to migrate to the new
+  model when possible as the old version is deprecated and will be removed in
+  a future version.
+
+- An option has been added to the `openldap` input plugin that reverses metric
+  name to improve grouping.  This change is enabled when `reverse_metric_names = true`
+  is set.  It is encouraged to enable this option when possible as the old
+  ordering is deprecated.
+
+- The new `http` input configured with `data_format = "json"` can perform the
+  same task as the, now deprecated, `httpjson` input.
+
+
+### New Inputs
+
+- [http](./plugins/inputs/http/README.md) - Thanks to @grange74
+- [ipset](./plugins/inputs/ipset/README.md) - Thanks to @sajoupa
+- [nats](./plugins/inputs/nats/README.md) - Thanks to @mjs & @levex
+
+### New Processors
+
+- [override](./plugins/processors/override/README.md) - Thanks to @KarstenSchnitter
+
+### New Parsers
+
+- [dropwizard](./docs/DATA_FORMATS_INPUT.md#dropwizard) - Thanks to @atzoum
+
+### Features
+
+- [#3551](https://github.com/influxdata/telegraf/pull/3551): Add health status mapping from string to int in elasticsearch input.
+- [#3580](https://github.com/influxdata/telegraf/pull/3580): Add control over which stats to gather in basicstats aggregator.
+- [#3596](https://github.com/influxdata/telegraf/pull/3596): Add messages_delivered_get to rabbitmq input.
+- [#3632](https://github.com/influxdata/telegraf/pull/3632): Add wired field to mem input.
+- [#3619](https://github.com/influxdata/telegraf/pull/3619): Add support for gathering exchange metrics to the rabbitmq input.
+- [#3565](https://github.com/influxdata/telegraf/pull/3565): Add support for additional metrics on Linux in zfs input.
+- [#3524](https://github.com/influxdata/telegraf/pull/3524): Add available_entropy field to kernel input plugin.
+- [#3643](https://github.com/influxdata/telegraf/pull/3643): Add user privilege level setting to IPMI sensors.
+- [#2701](https://github.com/influxdata/telegraf/pull/2701): Use persistent connection to postgresql database.
+- [#2846](https://github.com/influxdata/telegraf/pull/2846): Add support for dropwizard input format.
+- [#3666](https://github.com/influxdata/telegraf/pull/3666): Add container health metrics to docker input.
+- [#3687](https://github.com/influxdata/telegraf/pull/3687): Add support for using globs in devices list of diskio input plugin.
+- [#2754](https://github.com/influxdata/telegraf/pull/2754): Allow running as console application on Windows.
+- [#3703](https://github.com/influxdata/telegraf/pull/3703): Add listener counts and node running status to rabbitmq input.
+- [#3674](https://github.com/influxdata/telegraf/pull/3674): Add NATS Monitoring Input Plugin.
+- [#3702](https://github.com/influxdata/telegraf/pull/3702): Add ability to select which queues will be gathered in rabbitmq input.
+- [#3726](https://github.com/influxdata/telegraf/pull/3726): Add support for setting bsd source address to the ping input.
+- [#3346](https://github.com/influxdata/telegraf/pull/3346): Add Ipset input plugin.
+- [#3719](https://github.com/influxdata/telegraf/pull/3719): Add TLS and HTTP basic auth to prometheus_client output.
+- [#3618](https://github.com/influxdata/telegraf/pull/3618): Add new sqlserver output data model.
+- [#3559](https://github.com/influxdata/telegraf/pull/3559): Add native Go method for finding pids to procstat.
+- [#3722](https://github.com/influxdata/telegraf/pull/3722): Add additional metrics and reverse metric names option to openldap.
+- [#3769](https://github.com/influxdata/telegraf/pull/3769): Add TLS support to the mesos input plugin.
+- [#3546](https://github.com/influxdata/telegraf/pull/3546): Add http input plugin.
+- [#3781](https://github.com/influxdata/telegraf/pull/3781): Add keep alive support to the TCP mode of statsd.
+- [#3783](https://github.com/influxdata/telegraf/pull/3783): Support deadline in ping plugin.
+- [#3765](https://github.com/influxdata/telegraf/pull/3765): Add option to disable labels in prometheus output for string fields.
+- [#3808](https://github.com/influxdata/telegraf/pull/3808): Add shard server stats to the mongodb input plugin.
+- [#3713](https://github.com/influxdata/telegraf/pull/3713): Add server option to unbound plugin.
+- [#3804](https://github.com/influxdata/telegraf/pull/3804): Convert boolean metric values to float in datadog output.
+- [#3799](https://github.com/influxdata/telegraf/pull/3799): Add Solr 3 compatibility.
+- [#3797](https://github.com/influxdata/telegraf/pull/3797): Add sum stat to basicstats aggregator.
+- [#3626](https://github.com/influxdata/telegraf/pull/3626): Add ability to override proxy from environment in http response.
+- [#3853](https://github.com/influxdata/telegraf/pull/3853): Add host to ping timeout log message.
+- [#3773](https://github.com/influxdata/telegraf/pull/3773): Add override processor.
+- [#3814](https://github.com/influxdata/telegraf/pull/3814): Add status_code and result tags and result_type field to http_response input.
+- [#3880](https://github.com/influxdata/telegraf/pull/3880): Added config flag to skip collection of network protocol metrics.
+- [#3927](https://github.com/influxdata/telegraf/pull/3927): Add TLS support to kapacitor input.
+- [#3496](https://github.com/influxdata/telegraf/pull/3496): Add HTTP basic auth support to the http_listener input.
+- [#3452](https://github.com/influxdata/telegraf/issues/3452): Tags in output InfluxDB Line Protocol are now sorted.
+- [#3631](https://github.com/influxdata/telegraf/issues/3631): InfluxDB Line Protocol parser now accepts DOS line endings.
+- [#2496](https://github.com/influxdata/telegraf/issues/2496): An option has been added to skip database creation in the InfluxDB output.
+- [#3366](https://github.com/influxdata/telegraf/issues/3366): Add support for connecting to InfluxDB over a unix domain socket.
+- [#3946](https://github.com/influxdata/telegraf/pull/3946): Add optional unsigned integer support to the influx data format.
+- [#3811](https://github.com/influxdata/telegraf/pull/3811): Add TLS support to zookeeper input.
+- [#2737](https://github.com/influxdata/telegraf/issues/2737): Add filters for container state to docker input.
+
+### Bugfixes
+
+- [#1896](https://github.com/influxdata/telegraf/issues/1896): Fix various mysql data type conversions.
+- [#3810](https://github.com/influxdata/telegraf/issues/3810): Fix metric buffer limit in internal plugin after reload.
+- [#3801](https://github.com/influxdata/telegraf/issues/3801): Fix panic in http_response on invalid regex.
+- [#3973](https://github.com/influxdata/telegraf/issues/3873): Fix socket_listener setting ReadBufferSize on tcp sockets.
+- [#1575](https://github.com/influxdata/telegraf/issues/1575): Add tag for target url to phpfpm input.
+- [#3868](https://github.com/influxdata/telegraf/issues/3868): Fix cannot unmarshal object error in DC/OS input.
+- [#3648](https://github.com/influxdata/telegraf/issues/3648): Fix InfluxDB output not able to reconnect when server address changes.
+- [#3957](https://github.com/influxdata/telegraf/issues/3957): Fix parsing of dos line endings in the smart input.
+- [#3754](https://github.com/influxdata/telegraf/issues/3754): Fix precision truncation when no timestamp included.
+- [#3655](https://github.com/influxdata/telegraf/issues/3655): Fix SNMPv3 connection with Cisco ASA 5515 in snmp input.
+- [#3981](https://github.com/influxdata/telegraf/pull/3981): Export all vars defined in /etc/default/telegraf.
+- [#4004](https://github.com/influxdata/telegraf/issues/4004): Allow grok pattern to contain newlines.
+
+## v1.5.3 [2018-03-14]
+
+### Bugfixes
+
+- [#3729](https://github.com/influxdata/telegraf/issues/3729): Set path to / if HOST_MOUNT_PREFIX matches full path.
+- [#3739](https://github.com/influxdata/telegraf/issues/3739): Remove userinfo from url tag in prometheus input.
+- [#3778](https://github.com/influxdata/telegraf/issues/3778): Fix ping plugin not reporting zero durations.
+- [#3697](https://github.com/influxdata/telegraf/issues/3697): Disable keepalive in mqtt output to prevent deadlock.
+- [#3786](https://github.com/influxdata/telegraf/pull/3786): Fix collation difference in sqlserver input.
+- [#3871](https://github.com/influxdata/telegraf/pull/3871): Fix uptime metric in passenger input plugin.
+- [#3851](https://github.com/influxdata/telegraf/issues/3851): Add output of stderr in case of error to exec log message.
+
+## v1.5.2 [2018-01-30]
+
+### Bugfixes
+
+- [#3684](https://github.com/influxdata/telegraf/pull/3684): Ignore empty lines in Graphite plaintext.
+- [#3604](https://github.com/influxdata/telegraf/issues/3604): Fix index out of bounds error in solr input plugin.
+- [#3680](https://github.com/influxdata/telegraf/pull/3680): Reconnect before sending graphite metrics if disconnected.
+- [#3693](https://github.com/influxdata/telegraf/pull/3693): Align aggregator period with internal ticker to avoid skipping metrics.
+- [#3629](https://github.com/influxdata/telegraf/issues/3629): Fix a potential deadlock when using aggregators.
+- [#3697](https://github.com/influxdata/telegraf/issues/3697): Limit wait time for writes in mqtt output.
+- [#3698](https://github.com/influxdata/telegraf/issues/3698): Revert change in graphite output where dot in field key was replaced by underscore.
+- [#3710](https://github.com/influxdata/telegraf/issues/3710): Add timeout to wavefront output write.
+- [#3725](https://github.com/influxdata/telegraf/issues/3725): Exclude master_replid fields from redis input.
+
+## v1.5.1 [2018-01-10]
+
+### Bugfixes
+
+- [#3624](https://github.com/influxdata/telegraf/pull/3624): Fix name error in jolokia2_agent sample config.
+- [#3625](https://github.com/influxdata/telegraf/pull/3625): Fix DC/OS login expiration time.
+- [#3593](https://github.com/influxdata/telegraf/pull/3593): Set Content-Type charset in influxdb output and allow it be overridden.
+- [#3594](https://github.com/influxdata/telegraf/pull/3594): Document permissions setup for postfix input.
+- [#3633](https://github.com/influxdata/telegraf/pull/3633): Fix deliver_get field in rabbitmq input.
+- [#3607](https://github.com/influxdata/telegraf/issues/3607): Escape environment variables during config toml parsing.
+
+## v1.5 [2017-12-14]
+
+### New Plugins
+- [basicstats](./plugins/aggregators/basicstats/README.md) - Thanks to @toni-moreno
+- [bond](./plugins/inputs/bond/README.md) - Thanks to @ildarsv
+- [cratedb](./plugins/outputs/cratedb/README.md) - Thanks to @felixge
+- [dcos](./plugins/inputs/dcos/README.md) - Thanks to @influxdata
+- [jolokia2](./plugins/inputs/jolokia2/README.md) - Thanks to @dylanmei
+- [nginx_plus](./plugins/inputs/nginx_plus/README.md) - Thanks to @mplonka & @poblahblahblah
+- [opensmtpd](./plugins/inputs/opensmtpd/README.md) - Thanks to @aromeyer
+- [particle](./plugins/inputs/webhooks/particle/README.md) - Thanks to @davidgs
+- [pf](./plugins/inputs/pf/README.md) - Thanks to @nferch
+- [postfix](./plugins/inputs/postfix/README.md) - Thanks to @phemmer
+- [smart](./plugins/inputs/smart/README.md) - Thanks to @rickard-von-essen
+- [solr](./plugins/inputs/solr/README.md) - Thanks to @ljagiello
+- [teamspeak](./plugins/inputs/teamspeak/README.md) - Thanks to @p4ddy1
+- [unbound](./plugins/inputs/unbound/README.md) - Thanks to @aromeyer
+- [wavefront](./plugins/outputs/wavefront/README.md) - Thanks to @puckpuck
+
+### Release Notes
+
+- In the `kinesis` output, use of the `partition_key` and
+  `use_random_partitionkey` options has been deprecated in favor of the
+  `partition` subtable.  This allows for more flexible methods to set the
+  partition key such as by metric name or by tag.
+
+- With the release of the new improved `jolokia2` input, the legacy `jolokia`
+  plugin is deprecated and will be removed in a future release.  Users of this
+  plugin are encouraged to update to the new `jolokia2` plugin.
+
+- In the `postgresql` and `postgresql_extensible` plugins, the type of the oid
+  data type has changed from string to integer.  It is recommended to drop
+  affected fields until a new shard is started. For details on how to
+  workaround this issue please see [#3622](https://github.com/influxdata/telegraf/issues/3622).
+
+### Features
+
+- [#3170](https://github.com/influxdata/telegraf/pull/3170): Add support for sharding based on metric name.
+- [#3196](https://github.com/influxdata/telegraf/pull/3196): Add Kafka output plugin topic_suffix option.
+- [#3027](https://github.com/influxdata/telegraf/pull/3027): Include mount mode option in disk metrics.
+- [#3191](https://github.com/influxdata/telegraf/pull/3191): TLS and MTLS enhancements to HTTPListener input plugin.
+- [#3213](https://github.com/influxdata/telegraf/pull/3213): Add polling method to logparser and tail inputs.
+- [#3211](https://github.com/influxdata/telegraf/pull/3211): Add timeout option for kubernetes input.
+- [#3234](https://github.com/influxdata/telegraf/pull/3234): Add support for timing sums in statsd input.
+- [#2617](https://github.com/influxdata/telegraf/issues/2617): Add resource limit monitoring to procstat.
+- [#3236](https://github.com/influxdata/telegraf/pull/3236): Add support for k8s service DNS discovery to prometheus input.
+- [#3245](https://github.com/influxdata/telegraf/pull/3245): Add configurable metrics endpoint to prometheus output.
+- [#3214](https://github.com/influxdata/telegraf/pull/3214): Add new nginx_plus input plugin.
+- [#3215](https://github.com/influxdata/telegraf/pull/3215): Add support for NSQLookupd to nsq_consumer.
+- [#2278](https://github.com/influxdata/telegraf/pull/2278): Add redesigned Jolokia input plugin.
+- [#3106](https://github.com/influxdata/telegraf/pull/3106): Add configurable separator for metrics and fields in opentsdb output.
+- [#1692](https://github.com/influxdata/telegraf/pull/1692): Add support for the rollbar occurrence webhook event.
+- [#3160](https://github.com/influxdata/telegraf/pull/3160): Add Wavefront output plugin.
+- [#3281](https://github.com/influxdata/telegraf/pull/3281): Add extra wired tiger cache metrics to mongodb input.
+- [#3141](https://github.com/influxdata/telegraf/pull/3141): Collect Docker Swarm service metrics in docker input plugin.
+- [#2449](https://github.com/influxdata/telegraf/pull/2449): Add smart input plugin for collecting S.M.A.R.T. data.
+- [#3269](https://github.com/influxdata/telegraf/pull/3269): Add cluster health level configuration to elasticsearch input.
+- [#3304](https://github.com/influxdata/telegraf/pull/3304): Add ability to limit node stats in elasticsearch input.
+- [#2167](https://github.com/influxdata/telegraf/pull/2167): Add new basicstats aggregator.
+- [#3344](https://github.com/influxdata/telegraf/pull/3344): Add UDP IPv6 support to statsd input.
+- [#3350](https://github.com/influxdata/telegraf/pull/3350): Use labels in prometheus output for string fields.
+- [#3358](https://github.com/influxdata/telegraf/pull/3358): Add support for decimal timestamps to ts-epoch modifier.
+- [#3337](https://github.com/influxdata/telegraf/pull/3337): Add histogram and summary types and use in prometheus plugins.
+- [#3365](https://github.com/influxdata/telegraf/pull/3365): Gather concurrently from snmp agents.
+- [#3333](https://github.com/influxdata/telegraf/issues/3333): Perform DNS lookup before ping and report result.
+- [#3398](https://github.com/influxdata/telegraf/issues/3398): Add instance name option to varnish plugin.
+- [#3406](https://github.com/influxdata/telegraf/pull/3406):  Add support for SSL settings to ElasticSearch output plugin.
+- [#3315](https://github.com/influxdata/telegraf/pull/3315): Add Teamspeak 3 input plugin.
+- [#3305](https://github.com/influxdata/telegraf/pull/3305): Add modification_time field to filestat input plugin.
+- [#2019](https://github.com/influxdata/telegraf/pull/2019): Add Solr input plugin.
+- [#3210](https://github.com/influxdata/telegraf/pull/3210): Add CrateDB output plugin.
+- [#3459](https://github.com/influxdata/telegraf/pull/3459): Add systemd unit pid and cgroup matching to procstat.
+- [#3477](https://github.com/influxdata/telegraf/pull/3477): Add Particle Webhook Plugin.
+- [#3471](https://github.com/influxdata/telegraf/pull/3471): Use MAX() instead of SUM() for latency measurements in sqlserver.
+- [#3490](https://github.com/influxdata/telegraf/pull/3490): Add index by week number to Elasticsearch output.
+- [#3434](https://github.com/influxdata/telegraf/pull/3434): Add unbound input plugin.
+- [#3449](https://github.com/influxdata/telegraf/pull/3449): Add opensmtpd input plugin.
+- [#3470](https://github.com/influxdata/telegraf/pull/3470): Add support for tags in the index name in elasticsearch output.
+- [#2553](https://github.com/influxdata/telegraf/pull/2553): Add postfix input plugin.
+- [#3424](https://github.com/influxdata/telegraf/pull/3424): Add bond input plugin.
+- [#3518](https://github.com/influxdata/telegraf/pull/3518): Add slab to mem plugin.
+- [#3519](https://github.com/influxdata/telegraf/pull/3519): Add input plugin for DC/OS.
+- [#3140](https://github.com/influxdata/telegraf/pull/3140): Add support for glob patterns in net input plugin.
+- [#3405](https://github.com/influxdata/telegraf/pull/3405): Add input plugin for OpenBSD/FreeBSD pf.
+- [#3528](https://github.com/influxdata/telegraf/pull/3528): Add option to amqp output to publish persistent messages.
+- [#3530](https://github.com/influxdata/telegraf/pull/3530): Support I (idle) process state on procfs+Linux.
+
+### Bugfixes
+
+- [#3136](https://github.com/influxdata/telegraf/issues/3136): Fix webhooks input address in use during reload.
+- [#3258](https://github.com/influxdata/telegraf/issues/3258): Unlock Statsd when stopping to prevent deadlock.
+- [#3319](https://github.com/influxdata/telegraf/issues/3319): Fix cloudwatch output requires unneeded permissions.
+- [#3351](https://github.com/influxdata/telegraf/issues/3351): Fix prometheus passthrough for existing value types.
+- [#3430](https://github.com/influxdata/telegraf/issues/3430): Always ignore autofs filesystems in disk input.
+- [#3326](https://github.com/influxdata/telegraf/issues/3326): Fail metrics parsing on unescaped quotes.
+- [#3473](https://github.com/influxdata/telegraf/pull/3473): Whitelist allowed char classes for graphite output.
+- [#3488](https://github.com/influxdata/telegraf/pull/3488): Use hexadecimal ids and lowercase names in zipkin input.
+- [#3263](https://github.com/influxdata/telegraf/issues/3263): Fix snmp-tools output parsing with Windows EOLs.
+- [#3447](https://github.com/influxdata/telegraf/issues/3447): Add shadow-utils dependency to rpm package.
+- [#3448](https://github.com/influxdata/telegraf/issues/3448): Use deb-systemd-invoke to restart service.
+- [#3553](https://github.com/influxdata/telegraf/issues/3553): Fix kafka_consumer outside range of offsets error.
+- [#3568](https://github.com/influxdata/telegraf/issues/3568): Fix separation of multiple prometheus_client outputs.
+- [#3577](https://github.com/influxdata/telegraf/issues/3577): Don't add system input uptime_format as a counter.
+
+## v1.4.5 [2017-12-01]
+
+### Bugfixes
+
+- [#3500](https://github.com/influxdata/telegraf/issues/3500): Fix global variable collection when using interval_slow option in mysql input.
+- [#3486](https://github.com/influxdata/telegraf/issues/3486): Fix error getting net connections info in netstat input.
+- [#3529](https://github.com/influxdata/telegraf/issues/3529): Fix HOST_MOUNT_PREFIX in docker with disk input.
+
+## v1.4.4 [2017-11-08]
+
+### Bugfixes
+
+- [#3401](https://github.com/influxdata/telegraf/pull/3401): Use schema specified in mqtt_consumer input.
+- [#3419](https://github.com/influxdata/telegraf/issues/3419): Redact datadog API key in log output.
+- [#3311](https://github.com/influxdata/telegraf/issues/3311): Fix error getting pids in netstat input.
+- [#3339](https://github.com/influxdata/telegraf/issues/3339): Support HOST_VAR envvar to locate /var in system input.
+- [#3383](https://github.com/influxdata/telegraf/issues/3383): Use current time if docker container read time is zero value.
+
+## v1.4.3 [2017-10-25]
+
+### Bugfixes
+
+- [#3327](https://github.com/influxdata/telegraf/issues/3327): Fix container name filters in docker input.
+- [#3321](https://github.com/influxdata/telegraf/issues/3321): Fix snmpwalk address format in leofs input.
+- [#3329](https://github.com/influxdata/telegraf/issues/3329): Fix case sensitivity issue in sqlserver query.
+- [#3342](https://github.com/influxdata/telegraf/pull/3342): Fix CPU input plugin stuck after suspend on Linux.
+- [#3013](https://github.com/influxdata/telegraf/issues/3013): Fix mongodb input panic when restarting mongodb.
+- [#3224](https://github.com/influxdata/telegraf/pull/3224): Preserve url path prefix in influx output.
+- [#3354](https://github.com/influxdata/telegraf/pull/3354): Fix TELEGRAF_OPTS expansion in systemd service unit.
+- [#3357](https://github.com/influxdata/telegraf/issues/3357): Remove warning when JSON contains null value.
+- [#3375](https://github.com/influxdata/telegraf/issues/3375): Fix ACL token usage in consul input plugin.
+- [#3369](https://github.com/influxdata/telegraf/issues/3369): Fix unquoting error with Tomcat 6.
+- [#3373](https://github.com/influxdata/telegraf/issues/3373): Fix syscall panic in diskio on some Linux systems.
+
+## v1.4.2 [2017-10-10]
+
+### Bugfixes
+
+- [#3259](https://github.com/influxdata/telegraf/issues/3259): Fix error if int larger than 32-bit in /proc/vmstat.
+- [#3265](https://github.com/influxdata/telegraf/issues/3265): Fix parsing of JSON with a UTF8 BOM in httpjson.
+- [#2887](https://github.com/influxdata/telegraf/issues/2887): Allow JSON data format to contain zero metrics.
+- [#3284](https://github.com/influxdata/telegraf/issues/3284): Fix format of connection_timeout in mqtt_consumer.
+- [#3081](https://github.com/influxdata/telegraf/issues/3081): Fix case sensitivity error in sqlserver input.
+- [#3297](https://github.com/influxdata/telegraf/issues/3297): Add support for proxy environment variables to http_response.
+- [#1588](https://github.com/influxdata/telegraf/issues/1588): Add support for standard proxy env vars in outputs.
+- [#3282](https://github.com/influxdata/telegraf/issues/3282): Fix panic in cpu input if number of cpus changes.
+- [#2854](https://github.com/influxdata/telegraf/issues/2854): Use chunked transfer encoding in InfluxDB output.
+
+## v1.4.1 [2017-09-26]
+
+### Bugfixes
+
+- [#3167](https://github.com/influxdata/telegraf/issues/3167): Fix MQTT input exits if Broker is not available on startup.
+- [#3217](https://github.com/influxdata/telegraf/issues/3217): Fix optional field value conversions in fluentd input.
+- [#3227](https://github.com/influxdata/telegraf/issues/3227): Whitelist allowed char classes for opentsdb output.
+- [#3232](https://github.com/influxdata/telegraf/issues/3232): Fix counter and gauge metric types.
+- [#3235](https://github.com/influxdata/telegraf/issues/3235): Fix skipped line with empty target in iptables.
+- [#3175](https://github.com/influxdata/telegraf/issues/3175): Fix duplicate keys in perf counters sqlserver query.
+- [#3230](https://github.com/influxdata/telegraf/issues/3230): Fix panic in statsd p100 calculation.
+- [#3242](https://github.com/influxdata/telegraf/issues/3242): Fix arm64 packages contain 32-bit executable.
+
+## v1.4 [2017-09-05]
+
+### Release Notes
+
+- The `kafka_consumer` input has been updated to support Kafka 0.9 and
+  above style consumer offset handling.  The previous version of this plugin
+  supporting Kafka 0.8 and below is available as the `kafka_consumer_legacy`
+  plugin.
+
+- In the `aerospike` input the `node_name` field has been changed to be a tag
+  for both the `aerospike_node` and `aerospike_namespace` measurements.
+
+- The default prometheus_client port has been changed to 9273.
+
+### New Plugins
+
+- [fail2ban](./plugins/inputs/fail2ban/README.md) - Thanks to @grugrut
+- [fluentd](./plugins/inputs/fluentd/README.md) - Thanks to @DanKans
+- [histogram](./plugins/aggregators/histogram/README.md) - Thanks to @vlamug
+- [minecraft](./plugins/inputs/minecraft/README.md) - Thanks to @adamperlin & @Ayrdrie
+- [openldap](./plugins/inputs/openldap/README.md) - Thanks to @cobaugh
+- [salesforce](./plugins/inputs/salesforce/README.md) - Thanks to @rody
+- [tomcat](./plugins/inputs/tomcat/README.md) - Thanks to @mlindes
+- [win_services](./plugins/inputs/win_services/README.md) - Thanks to @vlastahajek
+- [zipkin](./plugins/inputs/zipkin/README.md) - Thanks to @adamperlin & @Ayrdrie
+
+### Features
+
+- [#2487](https://github.com/influxdata/telegraf/pull/2487): Add Kafka 0.9+ consumer support
+- [#2773](https://github.com/influxdata/telegraf/pull/2773): Add support for self-signed certs to InfluxDB input plugin
+- [#2293](https://github.com/influxdata/telegraf/pull/2293): Add TCP listener for statsd input
+- [#2581](https://github.com/influxdata/telegraf/pull/2581): Add Docker container environment variables as tags. Only whitelisted
+- [#2817](https://github.com/influxdata/telegraf/pull/2817): Add timeout option to IPMI sensor plugin
+- [#2883](https://github.com/influxdata/telegraf/pull/2883): Add support for an optional SSL/TLS configuration to nginx input plugin
+- [#2882](https://github.com/influxdata/telegraf/pull/2882): Add timezone support for logparser timestamps.
+- [#2814](https://github.com/influxdata/telegraf/pull/2814): Add result_type field for http_response input.
+- [#2734](https://github.com/influxdata/telegraf/pull/2734): Add include/exclude filters for docker containers.
+- [#2602](https://github.com/influxdata/telegraf/pull/2602): Add secure connection support to graphite output.
+- [#2908](https://github.com/influxdata/telegraf/pull/2908): Add min/max response time on linux/darwin to ping.
+- [#2929](https://github.com/influxdata/telegraf/pull/2929): Add HTTP Proxy support to influxdb output.
+- [#2933](https://github.com/influxdata/telegraf/pull/2933): Add standard SSL options to mysql input.
+- [#2875](https://github.com/influxdata/telegraf/pull/2875): Add input plugin for fail2ban.
+- [#2924](https://github.com/influxdata/telegraf/pull/2924): Support HOST_PROC in processes and linux_sysctl_fs inputs.
+- [#2960](https://github.com/influxdata/telegraf/pull/2960): Add Minecraft input plugin.
+- [#2963](https://github.com/influxdata/telegraf/pull/2963): Add support for RethinkDB 1.0 handshake protocol.
+- [#2943](https://github.com/influxdata/telegraf/pull/2943): Add optional usage_active and time_active CPU metrics.
+- [#2973](https://github.com/influxdata/telegraf/pull/2973): Change default prometheus_client port.
+- [#2661](https://github.com/influxdata/telegraf/pull/2661): Add fluentd input plugin.
+- [#2990](https://github.com/influxdata/telegraf/pull/2990): Add result_type field to net_response input plugin.
+- [#2571](https://github.com/influxdata/telegraf/pull/2571): Add read timeout to socket_listener
+- [#2612](https://github.com/influxdata/telegraf/pull/2612): Add input plugin for OpenLDAP.
+- [#3042](https://github.com/influxdata/telegraf/pull/3042): Add network option to dns_query.
+- [#3054](https://github.com/influxdata/telegraf/pull/3054): Add redis_version field to redis input.
+- [#3063](https://github.com/influxdata/telegraf/pull/3063): Add tls options to docker input.
+- [#2387](https://github.com/influxdata/telegraf/pull/2387): Add histogram aggregator plugin.
+- [#3080](https://github.com/influxdata/telegraf/pull/3080): Add zipkin input plugin.
+- [#3023](https://github.com/influxdata/telegraf/pull/3023): Add Windows Services input plugin.
+- [#3098](https://github.com/influxdata/telegraf/pull/3098): Add path tag to logparser containing path of logfile.
+- [#3075](https://github.com/influxdata/telegraf/pull/3075): Add salesforce input plugin.
+- [#3097](https://github.com/influxdata/telegraf/pull/3097): Add option to run varnish under sudo.
+- [#3119](https://github.com/influxdata/telegraf/pull/3119): Add weighted_io_time to diskio input.
+- [#2978](https://github.com/influxdata/telegraf/pull/2978): Add gzip content-encoding support to influxdb output.
+- [#3127](https://github.com/influxdata/telegraf/pull/3127): Allow using system plugin in Windows.
+- [#3112](https://github.com/influxdata/telegraf/pull/3112): Add tomcat input plugin.
+- [#3182](https://github.com/influxdata/telegraf/pull/3182): HTTP headers can be added to InfluxDB output.
+
+### Bugfixes
+
+- [#2607](https://github.com/influxdata/telegraf/issues/2607): Improve logging of errors in Cassandra input.
+- [#2819](https://github.com/influxdata/telegraf/pull/2819): [enh] set db_version at 0 if query version fails
+- [#2749](https://github.com/influxdata/telegraf/pull/2749): Fixed sqlserver input to work with case sensitive server collation.
+- [#2716](https://github.com/influxdata/telegraf/pull/2716): Systemd does not see all shutdowns as failures
+- [#2782](https://github.com/influxdata/telegraf/pull/2782): Reuse transports in input plugins
+- [#2815](https://github.com/influxdata/telegraf/issues/2815): Inputs processes fails with "no such process".
+- [#1137](https://github.com/influxdata/telegraf/issues/1137): Fix multiple plugin loading in win_perf_counters.
+- [#2855](https://github.com/influxdata/telegraf/pull/2855):  MySQL input: log and continue on field parse error.
+- [#2885](https://github.com/influxdata/telegraf/pull/2885): Fix timeout option in Windows ping input sample configuration.
+- [#2911](https://github.com/influxdata/telegraf/issues/2911): Fix Kinesis output plugin in govcloud.
+- [#2917](https://github.com/influxdata/telegraf/issues/2917): Fix Aerospike input adds all nodes to a single series.
+- [#2452](https://github.com/influxdata/telegraf/pull/2452): Improve Prometheus Client output documentation.
+- [#2984](https://github.com/influxdata/telegraf/pull/2984): Display error message if prometheus output fails to listen.
+- [#2997](https://github.com/influxdata/telegraf/issues/2997): Fix elasticsearch output content type detection warning.
+- [#2914](https://github.com/influxdata/telegraf/issues/2914): Prevent possible deadlock when using aggregators.
+- [#2860](https://github.com/influxdata/telegraf/issues/2860): Fix combined tagdrop/tagpass filtering.
+- [#3036](https://github.com/influxdata/telegraf/pull/3036): Fix filtering when both pass and drop match an item.
+- [#2964](https://github.com/influxdata/telegraf/issues/2964): Only report cpu usage for online cpus in docker input.
+- [#3050](https://github.com/influxdata/telegraf/pull/3050): Start first aggregator period at startup time.
+- [#2906](https://github.com/influxdata/telegraf/issues/2906): Fix panic in logparser if file cannot be opened.
+- [#2886](https://github.com/influxdata/telegraf/issues/2886): Default to localhost if zookeeper has no servers set.
+- [#2457](https://github.com/influxdata/telegraf/issues/2457): Fix docker memory and cpu reporting in Windows.
+- [#3058](https://github.com/influxdata/telegraf/issues/3058): Allow iptable entries with trailing text.
+- [#1680](https://github.com/influxdata/telegraf/issues/1680): Sanitize password from couchbase metric.
+- [#3104](https://github.com/influxdata/telegraf/issues/3104): Converge to typed value in prometheus output.
+- [#2899](https://github.com/influxdata/telegraf/issues/2899): Skip compilcation of logparser and tail on solaris.
+- [#2951](https://github.com/influxdata/telegraf/issues/2951): Discard logging from tail library.
+- [#3126](https://github.com/influxdata/telegraf/pull/3126): Remove log message on ping timeout.
+- [#3144](https://github.com/influxdata/telegraf/issues/3144): Don't retry points beyond retention policy.
+- [#3015](https://github.com/influxdata/telegraf/issues/3015): Don't start Telegraf on install in Amazon Linux.
+- [#3153](https://github.com/influxdata/telegraf/issues/3053): Enable hddtemp input on all platforms.
+- [#3142](https://github.com/influxdata/telegraf/issues/3142): Escape backslash within string fields.
+- [#3162](https://github.com/influxdata/telegraf/issues/3162): Fix parsing of SHM remotes in ntpq input
+- [#3149](https://github.com/influxdata/telegraf/issues/3149): Don't fail parsing zpool stats if pool health is UNAVAIL on FreeBSD.
+- [#2672](https://github.com/influxdata/telegraf/issues/2672): Fix NSQ input plugin when used with version 1.0.0-compat.
+- [#2523](https://github.com/influxdata/telegraf/issues/2523): Added CloudWatch metric constraint validation.
+- [#3179](https://github.com/influxdata/telegraf/issues/3179): Skip non-numerical values in graphite format.
+- [#3187](https://github.com/influxdata/telegraf/issues/3187): Fix panic when handling string fields with escapes.
+
+## v1.3.5 [2017-07-26]
+
+### Bugfixes
+
+- [#3049](https://github.com/influxdata/telegraf/issues/3049): Fix prometheus output cannot be reloaded.
+- [#3037](https://github.com/influxdata/telegraf/issues/3037): Fix filestat reporting exists when cannot list directory.
+- [#2386](https://github.com/influxdata/telegraf/issues/2386): Fix ntpq parse issue when using dns_lookup.
+- [#2554](https://github.com/influxdata/telegraf/issues/2554): Fix panic when agent.interval = "0s".
+
+## v1.3.4 [2017-07-12]
+
+### Bugfixes
+
+- [#3001](https://github.com/influxdata/telegraf/issues/3001): Fix handling of escape characters within fields.
+- [#2988](https://github.com/influxdata/telegraf/issues/2988): Fix chrony plugin does not track system time offset.
+- [#3004](https://github.com/influxdata/telegraf/issues/3004): Do not allow metrics with trailing slashes.
+- [#3011](https://github.com/influxdata/telegraf/issues/3011): Prevent Write from being called concurrently.
+
+## v1.3.3 [2017-06-28]
+
+### Bugfixes
+
+- [#2915](https://github.com/influxdata/telegraf/issues/2915): Allow dos line endings in tail and logparser.
+- [#2937](https://github.com/influxdata/telegraf/issues/2937): Remove label value sanitization in prometheus output.
+- [#2948](https://github.com/influxdata/telegraf/issues/2948): Fix bug parsing default timestamps with modified precision.
+- [#2954](https://github.com/influxdata/telegraf/issues/2954): Fix panic in elasticsearch input if cannot determine master.
+
+## v1.3.2 [2017-06-14]
+
+### Bugfixes
+
+- [#2862](https://github.com/influxdata/telegraf/issues/2862): Fix InfluxDB UDP metric splitting.
+- [#2888](https://github.com/influxdata/telegraf/issues/2888): Fix mongodb/leofs urls without scheme.
+- [#2822](https://github.com/influxdata/telegraf/issues/2822): Fix inconsistent label dimensions in prometheus output.
+
+## v1.3.1 [2017-05-31]
+
+### Bugfixes
+
+- [#2749](https://github.com/influxdata/telegraf/pull/2749): Fixed sqlserver input to work with case sensitive server collation.
+- [#2782](https://github.com/influxdata/telegraf/pull/2782): Reuse transports in input plugins
+- [#2815](https://github.com/influxdata/telegraf/issues/2815): Inputs processes fails with "no such process".
+- [#2851](https://github.com/influxdata/telegraf/pull/2851): Fix InfluxDB output database quoting.
+- [#2856](https://github.com/influxdata/telegraf/issues/2856): Fix net input on older Linux kernels.
+- [#2848](https://github.com/influxdata/telegraf/pull/2848): Fix panic in mongo input.
+- [#2869](https://github.com/influxdata/telegraf/pull/2869): Fix length calculation of split metric buffer.
+
+## v1.3 [2017-05-15]
+
+### Release Notes
+
+- Users of the windows `ping` plugin will need to drop or migrate their
+measurements in order to continue using the plugin. The reason for this is that
+the windows plugin was outputting a different type than the linux plugin. This
+made it impossible to use the `ping` plugin for both windows and linux
+machines.
+
+- Ceph: the `ceph_pgmap_state` metric content has been modified to use a unique field `count`, with each state expressed as a `state` tag.
+
+Telegraf < 1.3:
+
+```
+# field_name             value
+active+clean             123
+active+clean+scrubbing   3
+```
+
+Telegraf >= 1.3:
+
+```
+# field_name    value       tag
+count           123         state=active+clean
+count           3           state=active+clean+scrubbing
+```
+
+- The [Riemann output plugin](./plugins/outputs/riemann) has been rewritten
+and the previous riemann plugin is _incompatible_ with the new one. The reasons
+for this are outlined in issue [#1878](https://github.com/influxdata/telegraf/issues/1878).
+The previous riemann output will still be available using
+`outputs.riemann_legacy` if needed, but that will eventually be deprecated.
+It is highly recommended that all users migrate to the new riemann output plugin.
+
+- Generic [socket_listener](./plugins/inputs/socket_listener) and
+[socket_writer](./plugins/outputs/socket_writer) plugins have been implemented
+for receiving and sending UDP, TCP, unix, & unix-datagram data. These plugins
+will replace udp_listener and tcp_listener, which are still available but will
+be deprecated eventually.
+
+### Features
+
+- [#2721](https://github.com/influxdata/telegraf/pull/2721): Added SASL options for kafka output plugin.
+- [#2723](https://github.com/influxdata/telegraf/pull/2723): Added SSL configuration for input haproxy.
+- [#2494](https://github.com/influxdata/telegraf/pull/2494): Add interrupts input plugin.
+- [#2094](https://github.com/influxdata/telegraf/pull/2094): Add generic socket listener & writer.
+- [#2204](https://github.com/influxdata/telegraf/pull/2204): Extend http_response to support searching for a substring in response. Return 1 if found, else 0.
+- [#2137](https://github.com/influxdata/telegraf/pull/2137): Added userstats to mysql input plugin.
+- [#2179](https://github.com/influxdata/telegraf/pull/2179): Added more InnoDB metric to MySQL plugin.
+- [#2229](https://github.com/influxdata/telegraf/pull/2229): `ceph_pgmap_state` metric now uses a single field `count`, with PG state published as `state` tag.
+- [#2251](https://github.com/influxdata/telegraf/pull/2251): InfluxDB output: use own client for improved through-put and less allocations.
+- [#2330](https://github.com/influxdata/telegraf/pull/2330): Keep -config-directory when running as Windows service.
+- [#1900](https://github.com/influxdata/telegraf/pull/1900): Riemann plugin rewrite.
+- [#1453](https://github.com/influxdata/telegraf/pull/1453): diskio: add support for name templates and udev tags.
+- [#2277](https://github.com/influxdata/telegraf/pull/2277): add integer metrics for Consul check health state.
+- [#2201](https://github.com/influxdata/telegraf/pull/2201): Add lock option to the IPtables input plugin.
+- [#2244](https://github.com/influxdata/telegraf/pull/2244): Support ipmi_sensor plugin querying local ipmi sensors.
+- [#2339](https://github.com/influxdata/telegraf/pull/2339): Increment gather_errors for all errors emitted by inputs.
+- [#2071](https://github.com/influxdata/telegraf/issues/2071): Use official docker SDK.
+- [#1678](https://github.com/influxdata/telegraf/pull/1678): Add AMQP consumer input plugin
+- [#2512](https://github.com/influxdata/telegraf/pull/2512): Added pprof tool.
+- [#2501](https://github.com/influxdata/telegraf/pull/2501): Support DEAD(X) state in system input plugin.
+- [#2522](https://github.com/influxdata/telegraf/pull/2522): Add support for mongodb client certificates.
+- [#1948](https://github.com/influxdata/telegraf/pull/1948): Support adding SNMP table indexes as tags.
+- [#2332](https://github.com/influxdata/telegraf/pull/2332): Add Elasticsearch 5.x output
+- [#2587](https://github.com/influxdata/telegraf/pull/2587): Add json timestamp units configurability
+- [#2597](https://github.com/influxdata/telegraf/issues/2597): Add support for Linux sysctl-fs metrics.
+- [#2425](https://github.com/influxdata/telegraf/pull/2425): Support to include/exclude docker container labels as tags
+- [#1667](https://github.com/influxdata/telegraf/pull/1667): dmcache input plugin
+- [#2637](https://github.com/influxdata/telegraf/issues/2637): Add support for precision in http_listener
+- [#2636](https://github.com/influxdata/telegraf/pull/2636): Add `message_len_max` option to `kafka_consumer` input
+- [#1100](https://github.com/influxdata/telegraf/issues/1100): Add collectd parser
+- [#1820](https://github.com/influxdata/telegraf/issues/1820): easier plugin testing without outputs
+- [#2493](https://github.com/influxdata/telegraf/pull/2493): Check signature in the GitHub webhook plugin
+- [#2038](https://github.com/influxdata/telegraf/issues/2038): Add papertrail support to webhooks
+- [#2253](https://github.com/influxdata/telegraf/pull/2253): Change jolokia plugin to use bulk requests.
+- [#2575](https://github.com/influxdata/telegraf/issues/2575) Add diskio input for Darwin
+- [#2705](https://github.com/influxdata/telegraf/pull/2705): Kinesis output: add use_random_partitionkey option
+- [#2635](https://github.com/influxdata/telegraf/issues/2635): add tcp keep-alive to socket_listener & socket_writer
+- [#2031](https://github.com/influxdata/telegraf/pull/2031): Add Kapacitor input plugin
+- [#2732](https://github.com/influxdata/telegraf/pull/2732): Use go 1.8.1
+- [#2712](https://github.com/influxdata/telegraf/issues/2712): Documentation for rabbitmq input plugin
+- [#2141](https://github.com/influxdata/telegraf/pull/2141): Logparser handles newly-created files.
+
+### Bugfixes
+
+- [#2633](https://github.com/influxdata/telegraf/pull/2633): ipmi_sensor: allow @ symbol in password
+- [#2077](https://github.com/influxdata/telegraf/issues/2077): SQL Server Input - Arithmetic overflow error converting numeric to data type int.
+- [#2262](https://github.com/influxdata/telegraf/issues/2262): Flush jitter can inhibit metric collection.
+- [#2318](https://github.com/influxdata/telegraf/issues/2318): haproxy input - Add missing fields.
+- [#2287](https://github.com/influxdata/telegraf/issues/2287): Kubernetes input: Handle null startTime for stopped pods.
+- [#2356](https://github.com/influxdata/telegraf/issues/2356): cpu input panic when /proc/stat is empty.
+- [#2341](https://github.com/influxdata/telegraf/issues/2341): telegraf swallowing panics in --test mode.
+- [#2358](https://github.com/influxdata/telegraf/pull/2358): Create pidfile with 644 permissions & defer file deletion.
+- [#2360](https://github.com/influxdata/telegraf/pull/2360): Fixed install/remove of telegraf on non-systemd Debian/Ubuntu systems
+- [#2282](https://github.com/influxdata/telegraf/issues/2282): Reloading telegraf freezes prometheus output.
+- [#2390](https://github.com/influxdata/telegraf/issues/2390): Empty tag value causes error on InfluxDB output.
+- [#2380](https://github.com/influxdata/telegraf/issues/2380): buffer_size field value is negative number from "internal" plugin.
+- [#2414](https://github.com/influxdata/telegraf/issues/2414): Missing error handling in the MySQL plugin leads to segmentation violation.
+- [#2462](https://github.com/influxdata/telegraf/pull/2462): Fix type conflict in windows ping plugin.
+- [#2178](https://github.com/influxdata/telegraf/issues/2178): logparser: regexp with lookahead.
+- [#2466](https://github.com/influxdata/telegraf/issues/2466): Telegraf can crash in LoadDirectory on 0600 files.
+- [#2215](https://github.com/influxdata/telegraf/issues/2215): Iptables input: document better that rules without a comment are ignored.
+- [#2483](https://github.com/influxdata/telegraf/pull/2483): Fix win_perf_counters capping values at 100.
+- [#2498](https://github.com/influxdata/telegraf/pull/2498): Exporting Ipmi.Path to be set by config.
+- [#2500](https://github.com/influxdata/telegraf/pull/2500): Remove warning if parse empty content
+- [#2520](https://github.com/influxdata/telegraf/pull/2520): Update default value for Cloudwatch rate limit
+- [#2513](https://github.com/influxdata/telegraf/issues/2513): create /etc/telegraf/telegraf.d directory in tarball.
+- [#2541](https://github.com/influxdata/telegraf/issues/2541): Return error on unsupported serializer data format.
+- [#1827](https://github.com/influxdata/telegraf/issues/1827): Fix Windows Performance Counters multi instance identifier
+- [#2576](https://github.com/influxdata/telegraf/pull/2576): Add write timeout to Riemann output
+- [#2596](https://github.com/influxdata/telegraf/pull/2596): fix timestamp parsing on prometheus plugin
+- [#2610](https://github.com/influxdata/telegraf/pull/2610): Fix deadlock when output cannot write
+- [#2410](https://github.com/influxdata/telegraf/issues/2410): Fix connection leak in postgresql.
+- [#2628](https://github.com/influxdata/telegraf/issues/2628): Set default measurement name for snmp input.
+- [#2649](https://github.com/influxdata/telegraf/pull/2649): Improve performance of diskio with many disks
+- [#2671](https://github.com/influxdata/telegraf/issues/2671): The internal input plugin uses the wrong units for `heap_objects`
+- [#2684](https://github.com/influxdata/telegraf/pull/2684): Fix ipmi_sensor config is shared between all plugin instances
+- [#2450](https://github.com/influxdata/telegraf/issues/2450): Network statistics not collected when system has alias interfaces
+- [#1911](https://github.com/influxdata/telegraf/issues/1911): Sysstat plugin needs LANG=C or similar locale
+- [#2528](https://github.com/influxdata/telegraf/issues/2528): File output closes standard streams on reload.
+- [#2603](https://github.com/influxdata/telegraf/issues/2603): AMQP output disconnect blocks all outputs
+- [#2706](https://github.com/influxdata/telegraf/issues/2706): Improve documentation for redis input plugin
+
+## v1.2.1 [2017-02-01]
+
+### Bugfixes
+
+- [#2317](https://github.com/influxdata/telegraf/issues/2317): Fix segfault on nil metrics with influxdb output.
+- [#2324](https://github.com/influxdata/telegraf/issues/2324): Fix negative number handling.
+
+### Features
+
+- [#2348](https://github.com/influxdata/telegraf/pull/2348): Go version 1.7.4 -> 1.7.5
+
+## v1.2 [2017-01-00]
 
 ### Release Notes
 
@@ -54,7 +711,7 @@ plugins, not just statsd.
 - [#1775](https://github.com/influxdata/telegraf/issues/1775): Cache & expire metrics for delivery to prometheus.
 - [#2146](https://github.com/influxdata/telegraf/issues/2146): Fix potential panic in aggregator plugin metric maker.
 - [#1843](https://github.com/influxdata/telegraf/pull/1843) & [#1668](https://github.com/influxdata/telegraf/issues/1668): Add optional ability to define PID as a tag.
-- [#1730](https://github.com/influxdata/telegraf/issues/1730): Fix win_perf_counters not gathering non-English counters.
+- [#1730](https://github.com/influxdata/telegraf/issues/1730) & [#2261](https://github.com/influxdata/telegraf/pull/2261): Fix win_perf_counters not gathering non-English counters.
 - [#2061](https://github.com/influxdata/telegraf/issues/2061): Fix panic when file stat info cannot be collected due to permissions or other issue(s).
 - [#2045](https://github.com/influxdata/telegraf/issues/2045): Graylog output should set short_message field.
 - [#1904](https://github.com/influxdata/telegraf/issues/1904): Hddtemp always put the value in the field temperature.
@@ -67,7 +724,10 @@ plugins, not just statsd.
 - [#1973](https://github.com/influxdata/telegraf/issues/1973): Partial fix: logparser CLF pattern with IPv6 addresses.
 - [#1975](https://github.com/influxdata/telegraf/issues/1975) & [#2102](https://github.com/influxdata/telegraf/issues/2102): Fix thread-safety when using multiple instances of the statsd input plugin.
 - [#2027](https://github.com/influxdata/telegraf/issues/2027): docker input: interface conversion panic fix.
-- [#1814](https://github.com/influxdata/telegraf/issues/1814): snmp: ensure proper context is present on error messages
+- [#1814](https://github.com/influxdata/telegraf/issues/1814): snmp: ensure proper context is present on error messages.
+- [#2299](https://github.com/influxdata/telegraf/issues/2299): opentsdb: add tcp:// prefix if no scheme provided.
+- [#2297](https://github.com/influxdata/telegraf/issues/2297): influx parser: parse line-protocol without newlines.
+- [#2245](https://github.com/influxdata/telegraf/issues/2245): influxdb output: fix field type conflict blocking output buffer.
 
 ## v1.1.2 [2016-12-12]
 
@@ -218,8 +878,11 @@ which can be installed via
 evaluated at every flush interval, rather than once at startup. This makes it
 consistent with the behavior of `collection_jitter`.
 
+- postgresql plugins now handle oid and name typed columns seamlessly, previously they were ignored/skipped.
+
 ### Features
 
+- [#1617](https://github.com/influxdata/telegraf/pull/1617): postgresql_extensible now handles name and oid types correctly.
 - [#1413](https://github.com/influxdata/telegraf/issues/1413): Separate container_version from container_image tag.
 - [#1525](https://github.com/influxdata/telegraf/pull/1525): Support setting per-device and total metrics for Docker network and blockio.
 - [#1466](https://github.com/influxdata/telegraf/pull/1466): MongoDB input plugin: adding per DB stats from db.stats()

CONTRIBUTING.md

@@ -12,7 +12,7 @@ but any information you can provide on how the data will look is appreciated.
 See the [OpenTSDB output](https://github.com/influxdata/telegraf/tree/master/plugins/outputs/opentsdb)
 for a good example.
 1. **Optional:** Help users of your plugin by including example queries for populating dashboards. Include these sample queries in the `README.md` for the plugin.
-1. **Optional:** Write a [tickscript](https://docs.influxdata.com/kapacitor/v1.0/tick/syntax/) for your plugin and add it to [Kapacitor](https://github.com/influxdata/kapacitor/tree/master/examples/telegraf). Or mention @jackzampolin in a PR comment with some common queries that you would want to alert on and he will write one for you.
+1. **Optional:** Write a [tickscript](https://docs.influxdata.com/kapacitor/v1.0/tick/syntax/) for your plugin and add it to [Kapacitor](https://github.com/influxdata/kapacitor/tree/master/examples/telegraf).
 
 ## GoDoc
 
@@ -52,7 +52,7 @@ See below for a quick example.
 * Input Plugins must be added to the
 `github.com/influxdata/telegraf/plugins/inputs/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`.
+plugin can be configured. This is include in `telegraf config`.
 * The `Description` function should say in one line what this plugin does.
 
 Let's say you've written a plugin that emits metrics about processes on the
@@ -79,7 +79,10 @@ func (s *Simple) Description() string {
 }
 
 func (s *Simple) SampleConfig() string {
-    return "ok = true # indicate if everything is fine"
+    return `
+  ## Indicate if everything is fine
+  ok = true
+`
 }
 
 func (s *Simple) Gather(acc telegraf.Accumulator) error {
@@ -124,7 +127,7 @@ You should also add the following to your SampleConfig() return:
 
 ```toml
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "influx"
@@ -167,7 +170,7 @@ and `Stop()` methods.
 ### Service Plugin Guidelines
 
 * Same as the `Plugin` guidelines, except that they must conform to the
-`inputs.ServiceInput` interface.
+[`telegraf.ServiceInput`](https://godoc.org/github.com/influxdata/telegraf#ServiceInput) interface.
 
 ## Output Plugins
 
@@ -183,7 +186,7 @@ See below for a quick example.
 * To be available within Telegraf itself, plugins must add themselves to the
 `github.com/influxdata/telegraf/plugins/outputs/all/all.go` file.
 * The `SampleConfig` function should return valid toml that describes how the
-output can be configured. This is include in `telegraf -sample-config`.
+output can be configured. This is include in `telegraf config`.
 * The `Description` function should say in one line what this output does.
 
 ### Output Example
@@ -207,7 +210,9 @@ func (s *Simple) Description() string {
 }
 
 func (s *Simple) SampleConfig() string {
-    return "url = localhost"
+    return `
+  ok = true
+`
 }
 
 func (s *Simple) Connect() error {
@@ -254,7 +259,7 @@ You should also add the following to your SampleConfig() return:
 
 ```toml
   ## Data format to output.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
   data_format = "influx"
@@ -287,7 +292,7 @@ See below for a quick example.
 * To be available within Telegraf itself, plugins must add themselves to the
 `github.com/influxdata/telegraf/plugins/processors/all/all.go` file.
 * The `SampleConfig` function should return valid toml that describes how the
-processor can be configured. This is include in `telegraf -sample-config`.
+processor can be configured. This is include in the output of `telegraf config`.
 * The `Description` function should say in one line what this processor does.
 
 ### Processor Example
@@ -344,7 +349,7 @@ See below for a quick example.
 * To be available within Telegraf itself, plugins must add themselves to the
 `github.com/influxdata/telegraf/plugins/aggregators/all/all.go` file.
 * The `SampleConfig` function should return valid toml that describes how the
-aggregator can be configured. This is include in `telegraf -sample-config`.
+aggregator can be configured. This is include in `telegraf config`.
 * The `Description` function should say in one line what this aggregator does.
 * The Aggregator plugin will need to keep caches of metrics that have passed
 through it. This should be done using the builtin `HashID()` function of each
@@ -457,29 +462,28 @@ func init() {
 
 ## Unit Tests
 
+Before opening a pull request you should run the linter checks and
+the short tests.
+
+### Execute linter
+
+execute `make lint`
+
 ### Execute short tests
 
-execute `make test-short`
+execute `make test`
 
-### Execute long tests
+### Execute integration 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.
+Running the integration tests requires several docker containers to be
+running.  You can start the containers with:
+```
+make docker-run
+```
 
-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/influxdata/telegraf/blob/master/plugins/inputs/redis/redis_test.go)
-a simple mock will suffice.
+And run the full test suite with:
+```
+make test-all
+```
 
-To execute Telegraf tests follow these simple steps:
-
-- Install docker following [these](https://docs.docker.com/installation/)
-instructions
-- execute `make test`
-
-### Unit test troubleshooting
-
-Try cleaning up your test environment by executing `make docker-kill` and
-re-running
+Use `make docker-kill` to stop the containers.

Godeps

@@ -1,65 +1,100 @@
-github.com/Shopify/sarama 8aadb476e66ca998f2f6bb3c993e9a2daa3666b9
-github.com/Sirupsen/logrus 219c8cb75c258c552e999735be6df753ffc7afdc
-github.com/aerospike/aerospike-client-go 7f3a312c3b2a60ac083ec6da296091c52c795c63
-github.com/amir/raidman 53c1b967405155bfc8758557863bf2e14f814687
-github.com/aws/aws-sdk-go 13a12060f716145019378a10e2806c174356b857
-github.com/beorn7/perks 3ac7bf7a47d159a033b107610db8a1b6575507a4
-github.com/cenkalti/backoff 4dc77674aceaabba2c7e3da25d4c823edfb73f99
-github.com/couchbase/go-couchbase cb664315a324d87d19c879d9cc67fda6be8c2ac1
-github.com/couchbase/gomemcached a5ea6356f648fec6ab89add00edd09151455b4b2
+collectd.org 2ce144541b8903101fb8f1483cc0497a68798122
+github.com/Azure/go-autorest 9ad9326b278af8fa5cc67c30c0ce9a58cc0862b2
+github.com/Shopify/sarama 3b1b38866a79f06deddf0487d5c27ba0697ccd65
+github.com/Sirupsen/logrus 61e43dc76f7ee59a82bdf3d71033dc12bea4c77d
+github.com/aerospike/aerospike-client-go 95e1ad7791bdbca44707fedbb29be42024900d9c
+github.com/amir/raidman c74861fe6a7bb8ede0a010ce4485bdbb4fc4c985
+github.com/apache/thrift 4aaa92ece8503a6da9bc6701604f69acf2b99d07
+github.com/aws/aws-sdk-go c861d27d0304a79f727e9a8a4e2ac1e74602fdc0
+github.com/beorn7/perks 4c0e84591b9aa9e6dcfdf3e020114cd81f89d5f9
+github.com/bsm/sarama-cluster abf039439f66c1ce78017f560b490612552f6472
+github.com/cenkalti/backoff b02f2bbce11d7ea6b97f282ef1771b0fe2f65ef3
+github.com/couchbase/go-couchbase bfe555a140d53dc1adf390f1a1d4b0fd4ceadb28
+github.com/couchbase/gomemcached 4a25d2f4e1dea9ea7dd76dfd943407abf9b07d29
 github.com/couchbase/goutils 5823a0cbaaa9008406021dc5daf80125ea30bba6
-github.com/dancannon/gorethink e7cac92ea2bc52638791a021f212145acfedb1fc
-github.com/davecgh/go-spew 5215b55f46b2b919f50a1df0eaa5886afe4e3b3d
-github.com/docker/engine-api 8924d6900370b4c7e7984be5adc61f50a80d7537
-github.com/docker/go-connections f549a9393d05688dff0992ef3efd8bbe6c628aeb
-github.com/docker/go-units 5d2041e26a699eaca682e2ea41c8f891e1060444
+github.com/davecgh/go-spew 346938d642f2ec3594ed81d874461961cd0faa76
+github.com/dgrijalva/jwt-go dbeaa9332f19a944acb5736b4456cfcc02140e29
+github.com/docker/docker f5ec1e2936dcbe7b5001c2b817188b095c700c27
+github.com/docker/go-connections 990a1a1a70b0da4c4cb70e117971a4f0babfbf1a
 github.com/eapache/go-resiliency b86b1ec0dd4209a588dc1285cdd471e73525c0b3
-github.com/eapache/queue ded5959c0d4e360646dc9e9908cff48666781367
-github.com/eclipse/paho.mqtt.golang 0f7a459f04f13a41b7ed752d47944528d4bf9a86
-github.com/go-sql-driver/mysql 1fca743146605a172a266e1654e01e5cd5669bee
-github.com/gobwas/glob 49571a1557cd20e6a2410adc6421f85b66c730b5
-github.com/golang/protobuf 552c7b9542c194800fd493123b3798ef0a832032
-github.com/golang/snappy d9eb7a3d35ec988b8585d4a0068e462c27d28380
-github.com/gorilla/context 1ea25387ff6f684839d82767c1733ff4d4d15d0a
-github.com/gorilla/mux c9e326e2bdec29039a3761c07bece13133863e1e
+github.com/eapache/go-xerial-snappy bb955e01b9346ac19dc29eb16586c90ded99a98c
+github.com/eapache/queue 44cc805cf13205b55f69e14bcb69867d1ae92f98
+github.com/eclipse/paho.mqtt.golang aff15770515e3c57fc6109da73d42b0d46f7f483
+github.com/go-logfmt/logfmt 390ab7935ee28ec6b286364bba9b4dd6410cb3d5
+github.com/go-redis/redis 73b70592cdaa9e6abdfcfbf97b4a90d80728c836
+github.com/go-sql-driver/mysql 2e00b5cd70399450106cec6431c2e2ce3cae5034
+github.com/gobwas/glob bea32b9cd2d6f55753d94a28e959b13f0244797a
+github.com/go-ini/ini 9144852efba7c4daf409943ee90767da62d55438
+github.com/gogo/protobuf 7b6c6391c4ff245962047fc1e2c6e08b1cdfa0e8
+github.com/golang/protobuf 8ee79997227bf9b34611aee7946ae64735e6fd93
+github.com/golang/snappy 7db9049039a047d955fe8c19b83c8ff5abd765c7
+github.com/go-ole/go-ole be49f7c07711fcb603cff39e1de7c67926dc0ba7
+github.com/google/go-cmp f94e52cad91c65a63acc1e75d4be223ea22e99bc
+github.com/gorilla/mux 53c1911da2b537f792e7cafcb446b05ffe33b996
+github.com/go-redis/redis 73b70592cdaa9e6abdfcfbf97b4a90d80728c836
+github.com/go-sql-driver/mysql 2e00b5cd70399450106cec6431c2e2ce3cae5034
 github.com/hailocab/go-hostpool e80d13ce29ede4452c43dea11e79b9bc8a15b478
-github.com/hashicorp/consul 5aa90455ce78d4d41578bafc86305e6e6b28d7d2
-github.com/hpcloud/tail b2940955ab8b26e19d43a43c4da0475dd81bdb56
-github.com/influxdata/config b79f6829346b8d6e78ba73544b1e1038f1f1c9da
-github.com/influxdata/influxdb fc57c0f7c635df3873f3d64f0ed2100ddc94d5ae
-github.com/influxdata/toml af4df43894b16e3fd2b788d01bd27ad0776ef2d0
+github.com/hashicorp/consul 63d2fc68239b996096a1c55a0d4b400ea4c2583f
+github.com/influxdata/tail c43482518d410361b6c383d7aebce33d0471d7bc
+github.com/influxdata/toml 5d1d907f22ead1cd47adde17ceec5bda9cacaf8f
 github.com/influxdata/wlog 7c63b0a71ef8300adc255344d275e10e5c3a71ec
-github.com/kardianos/osext 29ae4ffbc9a6fe9fb2bc5029050ce6996ea1d3bc
-github.com/kardianos/service 5e335590050d6d00f3aa270217d288dda1c94d0a
+github.com/fsnotify/fsnotify c2828203cd70a50dcccfb2761f8b1f8ceef9a8e9
+github.com/jackc/pgx 63f58fd32edb5684b9e9f4cfaac847c6b42b3917
+github.com/jmespath/go-jmespath bd40a432e4c76585ef6b72d3fd96fb9b6dc7b68d
+github.com/kardianos/osext c2c54e542fb797ad986b31721e1baedf214ca413
+github.com/kardianos/service 6d3a0ee7d3425d9d835debc51a0ca1ffa28f4893
 github.com/kballard/go-shellquote d8ec1a69a250a17bb0e419c386eac1f3711dc142
-github.com/klauspost/crc32 19b0b332c9e4516a6370a0456e6182c3b5036720
-github.com/lib/pq e182dc4027e2ded4b19396d638610f2653295f36
-github.com/matttproud/golang_protobuf_extensions d0c3fe89de86839aecf2e0579c40ba3bb336a453
-github.com/miekg/dns cce6c130cdb92c752850880fd285bea1d64439dd
-github.com/mreiferson/go-snappystream 028eae7ab5c4c9e2d1cb4c4ca1e53259bbe7e504
+github.com/matttproud/golang_protobuf_extensions c12348ce28de40eed0136aa2b644d0ee0650e56c
+github.com/Microsoft/go-winio ce2922f643c8fd76b46cadc7f404a06282678b34
+github.com/miekg/dns 99f84ae56e75126dd77e5de4fae2ea034a468ca1
+github.com/mitchellh/mapstructure d0303fe809921458f417bcf828397a65db30a7e4
+github.com/multiplay/go-ts3 07477f49b8dfa3ada231afc7b7b17617d42afe8e
 github.com/naoina/go-stringutil 6b638e95a32d0c1131db0e7fe83775cbea4a0d0b
-github.com/nats-io/nats ea8b4fd12ebb823073c0004b9f09ac8748f4f165
-github.com/nats-io/nuid a5152d67cf63cbfb5d992a395458722a45194715
-github.com/nsqio/go-nsq 0b80d6f05e15ca1930e0c5e1d540ed627e299980
+github.com/nats-io/gnatsd 393bbb7c031433e68707c8810fda0bfcfbe6ab9b
+github.com/nats-io/go-nats ea9585611a4ab58a205b9b125ebd74c389a6b898
+github.com/nats-io/nats ea9585611a4ab58a205b9b125ebd74c389a6b898
+github.com/nats-io/nuid 289cccf02c178dc782430d534e3c1f5b72af807f
+github.com/nsqio/go-nsq eee57a3ac4174c55924125bb15eeeda8cffb6e6f
 github.com/opencontainers/runc 89ab7f2ccc1e45ddf6485eaa802c35dcf321dfc8
-github.com/prometheus/client_golang 18acf9993a863f4c4b40612e19cdd243e7c86831
+github.com/opentracing-contrib/go-observer a52f2342449246d5bcc273e65cbdcfa5f7d6c63c
+github.com/opentracing/opentracing-go 06f47b42c792fef2796e9681353e1d908c417827
+github.com/openzipkin/zipkin-go-opentracing 1cafbdfde94fbf2b373534764e0863aa3bd0bf7b
+github.com/pierrec/lz4 5c9560bfa9ace2bf86080bf40d46b34ae44604df
+github.com/pierrec/xxHash 5a004441f897722c627870a981d02b29924215fa
+github.com/pkg/errors 645ef00459ed84a119197bfb8d8205042c6df63d
+github.com/pmezard/go-difflib/difflib 792786c7400a136282c1664665ae0a8db921c6c2
+github.com/prometheus/client_golang c317fb74746eac4fc65fe3909195f4cf67c5562a
 github.com/prometheus/client_model fa8ad6fec33561be4280a8f0514318c79d7f6cb6
-github.com/prometheus/common e8eabff8812b05acf522b45fdcd725a785188e37
-github.com/prometheus/procfs 406e5b7bfd8201a36e2bb5f7bdae0b03380c2ce8
-github.com/samuel/go-zookeeper 218e9c81c0dd8b3b18172b2bbfad92cc7d6db55f
-github.com/shirou/gopsutil 1516eb9ddc5e61ba58874047a98f8b44b5e585e8
-github.com/soniah/gosnmp 3fe3beb30fa9700988893c56a63b1df8e1b68c26
-github.com/streadway/amqp b4f3ceab0337f013208d31348b578d83c0064744
-github.com/stretchr/testify 1f4a1643a57e798696635ea4c126e9127adb7d3c
-github.com/vjeantet/grok 83bfdfdfd1a8146795b28e547a8e3c8b28a466c2
+github.com/prometheus/common dd2f054febf4a6c00f2343686efb775948a8bff4
+github.com/prometheus/procfs 1878d9fbb537119d24b21ca07effd591627cd160
+github.com/rcrowley/go-metrics 1f30fe9094a513ce4c700b9a54458bbb0c96996c
+github.com/samuel/go-zookeeper 1d7be4effb13d2d908342d349d71a284a7542693
+github.com/satori/go.uuid 5bf94b69c6b68ee1b541973bb8e1144db23a194b
+github.com/shirou/gopsutil a5c2888e464b14fa882c2a059e0f95716bd45cf1
+github.com/shirou/w32 3c9377fc6748f222729a8270fe2775d149a249ad
+github.com/Shopify/sarama 3b1b38866a79f06deddf0487d5c27ba0697ccd65
+github.com/Sirupsen/logrus 61e43dc76f7ee59a82bdf3d71033dc12bea4c77d
+github.com/soniah/gosnmp f15472a4cd6f6ea7929e4c7d9f163c49f059924f
+github.com/StackExchange/wmi f3e2bae1e0cb5aef83e319133eabfee30013a4a5
+github.com/streadway/amqp 63795daa9a446c920826655f26ba31c81c860fd6
+github.com/stretchr/objx facf9a85c22f48d2f52f2380e4efce1768749a89
+github.com/stretchr/testify 12b6f73e6084dad08a7c6e575284b177ecafbc71
+github.com/tidwall/gjson 0623bd8fbdbf97cc62b98d15108832851a658e59
+github.com/tidwall/match 173748da739a410c5b0b813b956f89ff94730b4c
+github.com/vjeantet/grok d73e972b60935c7fec0b4ffbc904ed39ecaf7efe
 github.com/wvanbergen/kafka bc265fedb9ff5b5c5d3c0fdcef4a819b3523d3ee
-github.com/wvanbergen/kazoo-go 0f768712ae6f76454f987c3356177e138df258f8
-github.com/yuin/gopher-lua bf3808abd44b1e55143a2d7f08571aaa80db1808
+github.com/wvanbergen/kazoo-go 968957352185472eacb69215fa3dbfcfdbac1096
+github.com/yuin/gopher-lua 66c871e454fcf10251c61bf8eff02d0978cae75a
 github.com/zensqlmonitor/go-mssqldb ffe5510c6fa5e15e6d983210ab501c815b56b363
-golang.org/x/crypto c197bcf24cde29d3f73c7b4ac6fd41f4384e8af6
-golang.org/x/net 6acef71eb69611914f7a30939ea9f6e194c78172
-golang.org/x/text a71fd10341b064c10f4a81ceac72bcf70f26ea34
-gopkg.in/dancannon/gorethink.v1 7d1af5be49cb5ecc7b177bf387d232050299d6ef
-gopkg.in/fatih/pool.v2 cba550ebf9bce999a02e963296d4bc7a486cb715
-gopkg.in/mgo.v2 d90005c5262a3463800497ea5a89aed5fe22c886
-gopkg.in/yaml.v2 a83829b6f1293c91addabc89d0571c246397bbf4
+golang.org/x/crypto dc137beb6cce2043eb6b5f223ab8bf51c32459f4
+golang.org/x/net f2499483f923065a842d38eb4c7f1927e6fc6e6d
+golang.org/x/sys 739734461d1c916b6c72a63d7efda2b27edb369f
+golang.org/x/text 506f9d5c962f284575e88337e7d9296d27e729d3
+gopkg.in/asn1-ber.v1 4e86f4367175e39f69d9358a5f17b4dda270378d
+gopkg.in/fatih/pool.v2 6e328e67893eb46323ad06f0e92cb9536babbabc
+gopkg.in/gorethink/gorethink.v3 7ab832f7b65573104a555d84a27992ae9ea1f659
+gopkg.in/ldap.v2 8168ee085ee43257585e50c6441aadf54ecb2c9f
+gopkg.in/mgo.v2 3f83fa5005286a7fe593b055f0d7771a7dce4655
+gopkg.in/olivere/elastic.v5 3113f9b9ad37509fe5f8a0e5e91c96fdc4435e26
+gopkg.in/tomb.v1 dd632973f1e7218eb1089048e0798ec9ae7dceb8
+gopkg.in/yaml.v2 4c78c975fe7c825c6d1466c42be594d1d6f3aba6

Godeps_windows

@@ -1,11 +0,0 @@
-github.com/Microsoft/go-winio ce2922f643c8fd76b46cadc7f404a06282678b34
-github.com/StackExchange/wmi f3e2bae1e0cb5aef83e319133eabfee30013a4a5
-github.com/go-ole/go-ole be49f7c07711fcb603cff39e1de7c67926dc0ba7
-github.com/shirou/w32 3c9377fc6748f222729a8270fe2775d149a249ad
-golang.org/x/sys a646d33e2ee3172a661fc09bca23bb4889a41bc8
-github.com/go-ini/ini 9144852efba7c4daf409943ee90767da62d55438
-github.com/jmespath/go-jmespath bd40a432e4c76585ef6b72d3fd96fb9b6dc7b68d
-github.com/pmezard/go-difflib/difflib 792786c7400a136282c1664665ae0a8db921c6c2
-github.com/stretchr/objx 1a9d0bb9f541897e62256577b352fdbc1fb4fd94
-gopkg.in/fsnotify.v1 a8a77c9133d2d6fd8334f3260d06f60e8d80a5fb
-gopkg.in/tomb.v1 dd632973f1e7218eb1089048e0798ec9ae7dceb8

Makefile

@@ -1,96 +1,95 @@
-VERSION := $(shell sh -c 'git describe --always --tags')
-BRANCH := $(shell sh -c 'git rev-parse --abbrev-ref HEAD')
-COMMIT := $(shell sh -c 'git rev-parse --short HEAD')
+PREFIX := /usr/local
+VERSION := $(shell git describe --exact-match --tags 2>/dev/null)
+BRANCH := $(shell git rev-parse --abbrev-ref HEAD)
+COMMIT := $(shell git rev-parse --short HEAD)
+GOFILES ?= $(shell git ls-files '*.go')
+GOFMT ?= $(shell gofmt -l $(filter-out plugins/parsers/influx/machine.go, $(GOFILES)))
+BUILDFLAGS ?=
+
 ifdef GOBIN
 PATH := $(GOBIN):$(PATH)
 else
 PATH := $(subst :,/bin:,$(GOPATH))/bin:$(PATH)
 endif
 
-# Standard Telegraf build
-default: prepare build
+LDFLAGS := $(LDFLAGS) -X main.commit=$(COMMIT) -X main.branch=$(BRANCH)
+ifdef VERSION
+	LDFLAGS += -X main.version=$(VERSION)
+endif
 
-# Windows build
-windows: prepare-windows build-windows
+all:
+	$(MAKE) deps
+	$(MAKE) telegraf
 
-# Only run the build (no dependency grabbing)
-build:
-	go install -ldflags \
-		"-X main.version=$(VERSION) -X main.commit=$(COMMIT) -X main.branch=$(BRANCH)" ./...
-
-build-windows:
-	GOOS=windows GOARCH=amd64 go build -o telegraf.exe -ldflags \
-		"-X main.version=$(VERSION) -X main.commit=$(COMMIT) -X main.branch=$(BRANCH)" \
-		./cmd/telegraf/telegraf.go
-
-build-for-docker:
-	CGO_ENABLED=0 GOOS=linux go build -installsuffix cgo -o telegraf -ldflags \
-		"-s -X main.version=$(VERSION) -X main.commit=$(COMMIT) -X main.branch=$(BRANCH)" \
-		./cmd/telegraf/telegraf.go
-
-# run package script
-package:
-	./scripts/build.py --package --version="$(VERSION)" --platform=linux --arch=all --upload
-
-# Get dependencies and use gdm to checkout changesets
-prepare:
+deps:
+	go get -u github.com/golang/lint/golint
 	go get github.com/sparrc/gdm
 	gdm restore
 
-# Use the windows godeps file to prepare dependencies
-prepare-windows:
-	go get github.com/sparrc/gdm
-	gdm restore
-	gdm restore -f Godeps_windows
+telegraf:
+	go build -i -ldflags "$(LDFLAGS)" ./cmd/telegraf
 
-# Run all docker containers necessary for unit tests
-docker-run:
-	docker run --name aerospike -p "3000:3000" -d aerospike/aerospike-server:3.9.0
-	docker run --name kafka \
-		-e ADVERTISED_HOST=localhost \
-		-e ADVERTISED_PORT=9092 \
-		-p "2181:2181" -p "9092:9092" \
-		-d spotify/kafka
-	docker run --name mysql -p "3306:3306" -e MYSQL_ALLOW_EMPTY_PASSWORD=yes -d mysql
-	docker run --name memcached -p "11211:11211" -d memcached
-	docker run --name postgres -p "5432:5432" -d postgres
-	docker run --name rabbitmq -p "15672:15672" -p "5672:5672" -d rabbitmq:3-management
-	docker run --name redis -p "6379:6379" -d redis
-	docker run --name nsq -p "4150:4150" -d nsqio/nsq /nsqd
-	docker run --name mqtt -p "1883:1883" -d ncarlier/mqtt
-	docker run --name riemann -p "5555:5555" -d blalor/riemann
-	docker run --name nats -p "4222:4222" -d nats
+go-install:
+	go install -ldflags "-w -s $(LDFLAGS)" ./cmd/telegraf
 
-# Run docker containers necessary for CircleCI unit tests
-docker-run-circle:
-	docker run --name aerospike -p "3000:3000" -d aerospike/aerospike-server:3.9.0
-	docker run --name kafka \
-		-e ADVERTISED_HOST=localhost \
-		-e ADVERTISED_PORT=9092 \
-		-p "2181:2181" -p "9092:9092" \
-		-d spotify/kafka
-	docker run --name nsq -p "4150:4150" -d nsqio/nsq /nsqd
-	docker run --name mqtt -p "1883:1883" -d ncarlier/mqtt
-	docker run --name riemann -p "5555:5555" -d blalor/riemann
-	docker run --name nats -p "4222:4222" -d nats
+install: telegraf
+	mkdir -p $(DESTDIR)$(PREFIX)/bin/
+	cp $(TELEGRAF) $(DESTDIR)$(PREFIX)/bin/
 
-# Kill all docker containers, ignore errors
-docker-kill:
-	-docker kill nsq aerospike redis rabbitmq postgres memcached mysql kafka mqtt riemann nats
-	-docker rm nsq aerospike redis rabbitmq postgres memcached mysql kafka mqtt riemann nats
-
-# Run full unit tests using docker containers (includes setup and teardown)
-test: vet docker-kill docker-run
-	# Sleeping for kafka leadership election, TSDB setup, etc.
-	sleep 60
-	# SUCCESS, running tests
-	go test -race ./...
-
-# Run "short" unit tests
-test-short: vet
+test:
 	go test -short ./...
 
-vet:
-	go vet ./...
+fmt:
+	@gofmt -w $(filter-out plugins/parsers/influx/machine.go, $(GOFILES))
 
-.PHONY: test test-short vet build default
+fmtcheck:
+	@echo '[INFO] running gofmt to identify incorrectly formatted code...'
+	@if [ ! -z "$(GOFMT)" ]; then \
+		echo "[ERROR] gofmt has found errors in the following files:"  ; \
+		echo "$(GOFMT)" ; \
+		echo "" ;\
+		echo "Run make fmt to fix them." ; \
+		exit 1 ;\
+	fi
+	@echo '[INFO] done.'
+
+test-windows:
+	go test ./plugins/inputs/ping/...
+	go test ./plugins/inputs/win_perf_counters/...
+	go test ./plugins/inputs/win_services/...
+	go test ./plugins/inputs/procstat/...
+	go test ./plugins/inputs/ntpq/...
+
+# vet runs the Go source code static analysis tool `vet` to find
+# any common errors.
+vet:
+	@echo 'go vet $$(go list ./... | grep -v ./plugins/parsers/influx)'
+	@go vet $$(go list ./... | grep -v ./plugins/parsers/influx) ; if [ $$? -ne 0 ]; then \
+		echo ""; \
+		echo "go vet has found suspicious constructs. Please remediate any reported errors"; \
+		echo "to fix them before submitting code for review."; \
+		exit 1; \
+	fi
+
+test-ci: fmtcheck vet
+	go test -short ./...
+
+test-all: fmtcheck vet
+	go test ./...
+
+package:
+	./scripts/build.py --package --platform=all --arch=all
+
+clean:
+	rm -f telegraf
+	rm -f telegraf.exe
+
+docker-image:
+	./scripts/build.py --package --platform=linux --arch=amd64
+	cp build/telegraf*$(COMMIT)*.deb .
+	docker build -f scripts/dev.docker --build-arg "package=telegraf*$(COMMIT)*.deb" -t "telegraf-dev:$(COMMIT)" .
+
+plugins/parsers/influx/machine.go: plugins/parsers/influx/machine.go.rl
+	ragel -Z -G2 $^ -o $@
+
+.PHONY: deps telegraf install test test-windows lint vet test-all package clean docker-image fmtcheck uint64

README.md

@@ -5,8 +5,7 @@ and writing metrics.
 
 Design goals are to have a minimal memory footprint with a plugin system so
 that developers in the community can easily add support for collecting metrics
-from well known services (like Hadoop, Postgres, or Redis) and third party
-APIs (like Mailchimp, AWS CloudWatch, or Google Analytics).
+.  For an example configuration referencet from local or remote services.
 
 Telegraf is plugin-driven and has the concept of 4 distinct plugins:
 
@@ -20,113 +19,101 @@ For more information on Processor and Aggregator plugins please [read this](./do
 New plugins are designed to be easy to contribute,
 we'll eagerly accept pull
 requests and will manage the set of plugins that Telegraf supports.
-See the [contributing guide](CONTRIBUTING.md) for instructions on writing
-new plugins.
+
+## Contributing
+
+There are many ways to contribute:
+- Fix and [report bugs](https://github.com/influxdata/telegraf/issues/new)
+- [Improve documentation](https://github.com/influxdata/telegraf/issues?q=is%3Aopen+label%3Adocumentation)
+- [Review code and feature proposals](https://github.com/influxdata/telegraf/pulls)
+- Answer questions on github and on the [Community Site](https://community.influxdata.com/)
+- [Contribute plugins](CONTRIBUTING.md)
 
 ## Installation:
 
-### Linux deb and rpm Packages:
-
-Latest:
-* https://dl.influxdata.com/telegraf/releases/telegraf_1.1.1_amd64.deb
-* https://dl.influxdata.com/telegraf/releases/telegraf-1.1.1.x86_64.rpm
-
-Latest (arm):
-* https://dl.influxdata.com/telegraf/releases/telegraf_1.1.1_armhf.deb
-* https://dl.influxdata.com/telegraf/releases/telegraf-1.1.1.armhf.rpm
-
-##### Package Instructions:
-
-* Telegraf binary is installed in `/usr/bin/telegraf`
-* Telegraf daemon configuration file is in `/etc/telegraf/telegraf.conf`
-* On sysv systems, the telegraf daemon can be controlled via
-`service telegraf [action]`
-* On systemd systems (such as Ubuntu 15+), the telegraf daemon can be
-controlled via `systemctl [action] telegraf`
-
-### yum/apt Repositories:
-
-There is a yum/apt repo available for the whole InfluxData stack, see
-[here](https://docs.influxdata.com/influxdb/latest/introduction/installation/#installation)
-for instructions on setting up the repo. Once it is configured, you will be able
-to use this repo to install & update telegraf.
-
-### Linux tarballs:
-
-Latest:
-* https://dl.influxdata.com/telegraf/releases/telegraf-1.1.1_linux_amd64.tar.gz
-* https://dl.influxdata.com/telegraf/releases/telegraf-1.1.1_linux_i386.tar.gz
-* https://dl.influxdata.com/telegraf/releases/telegraf-1.1.1_linux_armhf.tar.gz
-
-### FreeBSD tarball:
-
-Latest:
-* https://dl.influxdata.com/telegraf/releases/telegraf-1.1.1_freebsd_amd64.tar.gz
+You can download the binaries directly from the [downloads](https://www.influxdata.com/downloads) page
+or from the [releases](https://github.com/influxdata/telegraf/releases) section.
 
 ### Ansible Role:
 
 Ansible role: https://github.com/rossmcdonald/telegraf
 
-### OSX via Homebrew:
-
-```
-brew update
-brew install telegraf
-```
-
-### Windows Binaries (EXPERIMENTAL)
-
-Latest:
-* https://dl.influxdata.com/telegraf/releases/telegraf-1.1.1_windows_amd64.zip
-
 ### From Source:
 
-Telegraf manages dependencies via [gdm](https://github.com/sparrc/gdm),
-which gets installed via the Makefile
-if you don't have it already. You also must build with golang version 1.5+.
+Telegraf requires golang version 1.8+, the Makefile requires GNU make.
+
+Dependencies are managed with [gdm](https://github.com/sparrc/gdm),
+which is installed by the Makefile if you don't have it already.
 
 1. [Install Go](https://golang.org/doc/install)
 2. [Setup your GOPATH](https://golang.org/doc/code.html#GOPATH)
-3. Run `go get github.com/influxdata/telegraf`
+3. Run `go get -d github.com/influxdata/telegraf`
 4. Run `cd $GOPATH/src/github.com/influxdata/telegraf`
 5. Run `make`
 
+### Nightly Builds
+
+These builds are generated from the master branch:
+- [telegraf_nightly_amd64.deb](https://dl.influxdata.com/telegraf/nightlies/telegraf_nightly_amd64.deb)
+- [telegraf_nightly_arm64.deb](https://dl.influxdata.com/telegraf/nightlies/telegraf_nightly_arm64.deb)
+- [telegraf-nightly.arm64.rpm](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly.arm64.rpm)
+- [telegraf_nightly_armel.deb](https://dl.influxdata.com/telegraf/nightlies/telegraf_nightly_armel.deb)
+- [telegraf-nightly.armel.rpm](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly.armel.rpm)
+- [telegraf_nightly_armhf.deb](https://dl.influxdata.com/telegraf/nightlies/telegraf_nightly_armhf.deb)
+- [telegraf-nightly.armv6hl.rpm](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly.armv6hl.rpm)
+- [telegraf-nightly_freebsd_amd64.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_freebsd_amd64.tar.gz)
+- [telegraf-nightly_freebsd_i386.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_freebsd_i386.tar.gz)
+- [telegraf_nightly_i386.deb](https://dl.influxdata.com/telegraf/nightlies/telegraf_nightly_i386.deb)
+- [telegraf-nightly.i386.rpm](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly.i386.rpm)
+- [telegraf-nightly_linux_amd64.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_linux_amd64.tar.gz)
+- [telegraf-nightly_linux_arm64.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_linux_arm64.tar.gz)
+- [telegraf-nightly_linux_armel.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_linux_armel.tar.gz)
+- [telegraf-nightly_linux_armhf.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_linux_armhf.tar.gz)
+- [telegraf-nightly_linux_i386.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_linux_i386.tar.gz)
+- [telegraf-nightly_linux_s390x.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_linux_s390x.tar.gz)
+- [telegraf_nightly_s390x.deb](https://dl.influxdata.com/telegraf/nightlies/telegraf_nightly_s390x.deb)
+- [telegraf-nightly.s390x.rpm](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly.s390x.rpm)
+- [telegraf-nightly_windows_amd64.zip](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_windows_amd64.zip)
+- [telegraf-nightly_windows_i386.zip](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly_windows_i386.zip)
+- [telegraf-nightly.x86_64.rpm](https://dl.influxdata.com/telegraf/nightlies/telegraf-nightly.x86_64.rpm)
+- [telegraf-static-nightly_linux_amd64.tar.gz](https://dl.influxdata.com/telegraf/nightlies/telegraf-static-nightly_linux_amd64.tar.gz)
+
 ## How to use it:
 
 See usage with:
 
 ```
-telegraf --help
+./telegraf --help
 ```
 
-### Generate a telegraf config file:
+#### Generate a telegraf config file:
 
 ```
-telegraf config > telegraf.conf
+./telegraf config > telegraf.conf
 ```
 
-### Generate config with only cpu input & influxdb output plugins defined
+#### Generate config with only cpu input & influxdb output plugins defined:
 
 ```
-telegraf --input-filter cpu --output-filter influxdb config
+./telegraf --input-filter cpu --output-filter influxdb config
 ```
 
-### Run a single telegraf collection, outputing metrics to stdout
+#### Run a single telegraf collection, outputing metrics to stdout:
 
 ```
-telegraf --config telegraf.conf -test
+./telegraf --config telegraf.conf --test
 ```
 
-### Run telegraf with all plugins defined in config file
+#### Run telegraf with all plugins defined in config file:
 
 ```
-telegraf --config telegraf.conf
+./telegraf --config telegraf.conf
 ```
 
-### Run telegraf, enabling the cpu & memory input, and influxdb output plugins
+#### Run telegraf, enabling the cpu & memory input, and influxdb output plugins:
 
 ```
-telegraf --config telegraf.conf -input-filter cpu:mem -output-filter influxdb
+./telegraf --config telegraf.conf --input-filter cpu:mem --output-filter influxdb
 ```
 
 
@@ -137,68 +124,99 @@ configuration options.
 
 ## Input Plugins
 
-* [aws cloudwatch](./plugins/inputs/cloudwatch)
 * [aerospike](./plugins/inputs/aerospike)
+* [amqp_consumer](./plugins/inputs/amqp_consumer) (rabbitmq)
 * [apache](./plugins/inputs/apache)
+* [aws cloudwatch](./plugins/inputs/cloudwatch)
 * [bcache](./plugins/inputs/bcache)
-* [cassandra](./plugins/inputs/cassandra)
+* [bond](./plugins/inputs/bond)
+* [cassandra](./plugins/inputs/cassandra) (deprecated, use [jolokia2](./plugins/inputs/jolokia2))
 * [ceph](./plugins/inputs/ceph)
+* [cgroup](./plugins/inputs/cgroup)
 * [chrony](./plugins/inputs/chrony)
 * [consul](./plugins/inputs/consul)
 * [conntrack](./plugins/inputs/conntrack)
 * [couchbase](./plugins/inputs/couchbase)
 * [couchdb](./plugins/inputs/couchdb)
+* [DC/OS](./plugins/inputs/dcos)
 * [disque](./plugins/inputs/disque)
+* [dmcache](./plugins/inputs/dmcache)
 * [dns query time](./plugins/inputs/dns_query)
 * [docker](./plugins/inputs/docker)
 * [dovecot](./plugins/inputs/dovecot)
 * [elasticsearch](./plugins/inputs/elasticsearch)
 * [exec](./plugins/inputs/exec) (generic executable plugin, support JSON, influx, graphite and nagios)
+* [fail2ban](./plugins/inputs/fail2ban)
+* [fibaro](./plugins/inputs/fibaro)
 * [filestat](./plugins/inputs/filestat)
+* [fluentd](./plugins/inputs/fluentd)
+* [graylog](./plugins/inputs/graylog)
 * [haproxy](./plugins/inputs/haproxy)
 * [hddtemp](./plugins/inputs/hddtemp)
+* [http](./plugins/inputs/http) (generic HTTP plugin, supports using input data formats)
 * [http_response](./plugins/inputs/http_response)
 * [httpjson](./plugins/inputs/httpjson) (generic JSON-emitting http service plugin)
 * [internal](./plugins/inputs/internal)
 * [influxdb](./plugins/inputs/influxdb)
+* [interrupts](./plugins/inputs/interrupts)
 * [ipmi_sensor](./plugins/inputs/ipmi_sensor)
 * [iptables](./plugins/inputs/iptables)
-* [jolokia](./plugins/inputs/jolokia)
+* [ipset](./plugins/inputs/ipset)
+* [jolokia](./plugins/inputs/jolokia) (deprecated, use [jolokia2](./plugins/inputs/jolokia2))
+* [jolokia2](./plugins/inputs/jolokia2) (java, cassandra, kafka)
+* [kapacitor](./plugins/inputs/kapacitor)
+* [kubernetes](./plugins/inputs/kubernetes)
 * [leofs](./plugins/inputs/leofs)
 * [lustre2](./plugins/inputs/lustre2)
 * [mailchimp](./plugins/inputs/mailchimp)
+* [mcrouter](./plugins/inputs/mcrouter)
 * [memcached](./plugins/inputs/memcached)
 * [mesos](./plugins/inputs/mesos)
+* [minecraft](./plugins/inputs/minecraft)
 * [mongodb](./plugins/inputs/mongodb)
 * [mysql](./plugins/inputs/mysql)
+* [nats](./plugins/inputs/nats)
 * [net_response](./plugins/inputs/net_response)
 * [nginx](./plugins/inputs/nginx)
+* [nginx_plus](./plugins/inputs/nginx_plus)
 * [nsq](./plugins/inputs/nsq)
 * [nstat](./plugins/inputs/nstat)
 * [ntpq](./plugins/inputs/ntpq)
+* [nvidia_smi](./plugins/inputs/nvidia_smi)
+* [openldap](./plugins/inputs/openldap)
+* [opensmtpd](./plugins/inputs/opensmtpd)
+* [pf](./plugins/inputs/pf)
 * [phpfpm](./plugins/inputs/phpfpm)
 * [phusion passenger](./plugins/inputs/passenger)
 * [ping](./plugins/inputs/ping)
-* [postgresql](./plugins/inputs/postgresql)
+* [postfix](./plugins/inputs/postfix)
 * [postgresql_extensible](./plugins/inputs/postgresql_extensible)
+* [postgresql](./plugins/inputs/postgresql)
 * [powerdns](./plugins/inputs/powerdns)
 * [procstat](./plugins/inputs/procstat)
-* [prometheus](./plugins/inputs/prometheus)
+* [prometheus](./plugins/inputs/prometheus) (can be used for [Caddy server](./plugins/inputs/prometheus/README.md#usage-for-caddy-http-server))
 * [puppetagent](./plugins/inputs/puppetagent)
 * [rabbitmq](./plugins/inputs/rabbitmq)
 * [raindrops](./plugins/inputs/raindrops)
 * [redis](./plugins/inputs/redis)
 * [rethinkdb](./plugins/inputs/rethinkdb)
 * [riak](./plugins/inputs/riak)
+* [salesforce](./plugins/inputs/salesforce)
 * [sensors](./plugins/inputs/sensors)
+* [smart](./plugins/inputs/smart)
 * [snmp](./plugins/inputs/snmp)
 * [snmp_legacy](./plugins/inputs/snmp_legacy)
+* [solr](./plugins/inputs/solr)
 * [sql server](./plugins/inputs/sqlserver) (microsoft)
+* [teamspeak](./plugins/inputs/teamspeak)
+* [tomcat](./plugins/inputs/tomcat)
 * [twemproxy](./plugins/inputs/twemproxy)
+* [unbound](./plugins/inputs/unbound)
 * [varnish](./plugins/inputs/varnish)
 * [zfs](./plugins/inputs/zfs)
 * [zookeeper](./plugins/inputs/zookeeper)
-* [win_perf_counters ](./plugins/inputs/win_perf_counters) (windows performance counters)
+* [win_perf_counters](./plugins/inputs/win_perf_counters) (windows performance counters)
+* [win_services](./plugins/inputs/win_services)
 * [sysstat](./plugins/inputs/sysstat)
 * [system](./plugins/inputs/system)
     * cpu
@@ -211,6 +229,7 @@ configuration options.
     * processes
     * kernel (/proc/stat)
     * kernel (/proc/vmstat)
+    * linux_sysctl_fs (/proc/sys/fs)
 
 Telegraf can also collect metrics via the following service plugins:
 
@@ -221,32 +240,52 @@ Telegraf can also collect metrics via the following service plugins:
 * [nsq_consumer](./plugins/inputs/nsq_consumer)
 * [logparser](./plugins/inputs/logparser)
 * [statsd](./plugins/inputs/statsd)
+* [socket_listener](./plugins/inputs/socket_listener)
 * [tail](./plugins/inputs/tail)
-* [tcp_listener](./plugins/inputs/tcp_listener)
-* [udp_listener](./plugins/inputs/udp_listener)
+* [tcp_listener](./plugins/inputs/socket_listener)
+* [udp_listener](./plugins/inputs/socket_listener)
 * [webhooks](./plugins/inputs/webhooks)
   * [filestack](./plugins/inputs/webhooks/filestack)
   * [github](./plugins/inputs/webhooks/github)
   * [mandrill](./plugins/inputs/webhooks/mandrill)
+  * [papertrail](./plugins/inputs/webhooks/papertrail)
+  * [particle](./plugins/inputs/webhooks/particle)
   * [rollbar](./plugins/inputs/webhooks/rollbar)
+* [zipkin](./plugins/inputs/zipkin)
+
+Telegraf is able to parse the following input data formats into metrics, these
+formats may be used with input plugins supporting the `data_format` option:
+
+* [InfluxDB Line Protocol](./docs/DATA_FORMATS_INPUT.md#influx)
+* [JSON](./docs/DATA_FORMATS_INPUT.md#json)
+* [Graphite](./docs/DATA_FORMATS_INPUT.md#graphite)
+* [Value](./docs/DATA_FORMATS_INPUT.md#value)
+* [Nagios](./docs/DATA_FORMATS_INPUT.md#nagios)
+* [Collectd](./docs/DATA_FORMATS_INPUT.md#collectd)
+* [Dropwizard](./docs/DATA_FORMATS_INPUT.md#dropwizard)
 
 ## Processor Plugins
 
 * [printer](./plugins/processors/printer)
+* [override](./plugins/processors/override)
 
 ## Aggregator Plugins
 
+* [basicstats](./plugins/aggregators/basicstats)
 * [minmax](./plugins/aggregators/minmax)
+* [histogram](./plugins/aggregators/histogram)
 
 ## Output Plugins
 
 * [influxdb](./plugins/outputs/influxdb)
 * [amon](./plugins/outputs/amon)
-* [amqp](./plugins/outputs/amqp)
+* [amqp](./plugins/outputs/amqp) (rabbitmq)
 * [aws kinesis](./plugins/outputs/kinesis)
 * [aws cloudwatch](./plugins/outputs/cloudwatch)
+* [cratedb](./plugins/outputs/cratedb)
 * [datadog](./plugins/outputs/datadog)
 * [discard](./plugins/outputs/discard)
+* [elasticsearch](./plugins/outputs/elasticsearch)
 * [file](./plugins/outputs/file)
 * [graphite](./plugins/outputs/graphite)
 * [graylog](./plugins/outputs/graylog)
@@ -259,9 +298,8 @@ Telegraf can also collect metrics via the following service plugins:
 * [opentsdb](./plugins/outputs/opentsdb)
 * [prometheus](./plugins/outputs/prometheus_client)
 * [riemann](./plugins/outputs/riemann)
-
-## Contributing
-
-Please see the
-[contributing guide](CONTRIBUTING.md)
-for details on contributing a plugin to Telegraf.
+* [riemann_legacy](./plugins/outputs/riemann_legacy)
+* [socket_writer](./plugins/outputs/socket_writer)
+* [tcp](./plugins/outputs/socket_writer)
+* [udp](./plugins/outputs/socket_writer)
+* [wavefront](./plugins/outputs/wavefront)

accumulator.go

@@ -28,6 +28,18 @@ type Accumulator interface {
 		tags map[string]string,
 		t ...time.Time)
 
+	// AddSummary is the same as AddFields, but will add the metric as a "Summary" type
+	AddSummary(measurement string,
+		fields map[string]interface{},
+		tags map[string]string,
+		t ...time.Time)
+
+	// AddHistogram is the same as AddFields, but will add the metric as a "Histogram" type
+	AddHistogram(measurement string,
+		fields map[string]interface{},
+		tags map[string]string,
+		t ...time.Time)
+
 	SetPrecision(precision, interval time.Duration)
 
 	AddError(err error)

agent/accumulator.go

@@ -26,7 +26,7 @@ type MetricMaker interface {
 func NewAccumulator(
 	maker MetricMaker,
 	metrics chan telegraf.Metric,
-) *accumulator {
+) telegraf.Accumulator {
 	acc := accumulator{
 		maker:     maker,
 		metrics:   metrics,
@@ -76,6 +76,28 @@ func (ac *accumulator) AddCounter(
 	}
 }
 
+func (ac *accumulator) AddSummary(
+	measurement string,
+	fields map[string]interface{},
+	tags map[string]string,
+	t ...time.Time,
+) {
+	if m := ac.maker.MakeMetric(measurement, fields, tags, telegraf.Summary, ac.getTime(t)); m != nil {
+		ac.metrics <- m
+	}
+}
+
+func (ac *accumulator) AddHistogram(
+	measurement string,
+	fields map[string]interface{},
+	tags map[string]string,
+	t ...time.Time,
+) {
+	if m := ac.maker.MakeMetric(measurement, fields, tags, telegraf.Histogram, ac.getTime(t)); m != nil {
+		ac.metrics <- m
+	}
+}
+
 // AddError passes a runtime error to the accumulator.
 // The error will be tagged with the plugin name and written to the log.
 func (ac *accumulator) AddError(err error) {

agent/accumulator_test.go

@@ -15,63 +15,36 @@ import (
 	"github.com/stretchr/testify/require"
 )
 
-func TestAdd(t *testing.T) {
-	now := time.Now()
-	metrics := make(chan telegraf.Metric, 10)
-	defer close(metrics)
-	a := NewAccumulator(&TestMetricMaker{}, metrics)
-
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-
-	testm := <-metrics
-	actual := testm.String()
-	assert.Contains(t, actual, "acctest value=101")
-
-	testm = <-metrics
-	actual = testm.String()
-	assert.Contains(t, actual, "acctest,acc=test value=101")
-
-	testm = <-metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", now.UnixNano()),
-		actual)
-}
-
 func TestAddFields(t *testing.T) {
-	now := time.Now()
 	metrics := make(chan telegraf.Metric, 10)
 	defer close(metrics)
 	a := NewAccumulator(&TestMetricMaker{}, metrics)
 
+	tags := map[string]string{"foo": "bar"}
 	fields := map[string]interface{}{
 		"usage": float64(99),
 	}
-	a.AddFields("acctest", fields, map[string]string{})
-	a.AddGauge("acctest", fields, map[string]string{"acc": "test"})
-	a.AddCounter("acctest", fields, map[string]string{"acc": "test"}, now)
+	now := time.Now()
+	a.AddCounter("acctest", fields, tags, now)
 
 	testm := <-metrics
-	actual := testm.String()
-	assert.Contains(t, actual, "acctest usage=99")
 
-	testm = <-metrics
-	actual = testm.String()
-	assert.Contains(t, actual, "acctest,acc=test usage=99")
+	require.Equal(t, "acctest", testm.Name())
+	actual, ok := testm.GetField("usage")
 
-	testm = <-metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test usage=99 %d\n", now.UnixNano()),
-		actual)
+	require.True(t, ok)
+	require.Equal(t, float64(99), actual)
+
+	actual, ok = testm.GetTag("foo")
+	require.True(t, ok)
+	require.Equal(t, "bar", actual)
+
+	tm := testm.Time()
+	// okay if monotonic clock differs
+	require.True(t, now.Equal(tm))
+
+	tp := testm.Type()
+	require.Equal(t, telegraf.Counter, tp)
 }
 
 func TestAccAddError(t *testing.T) {
@@ -98,215 +71,61 @@ func TestAccAddError(t *testing.T) {
 	assert.Contains(t, string(errs[2]), "baz")
 }
 
-func TestAddNoIntervalWithPrecision(t *testing.T) {
-	now := time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC)
-	metrics := make(chan telegraf.Metric, 10)
-	defer close(metrics)
-	a := NewAccumulator(&TestMetricMaker{}, metrics)
-	a.SetPrecision(0, time.Second)
+func TestSetPrecision(t *testing.T) {
+	tests := []struct {
+		name      string
+		unset     bool
+		precision time.Duration
+		interval  time.Duration
+		timestamp time.Time
+		expected  time.Time
+	}{
+		{
+			name:      "default precision is nanosecond",
+			unset:     true,
+			timestamp: time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC),
+			expected:  time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC),
+		},
+		{
+			name:      "second interval",
+			interval:  time.Second,
+			timestamp: time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC),
+			expected:  time.Date(2006, time.February, 10, 12, 0, 0, 0, time.UTC),
+		},
+		{
+			name:      "microsecond interval",
+			interval:  time.Microsecond,
+			timestamp: time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC),
+			expected:  time.Date(2006, time.February, 10, 12, 0, 0, 82913000, time.UTC),
+		},
+		{
+			name:      "2 second precision",
+			precision: 2 * time.Second,
+			timestamp: time.Date(2006, time.February, 10, 12, 0, 2, 4, time.UTC),
+			expected:  time.Date(2006, time.February, 10, 12, 0, 2, 0, time.UTC),
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			metrics := make(chan telegraf.Metric, 10)
 
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
+			a := NewAccumulator(&TestMetricMaker{}, metrics)
+			if !tt.unset {
+				a.SetPrecision(tt.precision, tt.interval)
+			}
 
-	testm := <-a.metrics
-	actual := testm.String()
-	assert.Contains(t, actual, "acctest value=101")
+			a.AddFields("acctest",
+				map[string]interface{}{"value": float64(101)},
+				map[string]string{},
+				tt.timestamp,
+			)
 
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Contains(t, actual, "acctest,acc=test value=101")
+			testm := <-metrics
+			require.Equal(t, tt.expected, testm.Time())
 
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", int64(1139572800000000000)),
-		actual)
-}
-
-func TestAddDisablePrecision(t *testing.T) {
-	now := time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC)
-	metrics := make(chan telegraf.Metric, 10)
-	defer close(metrics)
-	a := NewAccumulator(&TestMetricMaker{}, metrics)
-
-	a.SetPrecision(time.Nanosecond, 0)
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-
-	testm := <-a.metrics
-	actual := testm.String()
-	assert.Contains(t, actual, "acctest value=101")
-
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Contains(t, actual, "acctest,acc=test value=101")
-
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", int64(1139572800082912748)),
-		actual)
-}
-
-func TestAddNoPrecisionWithInterval(t *testing.T) {
-	now := time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC)
-	metrics := make(chan telegraf.Metric, 10)
-	defer close(metrics)
-	a := NewAccumulator(&TestMetricMaker{}, metrics)
-
-	a.SetPrecision(0, time.Second)
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"})
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-
-	testm := <-a.metrics
-	actual := testm.String()
-	assert.Contains(t, actual, "acctest value=101")
-
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Contains(t, actual, "acctest,acc=test value=101")
-
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", int64(1139572800000000000)),
-		actual)
-}
-
-func TestDifferentPrecisions(t *testing.T) {
-	now := time.Date(2006, time.February, 10, 12, 0, 0, 82912748, time.UTC)
-	metrics := make(chan telegraf.Metric, 10)
-	defer close(metrics)
-	a := NewAccumulator(&TestMetricMaker{}, metrics)
-
-	a.SetPrecision(0, time.Second)
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-	testm := <-a.metrics
-	actual := testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", int64(1139572800000000000)),
-		actual)
-
-	a.SetPrecision(0, time.Millisecond)
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", int64(1139572800083000000)),
-		actual)
-
-	a.SetPrecision(0, time.Microsecond)
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", int64(1139572800082913000)),
-		actual)
-
-	a.SetPrecision(0, time.Nanosecond)
-	a.AddFields("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-	testm = <-a.metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", int64(1139572800082912748)),
-		actual)
-}
-
-func TestAddGauge(t *testing.T) {
-	now := time.Now()
-	metrics := make(chan telegraf.Metric, 10)
-	defer close(metrics)
-	a := NewAccumulator(&TestMetricMaker{}, metrics)
-
-	a.AddGauge("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{})
-	a.AddGauge("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"})
-	a.AddGauge("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-
-	testm := <-metrics
-	actual := testm.String()
-	assert.Contains(t, actual, "acctest value=101")
-	assert.Equal(t, testm.Type(), telegraf.Gauge)
-
-	testm = <-metrics
-	actual = testm.String()
-	assert.Contains(t, actual, "acctest,acc=test value=101")
-	assert.Equal(t, testm.Type(), telegraf.Gauge)
-
-	testm = <-metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", now.UnixNano()),
-		actual)
-	assert.Equal(t, testm.Type(), telegraf.Gauge)
-}
-
-func TestAddCounter(t *testing.T) {
-	now := time.Now()
-	metrics := make(chan telegraf.Metric, 10)
-	defer close(metrics)
-	a := NewAccumulator(&TestMetricMaker{}, metrics)
-
-	a.AddCounter("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{})
-	a.AddCounter("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"})
-	a.AddCounter("acctest",
-		map[string]interface{}{"value": float64(101)},
-		map[string]string{"acc": "test"}, now)
-
-	testm := <-metrics
-	actual := testm.String()
-	assert.Contains(t, actual, "acctest value=101")
-	assert.Equal(t, testm.Type(), telegraf.Counter)
-
-	testm = <-metrics
-	actual = testm.String()
-	assert.Contains(t, actual, "acctest,acc=test value=101")
-	assert.Equal(t, testm.Type(), telegraf.Counter)
-
-	testm = <-metrics
-	actual = testm.String()
-	assert.Equal(t,
-		fmt.Sprintf("acctest,acc=test value=101 %d\n", now.UnixNano()),
-		actual)
-	assert.Equal(t, testm.Type(), telegraf.Counter)
+			close(metrics)
+		})
+	}
 }
 
 type TestMetricMaker struct {

agent/agent.go

@@ -143,7 +143,7 @@ func (a *Agent) gatherer(
 func gatherWithTimeout(
 	shutdown chan struct{},
 	input *models.RunningInput,
-	acc *accumulator,
+	acc telegraf.Accumulator,
 	timeout time.Duration,
 ) {
 	ticker := time.NewTicker(timeout)
@@ -157,13 +157,13 @@ func gatherWithTimeout(
 		select {
 		case err := <-done:
 			if err != nil {
-				log.Printf("E! ERROR in input [%s]: %s", input.Name(), err)
+				acc.AddError(err)
 			}
 			return
 		case <-ticker.C:
-			log.Printf("E! ERROR: input [%s] took longer to collect than "+
-				"collection interval (%s)",
-				input.Name(), timeout)
+			err := fmt.Errorf("took longer to collect than collection interval (%s)",
+				timeout)
+			acc.AddError(err)
 			continue
 		case <-shutdown:
 			return
@@ -191,17 +191,18 @@ func (a *Agent) Test() error {
 	}()
 
 	for _, input := range a.Config.Inputs {
+		if _, ok := input.Input.(telegraf.ServiceInput); ok {
+			fmt.Printf("\nWARNING: skipping plugin [[%s]]: service inputs not supported in --test mode\n",
+				input.Name())
+			continue
+		}
+
 		acc := NewAccumulator(input, metricC)
 		acc.SetPrecision(a.Config.Agent.Precision.Duration,
 			a.Config.Agent.Interval.Duration)
 		input.SetTrace(true)
 		input.SetDefaultTags(a.Config.Tags)
 
-		fmt.Printf("* Plugin: %s, Collection 1\n", input.Name())
-		if input.Config.Interval != 0 {
-			fmt.Printf("* Internal: %s\n", input.Config.Interval)
-		}
-
 		if err := input.Input.Gather(acc); err != nil {
 			return err
 		}
@@ -209,9 +210,8 @@ func (a *Agent) Test() error {
 		// Special instructions for some inputs. cpu, for example, needs to be
 		// run twice in order to return cpu usage percentages.
 		switch input.Name() {
-		case "cpu", "mongodb", "procstat":
+		case "inputs.cpu", "inputs.mongodb", "inputs.procstat":
 			time.Sleep(500 * time.Millisecond)
-			fmt.Printf("* Plugin: %s, Collection 2\n", input.Name())
 			if err := input.Input.Gather(acc); err != nil {
 				return err
 			}
@@ -241,12 +241,12 @@ func (a *Agent) flush() {
 }
 
 // flusher monitors the metrics input channel and flushes on the minimum interval
-func (a *Agent) flusher(shutdown chan struct{}, metricC chan telegraf.Metric) error {
+func (a *Agent) flusher(shutdown chan struct{}, metricC chan telegraf.Metric, aggC chan telegraf.Metric) error {
 	// Inelegant, but this sleep is to allow the Gather threads to run, so that
 	// the flusher will flush after metrics are collected.
 	time.Sleep(time.Millisecond * 300)
 
-	// create an output metric channel and a gorouting that continously passes
+	// create an output metric channel and a gorouting that continuously passes
 	// each metric onto the output plugins & aggregators.
 	outMetricC := make(chan telegraf.Metric, 100)
 	var wg sync.WaitGroup
@@ -265,11 +265,9 @@ func (a *Agent) flusher(shutdown chan struct{}, metricC chan telegraf.Metric) er
 				// if dropOriginal is set to true, then we will only send this
 				// metric to the aggregators, not the outputs.
 				var dropOriginal bool
-				if !m.IsAggregate() {
-					for _, agg := range a.Config.Aggregators {
-						if ok := agg.Add(m.Copy()); ok {
-							dropOriginal = true
-						}
+				for _, agg := range a.Config.Aggregators {
+					if ok := agg.Add(m.Copy()); ok {
+						dropOriginal = true
 					}
 				}
 				if !dropOriginal {
@@ -285,7 +283,37 @@ func (a *Agent) flusher(shutdown chan struct{}, metricC chan telegraf.Metric) er
 		}
 	}()
 
+	wg.Add(1)
+	go func() {
+		defer wg.Done()
+		for {
+			select {
+			case <-shutdown:
+				if len(aggC) > 0 {
+					// keep going until aggC is flushed
+					continue
+				}
+				return
+			case metric := <-aggC:
+				metrics := []telegraf.Metric{metric}
+				for _, processor := range a.Config.Processors {
+					metrics = processor.Apply(metrics...)
+				}
+				for _, m := range metrics {
+					for i, o := range a.Config.Outputs {
+						if i == len(a.Config.Outputs)-1 {
+							o.AddMetric(m)
+						} else {
+							o.AddMetric(m.Copy())
+						}
+					}
+				}
+			}
+		}
+	}()
+
 	ticker := time.NewTicker(a.Config.Agent.FlushInterval.Duration)
+	semaphore := make(chan struct{}, 1)
 	for {
 		select {
 		case <-shutdown:
@@ -295,8 +323,18 @@ func (a *Agent) flusher(shutdown chan struct{}, metricC chan telegraf.Metric) er
 			a.flush()
 			return nil
 		case <-ticker.C:
-			internal.RandomSleep(a.Config.Agent.FlushJitter.Duration, shutdown)
-			a.flush()
+			go func() {
+				select {
+				case semaphore <- struct{}{}:
+					internal.RandomSleep(a.Config.Agent.FlushJitter.Duration, shutdown)
+					a.flush()
+					<-semaphore
+				default:
+					// skipping this flush because one is already happening
+					log.Println("W! Skipping a scheduled flush because there is" +
+						" already a flush ongoing.")
+				}
+			}()
 		case metric := <-metricC:
 			// NOTE potential bottleneck here as we put each metric through the
 			// processors serially.
@@ -322,6 +360,7 @@ func (a *Agent) Run(shutdown chan struct{}) error {
 
 	// channel shared between all input threads for accumulating metrics
 	metricC := make(chan telegraf.Metric, 100)
+	aggC := make(chan telegraf.Metric, 100)
 
 	// Start all ServicePlugins
 	for _, input := range a.Config.Inputs {
@@ -350,7 +389,7 @@ func (a *Agent) Run(shutdown chan struct{}) error {
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
-		if err := a.flusher(shutdown, metricC); err != nil {
+		if err := a.flusher(shutdown, metricC, aggC); err != nil {
 			log.Printf("E! Flusher routine failed, exiting: %s\n", err.Error())
 			close(shutdown)
 		}
@@ -360,7 +399,7 @@ func (a *Agent) Run(shutdown chan struct{}) error {
 	for _, aggregator := range a.Config.Aggregators {
 		go func(agg *models.RunningAggregator) {
 			defer wg.Done()
-			acc := NewAccumulator(agg, metricC)
+			acc := NewAccumulator(agg, aggC)
 			acc.SetPrecision(a.Config.Agent.Precision.Duration,
 				a.Config.Agent.Interval.Duration)
 			agg.Run(acc, shutdown)
@@ -381,5 +420,6 @@ func (a *Agent) Run(shutdown chan struct{}) error {
 	}
 
 	wg.Wait()
+	a.Close()
 	return nil
 }

appveyor.yml

@@ -0,0 +1,34 @@
+image: Previous Visual Studio 2015
+version: "{build}"
+
+cache:
+ - C:\Cache
+
+clone_folder: C:\gopath\src\github.com\influxdata\telegraf
+
+environment:
+  GOPATH: C:\gopath
+
+platform: x64
+
+install:
+  - IF NOT EXIST "C:\Cache" mkdir C:\Cache
+  - IF NOT EXIST "C:\Cache\go1.10.1.msi" curl -o "C:\Cache\go1.10.1.msi" https://storage.googleapis.com/golang/go1.10.1.windows-amd64.msi
+  - IF NOT EXIST "C:\Cache\gnuwin32-bin.zip" curl -o "C:\Cache\gnuwin32-bin.zip" https://dl.influxdata.com/telegraf/ci/make-3.81-bin.zip
+  - IF NOT EXIST "C:\Cache\gnuwin32-dep.zip" curl -o "C:\Cache\gnuwin32-dep.zip" https://dl.influxdata.com/telegraf/ci/make-3.81-dep.zip
+  - IF EXIST "C:\Go" rmdir /S /Q C:\Go
+  - msiexec.exe /i "C:\Cache\go1.10.1.msi" /quiet
+  - 7z x "C:\Cache\gnuwin32-bin.zip" -oC:\GnuWin32 -y
+  - 7z x "C:\Cache\gnuwin32-dep.zip" -oC:\GnuWin32 -y
+  - go version
+  - go env
+
+build_script:
+  - cmd: C:\GnuWin32\bin\make deps
+  - cmd: C:\GnuWin32\bin\make telegraf
+
+test_script:
+  - cmd: C:\GnuWin32\bin\make test-windows
+
+artifacts:
+  - path: telegraf.exe

circle.yml

@@ -1,18 +0,0 @@
-machine:
-  services:
-    - docker
-  post:
-    - sudo service zookeeper stop
-    - go version
-    - go version | grep 1.7.4 || sudo rm -rf /usr/local/go
-    - wget https://storage.googleapis.com/golang/go1.7.4.linux-amd64.tar.gz
-    - sudo tar -C /usr/local -xzf go1.7.4.linux-amd64.tar.gz
-    - go version
-
-dependencies:
-  override:
-    - docker info
-
-test:
-  override:
-    - bash scripts/circle-test.sh

cmd/telegraf/telegraf.go

@@ -4,6 +4,8 @@ import (
 	"flag"
 	"fmt"
 	"log"
+	"net/http"
+	_ "net/http/pprof" // Comment this line to disable pprof endpoint.
 	"os"
 	"os/signal"
 	"runtime"
@@ -24,6 +26,8 @@ import (
 
 var fDebug = flag.Bool("debug", false,
 	"turn on debug logging")
+var pprofAddr = flag.String("pprof-addr", "",
+	"pprof address to listen on, not activate pprof if empty")
 var fQuiet = flag.Bool("quiet", false,
 	"run in quiet mode")
 var fTest = flag.Bool("test", false, "gather metrics, print them out, and exit")
@@ -47,16 +51,16 @@ var fAggregatorFilters = flag.String("aggregator-filter", "",
 var fProcessorFilters = flag.String("processor-filter", "",
 	"filter the processors to enable, separator is :")
 var fUsage = flag.String("usage", "",
-	"print usage for a plugin, ie, 'telegraf -usage mysql'")
+	"print usage for a plugin, ie, 'telegraf --usage mysql'")
 var fService = flag.String("service", "",
 	"operate on the service")
+var fRunAsConsole = flag.Bool("console", false, "run as console application (windows only)")
 
-// Telegraf version, populated linker.
-//   ie, -ldflags "-X main.version=`git describe --always --tags`"
 var (
-	version string
-	commit  string
-	branch  string
+	nextVersion = "1.7.0"
+	version     string
+	commit      string
+	branch      string
 )
 
 func init() {
@@ -69,134 +73,19 @@ func init() {
 	}
 }
 
-const usage = `Telegraf, The plugin-driven server agent for collecting and reporting metrics.
-
-Usage:
-
-  telegraf [commands|flags]
-
-The commands & flags are:
-
-  config             print out full sample configuration to stdout
-  version            print the version to stdout
-
-  --config <file>     configuration file to load
-  --test              gather metrics once, print them to stdout, and exit
-  --config-directory  directory containing additional *.conf files
-  --input-filter      filter the input plugins to enable, separator is :
-  --output-filter     filter the output plugins to enable, separator is :
-  --usage             print usage for a plugin, ie, 'telegraf --usage mysql'
-  --debug             print metrics as they're generated to stdout
-  --quiet             run in quiet mode
-
-Examples:
-
-  # generate a telegraf config file:
-  telegraf config > telegraf.conf
-
-  # generate config with only cpu input & influxdb output plugins defined
-  telegraf --input-filter cpu --output-filter influxdb config
-
-  # run a single telegraf collection, outputing metrics to stdout
-  telegraf --config telegraf.conf -test
-
-  # run telegraf with all plugins defined in config file
-  telegraf --config telegraf.conf
-
-  # run telegraf, enabling the cpu & memory input, and influxdb output plugins
-  telegraf --config telegraf.conf --input-filter cpu:mem --output-filter influxdb
-`
-
 var stop chan struct{}
 
-var srvc service.Service
-
-type program struct{}
-
-func reloadLoop(stop chan struct{}, s service.Service) {
-	defer func() {
-		if service.Interactive() {
-			os.Exit(0)
-		}
-		return
-	}()
+func reloadLoop(
+	stop chan struct{},
+	inputFilters []string,
+	outputFilters []string,
+	aggregatorFilters []string,
+	processorFilters []string,
+) {
 	reload := make(chan bool, 1)
 	reload <- true
 	for <-reload {
 		reload <- false
-		flag.Parse()
-		args := flag.Args()
-
-		var inputFilters []string
-		if *fInputFilters != "" {
-			inputFilter := strings.TrimSpace(*fInputFilters)
-			inputFilters = strings.Split(":"+inputFilter+":", ":")
-		}
-		var outputFilters []string
-		if *fOutputFilters != "" {
-			outputFilter := strings.TrimSpace(*fOutputFilters)
-			outputFilters = strings.Split(":"+outputFilter+":", ":")
-		}
-		var aggregatorFilters []string
-		if *fAggregatorFilters != "" {
-			aggregatorFilter := strings.TrimSpace(*fAggregatorFilters)
-			aggregatorFilters = strings.Split(":"+aggregatorFilter+":", ":")
-		}
-		var processorFilters []string
-		if *fProcessorFilters != "" {
-			processorFilter := strings.TrimSpace(*fProcessorFilters)
-			processorFilters = strings.Split(":"+processorFilter+":", ":")
-		}
-
-		if len(args) > 0 {
-			switch args[0] {
-			case "version":
-				fmt.Printf("Telegraf v%s (git: %s %s)\n", version, branch, commit)
-				return
-			case "config":
-				config.PrintSampleConfig(
-					inputFilters,
-					outputFilters,
-					aggregatorFilters,
-					processorFilters,
-				)
-				return
-			}
-		}
-
-		// switch for flags which just do something and exit immediately
-		switch {
-		case *fOutputList:
-			fmt.Println("Available Output Plugins:")
-			for k, _ := range outputs.Outputs {
-				fmt.Printf("  %s\n", k)
-			}
-			return
-		case *fInputList:
-			fmt.Println("Available Input Plugins:")
-			for k, _ := range inputs.Inputs {
-				fmt.Printf("  %s\n", k)
-			}
-			return
-		case *fVersion:
-			fmt.Printf("Telegraf v%s (git: %s %s)\n", version, branch, commit)
-			return
-		case *fSampleConfig:
-			config.PrintSampleConfig(
-				inputFilters,
-				outputFilters,
-				aggregatorFilters,
-				processorFilters,
-			)
-			return
-		case *fUsage != "":
-			if err := config.PrintInputConfig(*fUsage); err != nil {
-				if err2 := config.PrintOutputConfig(*fUsage); err2 != nil {
-					log.Fatalf("E! %s and %s", err, err2)
-				}
-			}
-			return
-		}
 
 		// If no other options are specified, load the config file and run.
 		c := config.NewConfig()
@@ -213,13 +102,23 @@ func reloadLoop(stop chan struct{}, s service.Service) {
 				log.Fatal("E! " + err.Error())
 			}
 		}
-		if len(c.Outputs) == 0 {
+		if !*fTest && len(c.Outputs) == 0 {
 			log.Fatalf("E! Error: no outputs found, did you provide a valid config file?")
 		}
 		if len(c.Inputs) == 0 {
 			log.Fatalf("E! Error: no inputs found, did you provide a valid config file?")
 		}
 
+		if int64(c.Agent.Interval.Duration) <= 0 {
+			log.Fatalf("E! Agent interval must be positive, found %s",
+				c.Agent.Interval.Duration)
+		}
+
+		if int64(c.Agent.FlushInterval.Duration) <= 0 {
+			log.Fatalf("E! Agent flush_interval must be positive; found %s",
+				c.Agent.Interval.Duration)
+		}
+
 		ag, err := agent.NewAgent(c)
 		if err != nil {
 			log.Fatal("E! " + err.Error())
@@ -237,7 +136,7 @@ func reloadLoop(stop chan struct{}, s service.Service) {
 			if err != nil {
 				log.Fatal("E! " + err.Error())
 			}
-			return
+			os.Exit(0)
 		}
 
 		err = ag.Connect()
@@ -265,20 +164,27 @@ func reloadLoop(stop chan struct{}, s service.Service) {
 			}
 		}()
 
-		log.Printf("I! Starting Telegraf (version %s)\n", version)
+		log.Printf("I! Starting Telegraf %s\n", displayVersion())
 		log.Printf("I! Loaded outputs: %s", strings.Join(c.OutputNames(), " "))
 		log.Printf("I! Loaded inputs: %s", strings.Join(c.InputNames(), " "))
 		log.Printf("I! Tags enabled: %s", c.ListTags())
 
 		if *fPidfile != "" {
-			f, err := os.Create(*fPidfile)
+			f, err := os.OpenFile(*fPidfile, os.O_CREATE|os.O_WRONLY, 0644)
 			if err != nil {
-				log.Fatalf("E! Unable to create pidfile: %s", err)
+				log.Printf("E! Unable to create pidfile: %s", err)
+			} else {
+				fmt.Fprintf(f, "%d\n", os.Getpid())
+
+				f.Close()
+
+				defer func() {
+					err := os.Remove(*fPidfile)
+					if err != nil {
+						log.Printf("E! Unable to remove pidfile: %s", err)
+					}
+				}()
 			}
-
-			fmt.Fprintf(f, "%d\n", os.Getpid())
-
-			f.Close()
 		}
 
 		ag.Run(shutdown)
@@ -290,47 +196,160 @@ func usageExit(rc int) {
 	os.Exit(rc)
 }
 
+type program struct {
+	inputFilters      []string
+	outputFilters     []string
+	aggregatorFilters []string
+	processorFilters  []string
+}
+
 func (p *program) Start(s service.Service) error {
-	srvc = s
 	go p.run()
 	return nil
 }
 func (p *program) run() {
 	stop = make(chan struct{})
-	reloadLoop(stop, srvc)
+	reloadLoop(
+		stop,
+		p.inputFilters,
+		p.outputFilters,
+		p.aggregatorFilters,
+		p.processorFilters,
+	)
 }
 func (p *program) Stop(s service.Service) error {
 	close(stop)
 	return nil
 }
 
+func displayVersion() string {
+	if version == "" {
+		return fmt.Sprintf("v%s~%s", nextVersion, commit)
+	}
+	return "v" + version
+}
+
 func main() {
 	flag.Usage = func() { usageExit(0) }
 	flag.Parse()
-	if runtime.GOOS == "windows" {
+	args := flag.Args()
+
+	inputFilters, outputFilters := []string{}, []string{}
+	if *fInputFilters != "" {
+		inputFilters = strings.Split(":"+strings.TrimSpace(*fInputFilters)+":", ":")
+	}
+	if *fOutputFilters != "" {
+		outputFilters = strings.Split(":"+strings.TrimSpace(*fOutputFilters)+":", ":")
+	}
+
+	aggregatorFilters, processorFilters := []string{}, []string{}
+	if *fAggregatorFilters != "" {
+		aggregatorFilters = strings.Split(":"+strings.TrimSpace(*fAggregatorFilters)+":", ":")
+	}
+	if *fProcessorFilters != "" {
+		processorFilters = strings.Split(":"+strings.TrimSpace(*fProcessorFilters)+":", ":")
+	}
+
+	if *pprofAddr != "" {
+		go func() {
+			pprofHostPort := *pprofAddr
+			parts := strings.Split(pprofHostPort, ":")
+			if len(parts) == 2 && parts[0] == "" {
+				pprofHostPort = fmt.Sprintf("localhost:%s", parts[1])
+			}
+			pprofHostPort = "http://" + pprofHostPort + "/debug/pprof"
+
+			log.Printf("I! Starting pprof HTTP server at: %s", pprofHostPort)
+
+			if err := http.ListenAndServe(*pprofAddr, nil); err != nil {
+				log.Fatal("E! " + err.Error())
+			}
+		}()
+	}
+
+	if len(args) > 0 {
+		switch args[0] {
+		case "version":
+			fmt.Printf("Telegraf %s (git: %s %s)\n", displayVersion(), branch, commit)
+			return
+		case "config":
+			config.PrintSampleConfig(
+				inputFilters,
+				outputFilters,
+				aggregatorFilters,
+				processorFilters,
+			)
+			return
+		}
+	}
+
+	// switch for flags which just do something and exit immediately
+	switch {
+	case *fOutputList:
+		fmt.Println("Available Output Plugins:")
+		for k, _ := range outputs.Outputs {
+			fmt.Printf("  %s\n", k)
+		}
+		return
+	case *fInputList:
+		fmt.Println("Available Input Plugins:")
+		for k, _ := range inputs.Inputs {
+			fmt.Printf("  %s\n", k)
+		}
+		return
+	case *fVersion:
+		fmt.Printf("Telegraf %s (git: %s %s)\n", displayVersion(), branch, commit)
+		return
+	case *fSampleConfig:
+		config.PrintSampleConfig(
+			inputFilters,
+			outputFilters,
+			aggregatorFilters,
+			processorFilters,
+		)
+		return
+	case *fUsage != "":
+		err := config.PrintInputConfig(*fUsage)
+		err2 := config.PrintOutputConfig(*fUsage)
+		if err != nil && err2 != nil {
+			log.Fatalf("E! %s and %s", err, err2)
+		}
+		return
+	}
+
+	if runtime.GOOS == "windows" && !(*fRunAsConsole) {
 		svcConfig := &service.Config{
 			Name:        "telegraf",
 			DisplayName: "Telegraf Data Collector Service",
 			Description: "Collects data using a series of plugins and publishes it to" +
 				"another series of plugins.",
-			Arguments: []string{"-config", "C:\\Program Files\\Telegraf\\telegraf.conf"},
+			Arguments: []string{"--config", "C:\\Program Files\\Telegraf\\telegraf.conf"},
 		}
 
-		prg := &program{}
+		prg := &program{
+			inputFilters:      inputFilters,
+			outputFilters:     outputFilters,
+			aggregatorFilters: aggregatorFilters,
+			processorFilters:  processorFilters,
+		}
 		s, err := service.New(prg, svcConfig)
 		if err != nil {
 			log.Fatal("E! " + err.Error())
 		}
-		// Handle the -service flag here to prevent any issues with tooling that
+		// Handle the --service flag here to prevent any issues with tooling that
 		// may not have an interactive session, e.g. installing from Ansible.
 		if *fService != "" {
 			if *fConfig != "" {
-				(*svcConfig).Arguments = []string{"-config", *fConfig}
+				(*svcConfig).Arguments = []string{"--config", *fConfig}
+			}
+			if *fConfigDirectory != "" {
+				(*svcConfig).Arguments = append((*svcConfig).Arguments, "--config-directory", *fConfigDirectory)
 			}
 			err := service.Control(s, *fService)
 			if err != nil {
 				log.Fatal("E! " + err.Error())
 			}
+			os.Exit(0)
 		} else {
 			err = s.Run()
 			if err != nil {
@@ -339,6 +358,12 @@ func main() {
 		}
 	} else {
 		stop = make(chan struct{})
-		reloadLoop(stop, nil)
+		reloadLoop(
+			stop,
+			inputFilters,
+			outputFilters,
+			aggregatorFilters,
+			processorFilters,
+		)
 	}
 }

cmd/telegraf/usage.go

@@ -0,0 +1,45 @@
+// +build !windows
+
+package main
+
+const usage = `Telegraf, The plugin-driven server agent for collecting and reporting metrics.
+
+Usage:
+
+  telegraf [commands|flags]
+
+The commands & flags are:
+
+  config              print out full sample configuration to stdout
+  version             print the version to stdout
+
+  --config <file>     configuration file to load
+  --test              gather metrics once, print them to stdout, and exit
+  --config-directory  directory containing additional *.conf files
+  --input-filter      filter the input plugins to enable, separator is :
+  --output-filter     filter the output plugins to enable, separator is :
+  --usage             print usage for a plugin, ie, 'telegraf --usage mysql'
+  --debug             print metrics as they're generated to stdout
+  --pprof-addr        pprof address to listen on, format: localhost:6060 or :6060
+  --quiet             run in quiet mode
+
+Examples:
+
+  # generate a telegraf config file:
+  telegraf config > telegraf.conf
+
+  # generate config with only cpu input & influxdb output plugins defined
+  telegraf --input-filter cpu --output-filter influxdb config
+
+  # run a single telegraf collection, outputing metrics to stdout
+  telegraf --config telegraf.conf --test
+
+  # run telegraf with all plugins defined in config file
+  telegraf --config telegraf.conf
+
+  # run telegraf, enabling the cpu & memory input, and influxdb output plugins
+  telegraf --config telegraf.conf --input-filter cpu:mem --output-filter influxdb
+
+  # run telegraf with pprof
+  telegraf --config telegraf.conf --pprof-addr localhost:6060
+`

cmd/telegraf/usage_windows.go

@@ -0,0 +1,54 @@
+// +build windows
+
+package main
+
+const usage = `Telegraf, The plugin-driven server agent for collecting and reporting metrics.
+
+Usage:
+
+  telegraf [commands|flags]
+
+The commands & flags are:
+
+  config              print out full sample configuration to stdout
+  version             print the version to stdout
+
+  --config <file>     configuration file to load
+  --test              gather metrics once, print them to stdout, and exit
+  --config-directory  directory containing additional *.conf files
+  --input-filter      filter the input plugins to enable, separator is :
+  --output-filter     filter the output plugins to enable, separator is :
+  --usage             print usage for a plugin, ie, 'telegraf --usage mysql'
+  --debug             print metrics as they're generated to stdout
+  --pprof-addr        pprof address to listen on, format: localhost:6060 or :6060
+  --quiet             run in quiet mode
+
+  --console           run as console application
+  --service           operate on service, one of: install, uninstall, start, stop
+
+Examples:
+
+  # generate a telegraf config file:
+  telegraf config > telegraf.conf
+
+  # generate config with only cpu input & influxdb output plugins defined
+  telegraf --input-filter cpu --output-filter influxdb config
+
+  # run a single telegraf collection, outputing metrics to stdout
+  telegraf --config telegraf.conf --test
+
+  # run telegraf with all plugins defined in config file
+  telegraf --config telegraf.conf
+
+  # run telegraf, enabling the cpu & memory input, and influxdb output plugins
+  telegraf --config telegraf.conf --input-filter cpu:mem --output-filter influxdb
+
+  # run telegraf with pprof
+  telegraf --config telegraf.conf --pprof-addr localhost:6060
+
+  # run telegraf without service controller
+  telegraf --console install --config "C:\Program Files\Telegraf\telegraf.conf"
+
+  # install telegraf service
+  telegraf --service install --config "C:\Program Files\Telegraf\telegraf.conf"
+`

docker-compose.yml

@@ -0,0 +1,93 @@
+version: '3'
+
+services:
+  aerospike:
+    image: aerospike/aerospike-server:3.9.0
+    ports:
+      - "3000:3000"
+  zookeeper:
+    image: wurstmeister/zookeeper
+    environment:
+      - JAVA_OPTS="-Xms256m -Xmx256m"
+    ports:
+      - "2181:2181"
+  kafka:
+    image: wurstmeister/kafka
+    environment:
+      - KAFKA_ADVERTISED_HOST_NAME=localhost
+      - KAFKA_ADVERTISED_PORT=9092
+      - KAFKA_ZOOKEEPER_CONNECT=zookeeper:2181
+      - KAFKA_CREATE_TOPICS="test:1:1"
+      - JAVA_OPTS="-Xms256m -Xmx256m"
+    ports:
+      - "9092:9092"
+    depends_on:
+      - zookeeper
+  elasticsearch:
+    image: elasticsearch:5
+    environment:
+    - JAVA_OPTS="-Xms256m -Xmx256m"
+    ports:
+      - "9200:9200"
+      - "9300:9300"
+  mysql:
+    image: mysql
+    environment:
+      - MYSQL_ALLOW_EMPTY_PASSWORD=yes
+    ports:
+      - "3306:3306"
+  memcached:
+    image: memcached
+    ports: 
+      - "11211:11211"
+  postgres:
+    image: postgres:alpine
+    ports:
+      - "5432:5432"
+  rabbitmq:
+    image: rabbitmq:3-management
+    ports:
+      - "15672:15672"
+      - "5672:5672"
+  redis:
+    image: redis:alpine
+    ports:
+      - "6379:6379"
+  nsq:
+    image: nsqio/nsq
+    ports:
+      - "4150:4150"
+    command: "/nsqd"
+  mqtt:
+    image: ncarlier/mqtt
+    ports:
+      - "1883:1883"
+  riemann:
+    image: stealthly/docker-riemann
+    ports:
+      - "5555:5555"
+  nats:
+    image: nats
+    ports:
+      - "4222:4222"
+  openldap:
+    image: cobaugh/openldap-alpine
+    environment:
+      - SLAPD_CONFIG_ROOTDN="cn=manager,cn=config"
+      - SLAPD_CONFIG_ROOTPW="secret"
+    ports:
+      - "389:389"
+      - "636:636"
+  crate:
+   image: crate/crate
+   ports:
+    - "4200:4200"
+    - "4230:4230"
+   command:
+    - crate
+    - -Cnetwork.host=0.0.0.0
+    - -Ctransport.host=localhost
+    - -Clicense.enterprise=false
+   environment:
+    - CRATE_HEAP_SIZE=128m
+    - JAVA_OPTS='-Xms256m -Xmx256m'

docs/AGGREGATORS_AND_PROCESSORS.md

@@ -39,6 +39,11 @@ metrics as they pass through Telegraf:
 
 Both Aggregators and Processors analyze metrics as they pass through Telegraf.
 
+Use [measurement filtering](CONFIGURATION.md#measurement-filtering)
+to control which metrics are passed through a processor or aggregator.  If a
+metric is filtered out the metric bypasses the plugin and is passed downstream
+to the next plugin.
+
 **Processor** plugins process metrics as they pass through and immediately emit
 results based on the values they process. For example, this could be printing
 all metrics or adding a tag to all metrics that pass through.

docs/CONFIGURATION.md

@@ -24,6 +24,22 @@ Environment variables can be used anywhere in the config file, simply prepend
 them with $. For strings the variable must be within quotes (ie, "$STR_VAR"),
 for numbers and booleans they should be plain (ie, $INT_VAR, $BOOL_VAR)
 
+When using the `.deb` or `.rpm` packages, you can define environment variables
+in the `/etc/default/telegraf` file.
+
+## Configuration file locations
+
+The location of the configuration file can be set via the `--config` command
+line flag.
+
+When the `--config-directory` command line flag is used files ending with
+`.conf` in the specified directory will also be included in the Telegraf
+configuration.
+
+On most systems, the default locations are `/etc/telegraf/telegraf.conf` for
+the main configuration file and `/etc/telegraf/telegraf.d` for the directory of
+configuration files.
+
 # Global Tags
 
 Global tags can be specified in the `[global_tags]` section of the config file
@@ -56,11 +72,14 @@ interval. Maximum flush_interval will be flush_interval + flush_jitter
 This is primarily to avoid
 large write spikes for users running a large number of telegraf instances.
 ie, a jitter of 5s and flush_interval 10s means flushes will happen every 10-15s.
-* **precision**: By default, precision will be set to the same timestamp order
-as the collection interval, with the maximum being 1s. Precision will NOT
-be used for service inputs, such as logparser and statsd. Valid values are
-"ns", "us" (or "µs"), "ms", "s".
-* **logfile**: Specify the log file name. The empty string means to log to stdout.
+* **precision**:
+   By default or when set to "0s", precision will be set to the same
+   timestamp order as the collection interval, with the maximum being 1s.
+   Precision will NOT be used for service inputs. It is up to each individual
+   service input to set the timestamp at the appropriate precision.
+   Valid time units are "ns", "us" (or "µs"), "ms", "s".
+
+* **logfile**: Specify the log file name. The empty string means to log to stderr.
 * **debug**: Run telegraf in debug mode.
 * **quiet**: Run telegraf in quiet mode (error messages only).
 * **hostname**: Override default hostname, if empty use os.Hostname().
@@ -79,9 +98,13 @@ you can configure that here.
 * **name_suffix**: Specifies a suffix to attach to the measurement name.
 * **tags**: A map of tags to apply to a specific input's measurements.
 
+The [measurement filtering](#measurement-filtering) parameters can be used to
+limit what metrics are emitted from the input plugin.
+
 ## Output Configuration
 
-There are no generic configuration options available for all outputs.
+The [measurement filtering](#measurement-filtering) parameters can be used to
+limit what metrics are emitted from the output plugin.
 
 ## Aggregator Configuration
 
@@ -102,6 +125,10 @@ aggregator and will not get sent to the output plugins.
 * **name_suffix**: Specifies a suffix to attach to the measurement name.
 * **tags**: A map of tags to apply to a specific input's measurements.
 
+The [measurement filtering](#measurement-filtering) parameters can be used to
+limit what metrics are handled by the aggregator.  Excluded metrics are passed
+downstream to the next aggregator.
+
 ## Processor Configuration
 
 The following config parameters are available for all processors:
@@ -109,36 +136,50 @@ The following config parameters are available for all processors:
 * **order**: This is the order in which the processor(s) get executed. If this
 is not specified then processor execution order will be random.
 
+The [measurement filtering](#measurement-filtering) parameters can be used
+to limit what metrics are handled by the processor.  Excluded metrics are
+passed downstream to the next processor.
+
 #### Measurement Filtering
 
 Filters can be configured per input, output, processor, or aggregator,
 see below for examples.
 
-* **namepass**: An array of strings that is used to filter metrics generated by the
-current input. Each string in the array is tested as a glob match against
-measurement names and if it matches, the field is emitted.
-* **namedrop**: The inverse of pass, if a measurement name matches, it is not emitted.
-* **fieldpass**: An array of strings that is used to filter metrics generated by the
-current input. Each string in the array is tested as a glob match against field names
-and if it matches, the field is emitted. fieldpass is not available for outputs.
-* **fielddrop**: The inverse of pass, if a field name matches, it is not emitted.
-fielddrop is not available for outputs.
-* **tagpass**: tag names and arrays of strings that are used to filter
-measurements by the current input. Each string in the array is tested as a glob
-match against the tag name, and if it matches the measurement is emitted.
-* **tagdrop**: The inverse of tagpass. If a tag matches, the measurement is not
-emitted. This is tested on measurements that have passed the tagpass test.
-* **tagexclude**: tagexclude can be used to exclude a tag from measurement(s).
-As opposed to tagdrop, which will drop an entire measurement based on it's
-tags, tagexclude simply strips the given tag keys from the measurement. This
-can be used on inputs & outputs, but it is _recommended_ to be used on inputs,
-as it is more efficient to filter out tags at the ingestion point.
-* **taginclude**: taginclude is the inverse of tagexclude. It will only include
-the tag keys in the final measurement.
+* **namepass**:
+An array of glob pattern strings.  Only points whose measurement name matches
+a pattern in this list are emitted.
+* **namedrop**:
+The inverse of `namepass`.  If a match is found the point is discarded. This
+is tested on points after they have passed the `namepass` test.
+* **fieldpass**:
+An array of glob pattern strings.  Only fields whose field key matches a
+pattern in this list are emitted.
+* **fielddrop**:
+The inverse of `fieldpass`.  Fields with a field key matching one of the
+patterns will be discarded from the point.  This is tested on points after
+they have passed the `fieldpass` test.
+* **tagpass**:
+A table mapping tag keys to arrays of glob pattern strings.  Only points
+that contain a tag key in the table and a tag value matching one of its
+patterns is emitted.
+* **tagdrop**:
+The inverse of `tagpass`.  If a match is found the point is discarded. This
+is tested on points after they have passed the `tagpass` test.
+* **taginclude**:
+An array of glob pattern strings.  Only tags with a tag key matching one of
+the patterns are emitted.  In contrast to `tagpass`, which will pass an entire
+point based on its tag, `taginclude` removes all non matching tags from the
+point.  This filter can be used on both inputs & outputs, but it is
+_recommended_ to be used on inputs, as it is more efficient to filter out tags
+at the ingestion point.
+* **tagexclude**:
+The inverse of `taginclude`. Tags with a tag key matching one of the patterns
+will be discarded from the point.
 
-**NOTE** `tagpass` and `tagdrop` parameters must be defined at the _end_ of
-the plugin definition, otherwise subsequent plugin config options will be
-interpreted as part of the tagpass/tagdrop map.
+**NOTE** Due to the way TOML is parsed, `tagpass` and `tagdrop` parameters
+must be defined at the _end_ of the plugin definition, otherwise subsequent
+plugin config options will be interpreted as part of the tagpass/tagdrop
+tables.
 
 #### Input Configuration Examples
 
@@ -158,7 +199,6 @@ fields which begin with `time_`.
 [[outputs.influxdb]]
   url = "http://192.168.59.103:8086" # required.
   database = "telegraf" # required.
-  precision = "s"
 
 # INPUTS
 [[inputs.cpu]]
@@ -297,21 +337,18 @@ to avoid measurement collisions:
 [[outputs.influxdb]]
   urls = [ "http://localhost:8086" ]
   database = "telegraf"
-  precision = "s"
   # Drop all measurements that start with "aerospike"
   namedrop = ["aerospike*"]
 
 [[outputs.influxdb]]
   urls = [ "http://localhost:8086" ]
   database = "telegraf-aerospike-data"
-  precision = "s"
   # Only accept aerospike data:
   namepass = ["aerospike*"]
 
 [[outputs.influxdb]]
   urls = [ "http://localhost:8086" ]
   database = "telegraf-cpu0-data"
-  precision = "s"
   # Only store measurements where the tag "cpu" matches the value "cpu0"
   [outputs.influxdb.tagpass]
     cpu = ["cpu0"]
@@ -351,4 +388,16 @@ to the system load metrics due to the `namepass` parameter.
 
 [[outputs.file]]
   files = ["stdout"]
-```
+```
+
+#### Processor Configuration Examples:
+
+Print only the metrics with `cpu` as the measurement name, all metrics are
+passed to the output:
+```toml
+[[processors.printer]]
+  namepass = "cpu"
+
+[[outputs.file]]
+  files = ["/tmp/metrics.out"]
+```

docs/DATA_FORMATS_INPUT.md

@@ -7,6 +7,8 @@ Telegraf is able to parse the following input data formats into metrics:
 1. [Graphite](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md#graphite)
 1. [Value](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md#value), ie: 45 or "booyah"
 1. [Nagios](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md#nagios) (exec input only)
+1. [Collectd](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md#collectd)
+1. [Dropwizard](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md#dropwizard)
 
 Telegraf metrics, like InfluxDB
 [points](https://docs.influxdata.com/influxdb/v0.10/write_protocols/line/),
@@ -40,7 +42,7 @@ example, in the exec plugin:
   name_suffix = "_mycollector"
 
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "json"
@@ -67,7 +69,7 @@ metrics are parsed directly into Telegraf metrics.
   name_suffix = "_mycollector"
 
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "influx"
@@ -117,7 +119,7 @@ For example, if you had this configuration:
   name_suffix = "_mycollector"
 
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "json"
@@ -161,7 +163,7 @@ For example, if the following configuration:
   name_suffix = "_mycollector"
 
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "json"
@@ -232,7 +234,7 @@ name of the plugin.
   name_override = "entropy_available"
 
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "value"
@@ -390,7 +392,7 @@ There are many more options available,
   name_suffix = "_mycollector"
 
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "graphite"
@@ -427,14 +429,227 @@ Note: Nagios Input Data Formats is only supported in `exec` input plugin.
 ```toml
 [[inputs.exec]]
   ## Commands array
-  commands = ["/usr/lib/nagios/plugins/check_load", "-w 5,6,7 -c 7,8,9"]
+  commands = ["/usr/lib/nagios/plugins/check_load -w 5,6,7 -c 7,8,9"]
 
   ## measurement name suffix (for separating different commands)
   name_suffix = "_mycollector"
 
   ## Data format to consume.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
   data_format = "nagios"
 ```
+
+# Collectd:
+
+The collectd format parses the collectd binary network protocol.  Tags are
+created for host, instance, type, and type instance.  All collectd values are
+added as float64 fields.
+
+For more information about the binary network protocol see
+[here](https://collectd.org/wiki/index.php/Binary_protocol).
+
+You can control the cryptographic settings with parser options.  Create an
+authentication file and set `collectd_auth_file` to the path of the file, then
+set the desired security level in `collectd_security_level`.
+
+Additional information including client setup can be found
+[here](https://collectd.org/wiki/index.php/Networking_introduction#Cryptographic_setup).
+
+You can also change the path to the typesdb or add additional typesdb using
+`collectd_typesdb`.
+
+#### Collectd Configuration:
+
+```toml
+[[inputs.socket_listener]]
+  service_address = "udp://127.0.0.1:25826"
+  name_prefix = "collectd_"
+
+  ## Data format to consume.
+  ## Each data format has its own unique set of configuration options, read
+  ## more about them here:
+  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
+  data_format = "collectd"
+
+  ## Authentication file for cryptographic security levels
+  collectd_auth_file = "/etc/collectd/auth_file"
+  ## One of none (default), sign, or encrypt
+  collectd_security_level = "encrypt"
+  ## Path of to TypesDB specifications
+  collectd_typesdb = ["/usr/share/collectd/types.db"]
+```
+
+# Dropwizard:
+
+The dropwizard format can parse the JSON representation of a single dropwizard metric registry. By default, tags are parsed from metric names as if they were actual influxdb line protocol keys (`measurement<,tag_set>`) which can be overriden by defining custom [measurement & tag templates](./DATA_FORMATS_INPUT.md#measurement--tag-templates). All field value types are supported, `string`, `number` and `boolean`.
+
+A typical JSON of a dropwizard metric registry:
+
+```json
+{
+	"version": "3.0.0",
+	"counters" : {
+		"measurement,tag1=green" : {
+			"count" : 1
+		}
+	},
+	"meters" : {
+		"measurement" : {
+			"count" : 1,
+			"m15_rate" : 1.0,
+			"m1_rate" : 1.0,
+			"m5_rate" : 1.0,
+			"mean_rate" : 1.0,
+			"units" : "events/second"
+		}
+	},
+	"gauges" : {
+		"measurement" : {
+			"value" : 1
+		}
+	},
+	"histograms" : {
+		"measurement" : {
+			"count" : 1,
+			"max" : 1.0,
+			"mean" : 1.0,
+			"min" : 1.0,
+			"p50" : 1.0,
+			"p75" : 1.0,
+			"p95" : 1.0,
+			"p98" : 1.0,
+			"p99" : 1.0,
+			"p999" : 1.0,
+			"stddev" : 1.0
+		}
+	},
+	"timers" : {
+		"measurement" : {
+			"count" : 1,
+			"max" : 1.0,
+			"mean" : 1.0,
+			"min" : 1.0,
+			"p50" : 1.0,
+			"p75" : 1.0,
+			"p95" : 1.0,
+			"p98" : 1.0,
+			"p99" : 1.0,
+			"p999" : 1.0,
+			"stddev" : 1.0,
+			"m15_rate" : 1.0,
+			"m1_rate" : 1.0,
+			"m5_rate" : 1.0,
+			"mean_rate" : 1.0,
+			"duration_units" : "seconds",
+			"rate_units" : "calls/second"
+		}
+	}
+}
+```
+
+Would get translated into 4 different measurements:
+
+```
+measurement,metric_type=counter,tag1=green count=1
+measurement,metric_type=meter count=1,m15_rate=1.0,m1_rate=1.0,m5_rate=1.0,mean_rate=1.0
+measurement,metric_type=gauge value=1
+measurement,metric_type=histogram count=1,max=1.0,mean=1.0,min=1.0,p50=1.0,p75=1.0,p95=1.0,p98=1.0,p99=1.0,p999=1.0
+measurement,metric_type=timer count=1,max=1.0,mean=1.0,min=1.0,p50=1.0,p75=1.0,p95=1.0,p98=1.0,p99=1.0,p999=1.0,stddev=1.0,m15_rate=1.0,m1_rate=1.0,m5_rate=1.0,mean_rate=1.0
+```
+
+You may also parse a dropwizard registry from any JSON document which contains a dropwizard registry in some inner field. 
+Eg. to parse the following JSON document:
+
+```json
+{
+	"time" : "2017-02-22T14:33:03.662+02:00",
+	"tags" : {
+		"tag1" : "green",
+		"tag2" : "yellow"
+	},
+	"metrics" : {
+		"counters" : 	{ 
+			"measurement" : {
+				"count" : 1
+			}
+		},
+		"meters" : {},
+		"gauges" : {},
+		"histograms" : {},
+		"timers" : {}
+	}
+}
+```
+and translate it into:
+
+```
+measurement,metric_type=counter,tag1=green,tag2=yellow count=1 1487766783662000000
+```
+
+you simply need to use the following additional configuration properties:
+
+```toml
+dropwizard_metric_registry_path = "metrics"
+dropwizard_time_path = "time"
+dropwizard_time_format = "2006-01-02T15:04:05Z07:00"
+dropwizard_tags_path = "tags"
+## tag paths per tag are supported too, eg.
+#[inputs.yourinput.dropwizard_tag_paths]
+#  tag1 = "tags.tag1"
+#  tag2 = "tags.tag2"
+```
+
+
+For more information about the dropwizard json format see
+[here](http://metrics.dropwizard.io/3.1.0/manual/json/).
+
+#### Dropwizard Configuration:
+
+```toml
+[[inputs.exec]]
+  ## Commands array
+  commands = ["curl http://localhost:8080/sys/metrics"]
+  timeout = "5s"
+
+  ## Data format to consume.
+  ## Each data format has its own unique set of configuration options, read
+  ## more about them here:
+  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
+  data_format = "dropwizard"
+
+  ## Used by the templating engine to join matched values when cardinality is > 1
+  separator = "_"
+
+  ## Each template line requires a template pattern. It can have an optional
+  ## filter before the template and separated by spaces. It can also have optional extra
+  ## tags following the template. Multiple tags should be separated by commas and no spaces
+  ## similar to the line protocol format. There can be only one default template.
+  ## Templates support below format:
+  ## 1. filter + template
+  ## 2. filter + template + extra tag(s)
+  ## 3. filter + template with field key
+  ## 4. default template
+  ## By providing an empty template array, templating is disabled and measurements are parsed as influxdb line protocol keys (measurement<,tag_set>)
+  templates = []
+
+  ## You may use an appropriate [gjson path](https://github.com/tidwall/gjson#path-syntax) 
+  ## to locate the metric registry within the JSON document
+  # dropwizard_metric_registry_path = "metrics"
+  
+  ## You may use an appropriate [gjson path](https://github.com/tidwall/gjson#path-syntax) 
+  ## to locate the default time of the measurements within the JSON document
+  # dropwizard_time_path = "time"
+  # dropwizard_time_format = "2006-01-02T15:04:05Z07:00"
+  
+  ## You may use an appropriate [gjson path](https://github.com/tidwall/gjson#path-syntax) 
+  ## to locate the tags map within the JSON document
+  # dropwizard_tags_path = "tags"
+
+  ## You may even use tag paths per tag
+  # [inputs.exec.dropwizard_tag_paths]
+  #   tag1 = "tags.tag1"
+  #   tag2 = "tags.tag2"
+
+```

docs/DATA_FORMATS_OUTPUT.md

@@ -2,12 +2,12 @@
 
 Telegraf is able to serialize metrics into the following output data formats:
 
-1. [InfluxDB Line Protocol](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md#influx)
-1. [JSON](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md#json)
-1. [Graphite](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md#graphite)
+1. [InfluxDB Line Protocol](#influx)
+1. [JSON](#json)
+1. [Graphite](#graphite)
 
 Telegraf metrics, like InfluxDB
-[points](https://docs.influxdata.com/influxdb/v0.10/write_protocols/line/),
+[points](https://docs.influxdata.com/influxdb/latest/concepts/glossary/#point),
 are a combination of four basic parts:
 
 1. Measurement Name
@@ -36,7 +36,7 @@ config option, for example, in the `file` output plugin:
   files = ["stdout"]
 
   ## Data format to output.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
   data_format = "influx"
@@ -49,8 +49,10 @@ I'll go over below.
 
 # Influx:
 
-There are no additional configuration options for InfluxDB line-protocol. The
-metrics are serialized directly into InfluxDB line-protocol.
+The `influx` format outputs data as
+[InfluxDB Line Protocol](https://docs.influxdata.com/influxdb/latest/write_protocols/line_protocol_tutorial/).
+This is the recommended format to use unless another format is required for
+interoperability.
 
 ### Influx Configuration:
 
@@ -60,10 +62,24 @@ metrics are serialized directly into InfluxDB line-protocol.
   files = ["stdout", "/tmp/metrics.out"]
 
   ## Data format to output.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
   data_format = "influx"
+
+  ## Maximum line length in bytes.  Useful only for debugging.
+  # influx_max_line_bytes = 0
+
+  ## When true, fields will be output in ascending lexical order.  Enabling
+  ## this option will result in decreased performance and is only recommended
+  ## when you need predictable ordering while debugging.
+  # influx_sort_fields = false
+
+  ## When true, Telegraf will output unsigned integers as unsigned values,
+  ## i.e.: `42u`.  You will need a version of InfluxDB supporting unsigned
+  ## integer values.  Enabling this option will result in field type errors if
+  ## existing data has been written.
+  # influx_uint_support = false
 ```
 
 # Graphite:
@@ -96,6 +112,9 @@ tars.cpu-total.us-east-1.cpu.usage_user 0.89 1455320690
 tars.cpu-total.us-east-1.cpu.usage_idle 98.09 1455320690
 ```
 
+Fields with string values will be skipped.  Boolean fields will be converted
+to 1 (true) or 0 (false).
+
 ### Graphite Configuration:
 
 ```toml
@@ -104,7 +123,7 @@ tars.cpu-total.us-east-1.cpu.usage_idle 98.09 1455320690
   files = ["stdout", "/tmp/metrics.out"]
 
   ## Data format to output.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
   data_format = "graphite"
@@ -143,8 +162,18 @@ The JSON data format serialized Telegraf metrics in json format. The format is:
   files = ["stdout", "/tmp/metrics.out"]
 
   ## Data format to output.
-  ## Each data format has it's own unique set of configuration options, read
+  ## Each data format has its own unique set of configuration options, read
   ## more about them here:
   ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
   data_format = "json"
+  json_timestamp_units = "1ns"
 ```
+
+By default, the timestamp that is output in JSON data format serialized Telegraf
+metrics is in seconds. The precision of this timestamp can be adjusted for any output
+by adding the optional `json_timestamp_units` parameter to the configuration for
+that output. This parameter can be used to set the timestamp units to  nanoseconds (`ns`),
+microseconds (`us` or `µs`), milliseconds (`ms`), or seconds (`s`). Note that this
+parameter will be truncated to the nearest power of 10 that, so if the `json_timestamp_units`
+are set to `15ms` the timestamps for the JSON format serialized Telegraf metrics will be
+output in hundredths of a second (`10ms`).

docs/FAQ.md

@@ -0,0 +1,46 @@
+# Frequently Asked Questions
+
+### Q: How can I monitor the Docker Engine Host from within a container?
+
+You will need to setup several volume mounts as well as some environment
+variables:
+```
+docker run --name telegraf
+	-v /:/hostfs:ro
+	-v /etc:/hostfs/etc:ro
+	-v /proc:/hostfs/proc:ro
+	-v /sys:/hostfs/sys:ro
+	-v /var/run/utmp:/var/run/utmp:ro
+	-e HOST_ETC=/hostfs/etc
+	-e HOST_PROC=/hostfs/proc
+	-e HOST_SYS=/hostfs/sys
+	-e HOST_MOUNT_PREFIX=/hostfs
+	telegraf
+```
+
+
+### Q: Why do I get a "no such host" error resolving hostnames that other
+programs can resolve?
+
+Go uses a pure Go resolver by default for [name resolution](https://golang.org/pkg/net/#hdr-Name_Resolution).
+This resolver behaves differently than the C library functions but is more
+efficient when used with the Go runtime.
+
+If you encounter problems or want to use more advanced name resolution methods
+that are unsupported by the pure Go resolver, you can switch to the cgo
+resolver.
+
+If running manually set:
+```
+export GODEBUG=netdns=cgo
+```
+
+If running as a service add the environment variable to `/etc/default/telegraf`:
+```
+GODEBUG=netdns=cgo
+```
+
+### Q: When will the next version be released?
+
+The latest release date estimate can be viewed on the
+[milestones](https://github.com/influxdata/telegraf/milestones) page.

docs/LICENSE_OF_DEPENDENCIES.md

@@ -1,33 +1,107 @@
-# List
-- github.com/Shopify/sarama [MIT LICENSE](https://github.com/Shopify/sarama/blob/master/MIT-LICENSE)
-- github.com/Sirupsen/logrus [MIT LICENSE](https://github.com/Sirupsen/logrus/blob/master/LICENSE)
-- github.com/armon/go-metrics [MIT LICENSE](https://github.com/armon/go-metrics/blob/master/LICENSE)
-- github.com/boltdb/bolt [MIT LICENSE](https://github.com/boltdb/bolt/blob/master/LICENSE)
-- github.com/cenkalti/backoff [MIT LICENSE](https://github.com/cenkalti/backoff/blob/master/LICENSE)
-- github.com/dancannon/gorethink [APACHE LICENSE](https://github.com/dancannon/gorethink/blob/master/LICENSE)
-- github.com/eapache/go-resiliency [MIT LICENSE](https://github.com/eapache/go-resiliency/blob/master/LICENSE)
-- github.com/eapache/queue [MIT LICENSE](https://github.com/eapache/queue/blob/master/LICENSE)
-- github.com/fsouza/go-dockerclient [BSD LICENSE](https://github.com/fsouza/go-dockerclient/blob/master/LICENSE)
-- github.com/go-sql-driver/mysql [MPL LICENSE](https://github.com/go-sql-driver/mysql/blob/master/LICENSE)
-- github.com/gogo/protobuf [BSD LICENSE](https://github.com/gogo/protobuf/blob/master/LICENSE)
-- github.com/golang/protobuf [BSD LICENSE](https://github.com/golang/protobuf/blob/master/LICENSE)
-- github.com/golang/snappy [BSD LICENSE](https://github.com/golang/snappy/blob/master/LICENSE)
-- github.com/hashicorp/go-msgpack [BSD LICENSE](https://github.com/hashicorp/go-msgpack/blob/master/LICENSE)
-- github.com/hashicorp/raft [MPL LICENSE](https://github.com/hashicorp/raft/blob/master/LICENSE)
-- github.com/hashicorp/raft-boltdb [MPL LICENSE](https://github.com/hashicorp/raft-boltdb/blob/master/LICENSE)
-- github.com/kardianos/service [ZLIB LICENSE](https://github.com/kardianos/service/blob/master/LICENSE) (License not named but matches word for word with ZLib)
-- github.com/kballard/go-shellquote [MIT LICENSE](https://github.com/kballard/go-shellquote/blob/master/LICENSE)
-- github.com/lib/pq [MIT LICENSE](https://github.com/lib/pq/blob/master/LICENSE.md)
-- github.com/matttproud/golang_protobuf_extensions [APACHE LICENSE](https://github.com/matttproud/golang_protobuf_extensions/blob/master/LICENSE)
-- github.com/naoina/go-stringutil [MIT LICENSE](https://github.com/naoina/go-stringutil/blob/master/LICENSE)
-- github.com/naoina/toml [MIT LICENSE](https://github.com/naoina/toml/blob/master/LICENSE)
-- github.com/prometheus/client_golang [APACHE LICENSE](https://github.com/prometheus/client_golang/blob/master/LICENSE)
-- github.com/samuel/go-zookeeper [BSD LICENSE](https://github.com/samuel/go-zookeeper/blob/master/LICENSE)
-- github.com/stretchr/objx [MIT LICENSE](github.com/stretchr/objx)
-- github.com/stretchr/testify [MIT LICENSE](https://github.com/stretchr/testify/blob/master/LICENCE.txt)
-- github.com/wvanbergen/kafka [MIT LICENSE](https://github.com/wvanbergen/kafka/blob/master/LICENSE)
-- github.com/wvanbergen/kazoo-go [MIT LICENSE](https://github.com/wvanbergen/kazoo-go/blob/master/MIT-LICENSE)
-- gopkg.in/dancannon/gorethink.v1 [APACHE LICENSE](https://github.com/dancannon/gorethink/blob/v1.1.2/LICENSE)
-- gopkg.in/mgo.v2 [BSD LICENSE](https://github.com/go-mgo/mgo/blob/v2/LICENSE)
-- golang.org/x/crypto/ [BSD LICENSE](https://github.com/golang/crypto/blob/master/LICENSE)
+# Licenses of dependencies
 
+When distributed in a binary form, Telegraf may contain portions of the
+following works:
+
+- collectd.org [MIT](https://github.com/collectd/go-collectd/blob/master/LICENSE)
+- github.com/aerospike/aerospike-client-go [APACHE](https://github.com/aerospike/aerospike-client-go/blob/master/LICENSE)
+- github.com/amir/raidman [PUBLIC DOMAIN](https://github.com/amir/raidman/blob/master/UNLICENSE)
+- github.com/armon/go-metrics [MIT](https://github.com/armon/go-metrics/blob/master/LICENSE)
+- github.com/aws/aws-sdk-go [APACHE](https://github.com/aws/aws-sdk-go/blob/master/LICENSE.txt)
+- github.com/beorn7/perks [MIT](https://github.com/beorn7/perks/blob/master/LICENSE)
+- github.com/boltdb/bolt [MIT](https://github.com/boltdb/bolt/blob/master/LICENSE)
+- github.com/bsm/sarama-cluster [MIT](https://github.com/bsm/sarama-cluster/blob/master/LICENSE)
+- github.com/cenkalti/backoff [MIT](https://github.com/cenkalti/backoff/blob/master/LICENSE)
+- github.com/chuckpreslar/rcon [MIT](https://github.com/chuckpreslar/rcon#license)
+- github.com/couchbase/go-couchbase [MIT](https://github.com/couchbase/go-couchbase/blob/master/LICENSE)
+- github.com/couchbase/gomemcached [MIT](https://github.com/couchbase/gomemcached/blob/master/LICENSE)
+- github.com/couchbase/goutils [MIT](https://github.com/couchbase/go-couchbase/blob/master/LICENSE)
+- github.com/dancannon/gorethink [APACHE](https://github.com/dancannon/gorethink/blob/master/LICENSE)
+- github.com/davecgh/go-spew [ISC](https://github.com/davecgh/go-spew/blob/master/LICENSE)
+- github.com/docker/docker [APACHE](https://github.com/docker/docker/blob/master/LICENSE)
+- github.com/docker/cli [APACHE](https://github.com/docker/cli/blob/master/LICENSE)
+- github.com/eapache/go-resiliency [MIT](https://github.com/eapache/go-resiliency/blob/master/LICENSE)
+- github.com/eapache/go-xerial-snappy [MIT](https://github.com/eapache/go-xerial-snappy/blob/master/LICENSE)
+- github.com/eapache/queue [MIT](https://github.com/eapache/queue/blob/master/LICENSE)
+- github.com/eclipse/paho.mqtt.golang [ECLIPSE](https://github.com/eclipse/paho.mqtt.golang/blob/master/LICENSE)
+- github.com/fsnotify/fsnotify [BSD](https://github.com/fsnotify/fsnotify/blob/master/LICENSE)
+- github.com/fsouza/go-dockerclient [BSD](https://github.com/fsouza/go-dockerclient/blob/master/LICENSE)
+- github.com/gobwas/glob [MIT](https://github.com/gobwas/glob/blob/master/LICENSE)
+- github.com/google/go-cmp [BSD](https://github.com/google/go-cmp/blob/master/LICENSE)
+- github.com/gogo/protobuf [BSD](https://github.com/gogo/protobuf/blob/master/LICENSE)
+- github.com/golang/protobuf [BSD](https://github.com/golang/protobuf/blob/master/LICENSE)
+- github.com/golang/snappy [BSD](https://github.com/golang/snappy/blob/master/LICENSE)
+- github.com/go-logfmt/logfmt [MIT](https://github.com/go-logfmt/logfmt/blob/master/LICENSE)
+- github.com/gorilla/mux [BSD](https://github.com/gorilla/mux/blob/master/LICENSE)
+- github.com/go-ini/ini [APACHE](https://github.com/go-ini/ini/blob/master/LICENSE)
+- github.com/go-ole/go-ole [MPL](http://mattn.mit-license.org/2013)
+- github.com/go-sql-driver/mysql [MPL](https://github.com/go-sql-driver/mysql/blob/master/LICENSE)
+- github.com/hailocab/go-hostpool [MIT](https://github.com/hailocab/go-hostpool/blob/master/LICENSE)
+- github.com/hashicorp/consul [MPL](https://github.com/hashicorp/consul/blob/master/LICENSE)
+- github.com/hashicorp/go-msgpack [BSD](https://github.com/hashicorp/go-msgpack/blob/master/LICENSE)
+- github.com/hashicorp/raft-boltdb [MPL](https://github.com/hashicorp/raft-boltdb/blob/master/LICENSE)
+- github.com/hashicorp/raft [MPL](https://github.com/hashicorp/raft/blob/master/LICENSE)
+- github.com/influxdata/tail [MIT](https://github.com/influxdata/tail/blob/master/LICENSE.txt)
+- github.com/influxdata/toml [MIT](https://github.com/influxdata/toml/blob/master/LICENSE)
+- github.com/influxdata/wlog [MIT](https://github.com/influxdata/wlog/blob/master/LICENSE)
+- github.com/jackc/pgx [MIT](https://github.com/jackc/pgx/blob/master/LICENSE)
+- github.com/jmespath/go-jmespath [APACHE](https://github.com/jmespath/go-jmespath/blob/master/LICENSE)
+- github.com/kardianos/osext [BSD](https://github.com/kardianos/osext/blob/master/LICENSE)
+- github.com/kardianos/service [ZLIB](https://github.com/kardianos/service/blob/master/LICENSE) (License not named but matches word for word with ZLib)
+- github.com/kballard/go-shellquote [MIT](https://github.com/kballard/go-shellquote/blob/master/LICENSE)
+- github.com/lib/pq [MIT](https://github.com/lib/pq/blob/master/LICENSE.md)
+- github.com/matttproud/golang_protobuf_extensions [APACHE](https://github.com/matttproud/golang_protobuf_extensions/blob/master/LICENSE)
+- github.com/Microsoft/go-winio [MIT](https://github.com/Microsoft/go-winio/blob/master/LICENSE)
+- github.com/miekg/dns [BSD](https://github.com/miekg/dns/blob/master/LICENSE)
+- github.com/naoina/go-stringutil [MIT](https://github.com/naoina/go-stringutil/blob/master/LICENSE)
+- github.com/naoina/toml [MIT](https://github.com/naoina/toml/blob/master/LICENSE)
+- github.com/nats-io/gnatsd [MIT](https://github.com/nats-io/gnatsd/blob/master/LICENSE)
+- github.com/nats-io/go-nats [MIT](https://github.com/nats-io/go-nats/blob/master/LICENSE)
+- github.com/nats-io/nats [MIT](https://github.com/nats-io/nats/blob/master/LICENSE)
+- github.com/nats-io/nuid [MIT](https://github.com/nats-io/nuid/blob/master/LICENSE)
+- github.com/nsqio/go-nsq [MIT](https://github.com/nsqio/go-nsq/blob/master/LICENSE)
+- github.com/opentracing-contrib/go-observer [APACHE](https://github.com/opentracing-contrib/go-observer/blob/master/LICENSE)
+- github.com/opentracing/opentracing-go [MIT](https://github.com/opentracing/opentracing-go/blob/master/LICENSE)
+- github.com/openzipkin/zipkin-go-opentracing [MIT](https://github.com/openzipkin/zipkin-go-opentracing/blob/master/LICENSE)
+- github.com/pierrec/lz4 [BSD](https://github.com/pierrec/lz4/blob/master/LICENSE)
+- github.com/pierrec/xxHash [BSD](https://github.com/pierrec/xxHash/blob/master/LICENSE)
+- github.com/pkg/errors [BSD](https://github.com/pkg/errors/blob/master/LICENSE)
+- github.com/pmezard/go-difflib [BSD](https://github.com/pmezard/go-difflib/blob/master/LICENSE)
+- github.com/prometheus/client_golang [APACHE](https://github.com/prometheus/client_golang/blob/master/LICENSE)
+- github.com/prometheus/client_model [APACHE](https://github.com/prometheus/client_model/blob/master/LICENSE)
+- github.com/prometheus/common [APACHE](https://github.com/prometheus/common/blob/master/LICENSE)
+- github.com/prometheus/procfs [APACHE](https://github.com/prometheus/procfs/blob/master/LICENSE)
+- github.com/rcrowley/go-metrics [BSD](https://github.com/rcrowley/go-metrics/blob/master/LICENSE)
+- github.com/samuel/go-zookeeper [BSD](https://github.com/samuel/go-zookeeper/blob/master/LICENSE)
+- github.com/satori/go.uuid [MIT](https://github.com/satori/go.uuid/blob/master/LICENSE)
+- github.com/shirou/gopsutil [BSD](https://github.com/shirou/gopsutil/blob/master/LICENSE)
+- github.com/shirou/w32 [BSD](https://github.com/shirou/w32/blob/master/LICENSE)
+- github.com/Shopify/sarama [MIT](https://github.com/Shopify/sarama/blob/master/MIT-LICENSE)
+- github.com/Sirupsen/logrus [MIT](https://github.com/Sirupsen/logrus/blob/master/LICENSE)
+- github.com/StackExchange/wmi [MIT](https://github.com/StackExchange/wmi/blob/master/LICENSE)
+- github.com/stretchr/objx [MIT](https://github.com/stretchr/objx/blob/master/LICENSE.md)
+- github.com/soniah/gosnmp [BSD](https://github.com/soniah/gosnmp/blob/master/LICENSE)
+- github.com/streadway/amqp [BSD](https://github.com/streadway/amqp/blob/master/LICENSE)
+- github.com/stretchr/objx [MIT](https://github.com/stretchr/objx/blob/master/LICENSE.md)
+- github.com/stretchr/testify [MIT](https://github.com/stretchr/testify/blob/master/LICENCE.txt)
+- github.com/tidwall/gjson [MIT](https://github.com/tidwall/gjson/blob/master/LICENSE)
+- github.com/tidwall/match [MIT](https://github.com/tidwall/match/blob/master/LICENSE)
+- github.com/mitchellh/mapstructure [MIT](https://github.com/mitchellh/mapstructure/blob/master/LICENSE)
+- github.com/multiplay/go-ts3 [BSD](https://github.com/multiplay/go-ts3/blob/master/LICENSE)
+- github.com/vjeantet/grok [APACHE](https://github.com/vjeantet/grok/blob/master/LICENSE)
+- github.com/wvanbergen/kafka [MIT](https://github.com/wvanbergen/kafka/blob/master/LICENSE)
+- github.com/wvanbergen/kazoo-go [MIT](https://github.com/wvanbergen/kazoo-go/blob/master/MIT-LICENSE)
+- github.com/yuin/gopher-lua [MIT](https://github.com/yuin/gopher-lua/blob/master/LICENSE)
+- github.com/zensqlmonitor/go-mssqldb [BSD](https://github.com/zensqlmonitor/go-mssqldb/blob/master/LICENSE.txt)
+- golang.org/x/crypto [BSD](https://github.com/golang/crypto/blob/master/LICENSE)
+- golang.org/x/net [BSD](https://go.googlesource.com/net/+/master/LICENSE)
+- golang.org/x/text [BSD](https://go.googlesource.com/text/+/master/LICENSE)
+- golang.org/x/sys [BSD](https://go.googlesource.com/sys/+/master/LICENSE)
+- gopkg.in/asn1-ber.v1 [MIT](https://github.com/go-asn1-ber/asn1-ber/blob/v1.2/LICENSE)
+- gopkg.in/dancannon/gorethink.v1 [APACHE](https://github.com/dancannon/gorethink/blob/v1.1.2/LICENSE)
+- gopkg.in/fatih/pool.v2 [MIT](https://github.com/fatih/pool/blob/v2.0.0/LICENSE)
+- gopkg.in/ldap.v2 [MIT](https://github.com/go-ldap/ldap/blob/v2.5.0/LICENSE)
+- gopkg.in/mgo.v2 [BSD](https://github.com/go-mgo/mgo/blob/v2/LICENSE)
+- gopkg.in/olivere/elastic.v5 [MIT](https://github.com/olivere/elastic/blob/v5.0.38/LICENSE)
+- gopkg.in/tomb.v1 [BSD](https://github.com/go-tomb/tomb/blob/v1/LICENSE)
+- gopkg.in/yaml.v2 [APACHE](https://github.com/go-yaml/yaml/blob/v2/LICENSE)

docs/PROFILING.md

@@ -0,0 +1,24 @@
+# Telegraf profiling
+
+Telegraf uses the standard package `net/http/pprof`. This package serves via its HTTP server runtime profiling data in the format expected by the pprof visualization tool.
+
+By default, the profiling is turned off.
+
+To enable profiling you need to specify address to config parameter `pprof-addr`, for example:
+
+```
+telegraf --config telegraf.conf --pprof-addr localhost:6060
+```
+
+There are several paths to get different profiling information:
+
+To look at the heap profile:
+
+`go tool pprof http://localhost:6060/debug/pprof/heap`
+
+or to look at a 30-second CPU profile:
+
+`go tool pprof http://localhost:6060/debug/pprof/profile?seconds=30`
+
+To view all available profiles, open `http://localhost:6060/debug/pprof/` in your browser.
+

docs/WINDOWS_SERVICE.md

@@ -5,7 +5,7 @@ the general steps to set it up.
 
 1. Obtain the telegraf windows distribution
 2. Create the directory `C:\Program Files\Telegraf` (if you install in a different
-   location simply specify the `-config` parameter with the desired location)
+   location simply specify the `--config` parameter with the desired location)
 3. Place the telegraf.exe and the telegraf.conf config file into `C:\Program Files\Telegraf`
 4. To install the service into the Windows Service Manager, run the following in PowerShell as an administrator (If necessary, you can wrap any spaces in the file paths in double quotes ""):
 
@@ -26,6 +26,15 @@ the general steps to set it up.
    > net start telegraf
    ```
 
+## Config Directory
+
+You can also specify a `--config-directory` for the service to use:
+1. Create a directory for config snippets: `C:\Program Files\Telegraf\telegraf.d`
+2. Include the `--config-directory` option when registering the service:
+   ```
+   > C:\"Program Files"\Telegraf\telegraf.exe --service install --config C:\"Program Files"\Telegraf\telegraf.conf --config-directory C:\"Program Files"\Telegraf\telegraf.d
+   ```
+
 ## Other supported operations
 
 Telegraf can manage its own service through the --service flag:
@@ -37,3 +46,8 @@ Telegraf can manage its own service through the --service flag:
 | `telegraf.exe --service start`     | Start the telegraf service    |
 | `telegraf.exe --service stop`      | Stop the telegraf service     |
 
+Troubleshooting  common error #1067
+
+When installing as service in Windows, always double check to specify full path of the config file, otherwise windows service will fail to start
+
+ --config C:\"Program Files"\Telegraf\telegraf.conf

etc/telegraf_windows.conf

@@ -63,8 +63,8 @@
   # The full HTTP or UDP endpoint URL for your InfluxDB instance.
   # Multiple urls can be specified but it is assumed that they are part of the same
   # cluster, this means that only ONE of the urls will be written to each interval.
-  # urls = ["udp://localhost:8089"] # UDP endpoint example
-  urls = ["http://localhost:8086"] # required
+  # urls = ["udp://127.0.0.1:8089"] # UDP endpoint example
+  urls = ["http://127.0.0.1:8086"] # required
   # The target database for metrics (telegraf will create it if not exists)
   database = "telegraf" # required
   # Precision of writes, valid values are "ns", "us" (or "µs"), "ms", "s", "m", "h".
@@ -105,10 +105,11 @@
       "% Privileged Time",
       "% User Time",
       "% Processor Time",
+      "% DPC Time",
     ]
     Measurement = "win_cpu"
     # Set to true to include _Total instance when querying for all (*).
-    #IncludeTotal=false
+    IncludeTotal=true
 
   [[inputs.win_perf_counters.object]]
     # Disk times and queues
@@ -116,21 +117,54 @@
     Instances = ["*"]
     Counters = [
       "% Idle Time",
-      "% Disk Time","% Disk Read Time",
+      "% Disk Time",
+      "% Disk Read Time",
       "% Disk Write Time",
-      "% User Time",
       "Current Disk Queue Length",
+      "% Free Space",
+      "Free Megabytes",
     ]
     Measurement = "win_disk"
     # Set to true to include _Total instance when querying for all (*).
     #IncludeTotal=false
 
+  [[inputs.win_perf_counters.object]]
+    ObjectName = "PhysicalDisk"
+    Instances = ["*"]
+    Counters = [
+      "Disk Read Bytes/sec",
+      "Disk Write Bytes/sec",
+      "Current Disk Queue Length",
+      "Disk Reads/sec",
+      "Disk Writes/sec",
+      "% Disk Time",
+      "% Disk Read Time",
+      "% Disk Write Time",
+    ]
+    Measurement = "win_diskio"
+
+  [[inputs.win_perf_counters.object]]
+    ObjectName = "Network Interface"
+    Instances = ["*"]
+    Counters = [
+      "Bytes Received/sec",
+      "Bytes Sent/sec",
+      "Packets Received/sec",
+      "Packets Sent/sec",
+      "Packets Received Discarded",
+      "Packets Outbound Discarded",
+      "Packets Received Errors",
+      "Packets Outbound Errors",
+    ]
+    Measurement = "win_net"
+
   [[inputs.win_perf_counters.object]]
     ObjectName = "System"
     Counters = [
       "Context Switches/sec",
       "System Calls/sec",
       "Processor Queue Length",
+      "System Up Time",
     ]
     Instances = ["------"]
     Measurement = "win_system"
@@ -150,6 +184,10 @@
       "Transition Faults/sec",
       "Pool Nonpaged Bytes",
       "Pool Paged Bytes",
+      "Standby Cache Reserve Bytes",
+      "Standby Cache Normal Priority Bytes",
+      "Standby Cache Core Bytes",
+
     ]
     # Use 6 x - to remove the Instance bit from the query.
     Instances = ["------"]
@@ -157,6 +195,31 @@
     # Set to true to include _Total instance when querying for all (*).
     #IncludeTotal=false
 
+  [[inputs.win_perf_counters.object]]
+    # Example query where the Instance portion must be removed to get data back,
+    # such as from the Paging File object.
+    ObjectName = "Paging File"
+    Counters = [
+      "% Usage",
+    ]
+    Instances = ["_Total"]
+    Measurement = "win_swap"
+
+  [[inputs.win_perf_counters.object]]
+    ObjectName = "Network Interface"
+    Instances = ["*"]
+    Counters = [
+      "Bytes Sent/sec",
+      "Bytes Received/sec",
+      "Packets Sent/sec",
+      "Packets Received/sec",
+      "Packets Received Discarded",
+      "Packets Received Errors",
+      "Packets Outbound Discarded",
+      "Packets Outbound Errors",
+    ]
+
+
 
 # Windows system plugins using WMI (disabled by default, using
 # win_perf_counters over WMI is recommended)

filter/filter.go

@@ -77,3 +77,40 @@ func compileFilterNoGlob(filters []string) Filter {
 	}
 	return &out
 }
+
+type IncludeExcludeFilter struct {
+	include Filter
+	exclude Filter
+}
+
+func NewIncludeExcludeFilter(
+	include []string,
+	exclude []string,
+) (Filter, error) {
+	in, err := Compile(include)
+	if err != nil {
+		return nil, err
+	}
+
+	ex, err := Compile(exclude)
+	if err != nil {
+		return nil, err
+	}
+
+	return &IncludeExcludeFilter{in, ex}, nil
+}
+
+func (f *IncludeExcludeFilter) Match(s string) bool {
+	if f.include != nil {
+		if !f.include.Match(s) {
+			return false
+		}
+	}
+
+	if f.exclude != nil {
+		if f.exclude.Match(s) {
+			return false
+		}
+	}
+	return true
+}

internal/buffer/buffer.go

@@ -45,9 +45,11 @@ func (b *Buffer) Add(metrics ...telegraf.Metric) {
 		select {
 		case b.buf <- metrics[i]:
 		default:
+			b.mu.Lock()
 			MetricsDropped.Incr(1)
 			<-b.buf
 			b.buf <- metrics[i]
+			b.mu.Unlock()
 		}
 	}
 }

internal/config/config.go

@@ -6,8 +6,10 @@ import (
 	"fmt"
 	"io/ioutil"
 	"log"
+	"math"
 	"os"
 	"path/filepath"
+
 	"regexp"
 	"runtime"
 	"sort"
@@ -25,7 +27,6 @@ import (
 	"github.com/influxdata/telegraf/plugins/processors"
 	"github.com/influxdata/telegraf/plugins/serializers"
 
-	"github.com/influxdata/config"
 	"github.com/influxdata/toml"
 	"github.com/influxdata/toml/ast"
 )
@@ -40,6 +41,11 @@ var (
 
 	// envVarRe is a regex to find environment variables in the config file
 	envVarRe = regexp.MustCompile(`\$\w+`)
+
+	envVarEscaper = strings.NewReplacer(
+		`"`, `\"`,
+		`\`, `\\`,
+	)
 )
 
 // Config specifies the URL/user/password for the database that telegraf
@@ -85,8 +91,8 @@ type AgentConfig struct {
 	//     ie, if Interval=10s then always collect on :00, :10, :20, etc.
 	RoundInterval bool
 
-	// By default, precision will be set to the same timestamp order as the
-	// collection interval, with the maximum being 1s.
+	// By default or when set to "0s", precision will be set to the same
+	// timestamp order as the collection interval, with the maximum being 1s.
 	//   ie, when interval = "10s", precision will be "1s"
 	//       when interval = "250ms", precision will be "1ms"
 	// Precision will NOT be used for service inputs. It is up to each individual
@@ -126,7 +132,7 @@ type AgentConfig struct {
 
 	// TODO(cam): Remove UTC and parameter, they are no longer
 	// valid for the agent config. Leaving them here for now for backwards-
-	// compatability
+	// compatibility
 	UTC bool `toml:"utc"`
 
 	// Debug is the option for running in debug mode
@@ -230,10 +236,13 @@ var header = `# Telegraf Configuration
   ## ie, a jitter of 5s and interval 10s means flushes will happen every 10-15s
   flush_jitter = "0s"
 
-  ## By default, precision will be set to the same timestamp order as the
-  ## collection interval, with the maximum being 1s.
-  ## Precision will NOT be used for service inputs, such as logparser and statsd.
-  ## Valid values are "ns", "us" (or "µs"), "ms", "s".
+  ## By default or when set to "0s", precision will be set to the same
+  ## timestamp order as the collection interval, with the maximum being 1s.
+  ##   ie, when interval = "10s", precision will be "1s"
+  ##       when interval = "250ms", precision will be "1ms"
+  ## Precision will NOT be used for service inputs. It is up to each individual
+  ## service input to set the timestamp at the appropriate precision.
+  ## Valid time units are "ns", "us" (or "µs"), "ms", "s".
   precision = ""
 
   ## Logging configuration:
@@ -506,7 +515,17 @@ func PrintOutputConfig(name string) error {
 
 func (c *Config) LoadDirectory(path string) error {
 	walkfn := func(thispath string, info os.FileInfo, _ error) error {
+		if info == nil {
+			log.Printf("W! Telegraf is not permitted to read %s", thispath)
+			return nil
+		}
+
 		if info.IsDir() {
+			if strings.HasPrefix(info.Name(), "..") {
+				// skip Kubernetes mounts, prevening loading the same config twice
+				return filepath.SkipDir
+			}
+
 			return nil
 		}
 		name := info.Name()
@@ -566,7 +585,7 @@ func (c *Config) LoadConfig(path string) error {
 			if !ok {
 				return fmt.Errorf("%s: invalid configuration", path)
 			}
-			if err = config.UnmarshalTable(subTable, c.Tags); err != nil {
+			if err = toml.UnmarshalTable(subTable, c.Tags); err != nil {
 				log.Printf("E! Could not parse [global_tags] config\n")
 				return fmt.Errorf("Error parsing %s, %s", path, err)
 			}
@@ -579,7 +598,7 @@ func (c *Config) LoadConfig(path string) error {
 		if !ok {
 			return fmt.Errorf("%s: invalid configuration", path)
 		}
-		if err = config.UnmarshalTable(subTable, c.Agent); err != nil {
+		if err = toml.UnmarshalTable(subTable, c.Agent); err != nil {
 			log.Printf("E! Could not parse [agent] config\n")
 			return fmt.Errorf("Error parsing %s, %s", path, err)
 		}
@@ -676,12 +695,17 @@ func (c *Config) LoadConfig(path string) error {
 }
 
 // trimBOM trims the Byte-Order-Marks from the beginning of the file.
-// this is for Windows compatability only.
+// this is for Windows compatibility only.
 // see https://github.com/influxdata/telegraf/issues/1378
 func trimBOM(f []byte) []byte {
 	return bytes.TrimPrefix(f, []byte("\xef\xbb\xbf"))
 }
 
+// escapeEnv escapes a value for inserting into a TOML string.
+func escapeEnv(value string) string {
+	return envVarEscaper.Replace(value)
+}
+
 // parseFile loads a TOML configuration from a provided path and
 // returns the AST produced from the TOML parser. When loading the file, it
 // will find environment variables and replace them.
@@ -695,8 +719,9 @@ func parseFile(fpath string) (*ast.Table, error) {
 
 	env_vars := envVarRe.FindAll(contents, -1)
 	for _, env_var := range env_vars {
-		env_val := os.Getenv(strings.TrimPrefix(string(env_var), "$"))
-		if env_val != "" {
+		env_val, ok := os.LookupEnv(strings.TrimPrefix(string(env_var), "$"))
+		if ok {
+			env_val = escapeEnv(env_val)
 			contents = bytes.Replace(contents, env_var, []byte(env_val), 1)
 		}
 	}
@@ -716,7 +741,7 @@ func (c *Config) addAggregator(name string, table *ast.Table) error {
 		return err
 	}
 
-	if err := config.UnmarshalTable(table, aggregator); err != nil {
+	if err := toml.UnmarshalTable(table, aggregator); err != nil {
 		return err
 	}
 
@@ -736,7 +761,7 @@ func (c *Config) addProcessor(name string, table *ast.Table) error {
 		return err
 	}
 
-	if err := config.UnmarshalTable(table, processor); err != nil {
+	if err := toml.UnmarshalTable(table, processor); err != nil {
 		return err
 	}
 
@@ -776,7 +801,7 @@ func (c *Config) addOutput(name string, table *ast.Table) error {
 		return err
 	}
 
-	if err := config.UnmarshalTable(table, output); err != nil {
+	if err := toml.UnmarshalTable(table, output); err != nil {
 		return err
 	}
 
@@ -817,7 +842,7 @@ func (c *Config) addInput(name string, table *ast.Table) error {
 		return err
 	}
 
-	if err := config.UnmarshalTable(table, input); err != nil {
+	if err := toml.UnmarshalTable(table, input); err != nil {
 		return err
 	}
 
@@ -909,7 +934,7 @@ func buildAggregator(name string, tbl *ast.Table) (*models.AggregatorConfig, err
 	conf.Tags = make(map[string]string)
 	if node, ok := tbl.Fields["tags"]; ok {
 		if subtbl, ok := node.(*ast.Table); ok {
-			if err := config.UnmarshalTable(subtbl, conf.Tags); err != nil {
+			if err := toml.UnmarshalTable(subtbl, conf.Tags); err != nil {
 				log.Printf("Could not parse tags for input %s\n", name)
 			}
 		}
@@ -1146,7 +1171,7 @@ func buildInput(name string, tbl *ast.Table) (*models.InputConfig, error) {
 	cp.Tags = make(map[string]string)
 	if node, ok := tbl.Fields["tags"]; ok {
 		if subtbl, ok := node.(*ast.Table); ok {
-			if err := config.UnmarshalTable(subtbl, cp.Tags); err != nil {
+			if err := toml.UnmarshalTable(subtbl, cp.Tags); err != nil {
 				log.Printf("E! Could not parse tags for input %s\n", name)
 			}
 		}
@@ -1226,6 +1251,75 @@ func buildParser(name string, tbl *ast.Table) (parsers.Parser, error) {
 		}
 	}
 
+	if node, ok := tbl.Fields["collectd_auth_file"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if str, ok := kv.Value.(*ast.String); ok {
+				c.CollectdAuthFile = str.Value
+			}
+		}
+	}
+
+	if node, ok := tbl.Fields["collectd_security_level"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if str, ok := kv.Value.(*ast.String); ok {
+				c.CollectdSecurityLevel = str.Value
+			}
+		}
+	}
+
+	if node, ok := tbl.Fields["collectd_typesdb"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if ary, ok := kv.Value.(*ast.Array); ok {
+				for _, elem := range ary.Value {
+					if str, ok := elem.(*ast.String); ok {
+						c.CollectdTypesDB = append(c.CollectdTypesDB, str.Value)
+					}
+				}
+			}
+		}
+	}
+
+	if node, ok := tbl.Fields["dropwizard_metric_registry_path"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if str, ok := kv.Value.(*ast.String); ok {
+				c.DropwizardMetricRegistryPath = str.Value
+			}
+		}
+	}
+	if node, ok := tbl.Fields["dropwizard_time_path"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if str, ok := kv.Value.(*ast.String); ok {
+				c.DropwizardTimePath = str.Value
+			}
+		}
+	}
+	if node, ok := tbl.Fields["dropwizard_time_format"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if str, ok := kv.Value.(*ast.String); ok {
+				c.DropwizardTimeFormat = str.Value
+			}
+		}
+	}
+	if node, ok := tbl.Fields["dropwizard_tags_path"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if str, ok := kv.Value.(*ast.String); ok {
+				c.DropwizardTagsPath = str.Value
+			}
+		}
+	}
+	c.DropwizardTagPathsMap = make(map[string]string)
+	if node, ok := tbl.Fields["dropwizard_tag_paths"]; ok {
+		if subtbl, ok := node.(*ast.Table); ok {
+			for name, val := range subtbl.Fields {
+				if kv, ok := val.(*ast.KeyValue); ok {
+					if str, ok := kv.Value.(*ast.String); ok {
+						c.DropwizardTagPathsMap[name] = str.Value
+					}
+				}
+			}
+		}
+	}
+
 	c.MetricName = name
 
 	delete(tbl.Fields, "data_format")
@@ -1233,6 +1327,14 @@ func buildParser(name string, tbl *ast.Table) (parsers.Parser, error) {
 	delete(tbl.Fields, "templates")
 	delete(tbl.Fields, "tag_keys")
 	delete(tbl.Fields, "data_type")
+	delete(tbl.Fields, "collectd_auth_file")
+	delete(tbl.Fields, "collectd_security_level")
+	delete(tbl.Fields, "collectd_typesdb")
+	delete(tbl.Fields, "dropwizard_metric_registry_path")
+	delete(tbl.Fields, "dropwizard_time_path")
+	delete(tbl.Fields, "dropwizard_time_format")
+	delete(tbl.Fields, "dropwizard_tags_path")
+	delete(tbl.Fields, "dropwizard_tag_paths")
 
 	return parsers.NewParser(c)
 }
@@ -1241,7 +1343,7 @@ func buildParser(name string, tbl *ast.Table) (parsers.Parser, error) {
 // a serializers.Serializer object, and creates it, which can then be added onto
 // an Output object.
 func buildSerializer(name string, tbl *ast.Table) (serializers.Serializer, error) {
-	c := &serializers.Config{}
+	c := &serializers.Config{TimestampUnits: time.Duration(1 * time.Second)}
 
 	if node, ok := tbl.Fields["data_format"]; ok {
 		if kv, ok := node.(*ast.KeyValue); ok {
@@ -1271,9 +1373,65 @@ func buildSerializer(name string, tbl *ast.Table) (serializers.Serializer, error
 		}
 	}
 
+	if node, ok := tbl.Fields["influx_max_line_bytes"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if integer, ok := kv.Value.(*ast.Integer); ok {
+				v, err := integer.Int()
+				if err != nil {
+					return nil, err
+				}
+				c.InfluxMaxLineBytes = int(v)
+			}
+		}
+	}
+
+	if node, ok := tbl.Fields["influx_sort_fields"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if b, ok := kv.Value.(*ast.Boolean); ok {
+				var err error
+				c.InfluxSortFields, err = b.Boolean()
+				if err != nil {
+					return nil, err
+				}
+			}
+		}
+	}
+
+	if node, ok := tbl.Fields["influx_uint_support"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if b, ok := kv.Value.(*ast.Boolean); ok {
+				var err error
+				c.InfluxUintSupport, err = b.Boolean()
+				if err != nil {
+					return nil, err
+				}
+			}
+		}
+	}
+
+	if node, ok := tbl.Fields["json_timestamp_units"]; ok {
+		if kv, ok := node.(*ast.KeyValue); ok {
+			if str, ok := kv.Value.(*ast.String); ok {
+				timestampVal, err := time.ParseDuration(str.Value)
+				if err != nil {
+					return nil, fmt.Errorf("Unable to parse json_timestamp_units as a duration, %s", err)
+				}
+				// now that we have a duration, truncate it to the nearest
+				// power of ten (just in case)
+				nearest_exponent := int64(math.Log10(float64(timestampVal.Nanoseconds())))
+				new_nanoseconds := int64(math.Pow(10.0, float64(nearest_exponent)))
+				c.TimestampUnits = time.Duration(new_nanoseconds)
+			}
+		}
+	}
+
+	delete(tbl.Fields, "influx_max_line_bytes")
+	delete(tbl.Fields, "influx_sort_fields")
+	delete(tbl.Fields, "influx_uint_support")
 	delete(tbl.Fields, "data_format")
 	delete(tbl.Fields, "prefix")
 	delete(tbl.Fields, "template")
+	delete(tbl.Fields, "json_timestamp_units")
 	return serializers.NewSerializer(c)
 }
 

internal/config/testdata/subconfig/..4984_10_04_08_28_06.119/invalid-config.conf

@@ -0,0 +1,4 @@
+# This invalid config file should be skipped during testing
+# as it is an ..data folder
+
+[[outputs.influxdb

internal/config/testdata/telegraf-agent.toml

@@ -60,7 +60,7 @@
   # Kafka topic for producer messages
   topic = "telegraf"
   # Telegraf tag to use as a routing key
-  #  ie, if this tag exists, it's value will be used as the routing key
+  #  ie, if this tag exists, its value will be used as the routing key
   routing_tag = "host"
 
 
@@ -143,19 +143,31 @@
 [[inputs.diskio]]
   # no configuration
 
-# read metrics from a Kafka topic
+# read metrics from a Kafka 0.9+ topic
 [[inputs.kafka_consumer]]
-  # topic(s) to consume
+  ## kafka brokers
+  brokers = ["localhost:9092"]
+  ## topic(s) to consume
+  topics = ["telegraf"]
+  ## the name of the consumer group
+  consumer_group = "telegraf_metrics_consumers"
+  ## Offset (must be either "oldest" or "newest")
+  offset = "oldest"
+
+# read metrics from a Kafka legacy topic
+[[inputs.kafka_consumer_legacy]]
+  ## topic(s) to consume
   topics = ["telegraf"]
   # an array of Zookeeper connection strings
   zookeeper_peers = ["localhost:2181"]
-  # the name of the consumer group
+  ## the name of the consumer group
   consumer_group = "telegraf_metrics_consumers"
   # Maximum number of points to buffer between collection intervals
   point_buffer = 100000
-  # Offset (must be either "oldest" or "newest")
+  ## Offset (must be either "oldest" or "newest")
   offset = "oldest"
 
+
 # Read metrics from a LeoFS Server via SNMP
 [[inputs.leofs]]
   # An array of URI to gather stats about LeoFS.

internal/errchan/errchan.go

@@ -1,37 +0,0 @@
-package errchan
-
-import (
-	"fmt"
-	"strings"
-)
-
-type ErrChan struct {
-	C chan error
-}
-
-// New returns an error channel of max length 'n'
-// errors can be sent to the ErrChan.C channel, and will be returned when
-// ErrChan.Error() is called.
-func New(n int) *ErrChan {
-	return &ErrChan{
-		C: make(chan error, n),
-	}
-}
-
-// Error closes the ErrChan.C channel and returns an error if there are any
-// non-nil errors, otherwise returns nil.
-func (e *ErrChan) Error() error {
-	close(e.C)
-
-	var out string
-	for err := range e.C {
-		if err != nil {
-			out += "[" + err.Error() + "], "
-		}
-	}
-
-	if out != "" {
-		return fmt.Errorf("Errors encountered: " + strings.TrimRight(out, ", "))
-	}
-	return nil
-}

internal/globpath/globpath.go

@@ -45,7 +45,7 @@ func (g *GlobPath) Match() map[string]os.FileInfo {
 	if !g.hasMeta {
 		out := make(map[string]os.FileInfo)
 		info, err := os.Stat(g.path)
-		if !os.IsNotExist(err) {
+		if err == nil {
 			out[g.path] = info
 		}
 		return out
@@ -55,7 +55,7 @@ func (g *GlobPath) Match() map[string]os.FileInfo {
 		files, _ := filepath.Glob(g.path)
 		for _, file := range files {
 			info, err := os.Stat(file)
-			if !os.IsNotExist(err) {
+			if err == nil {
 				out[file] = info
 			}
 		}

internal/globpath/globpath_test.go

@@ -1,6 +1,7 @@
 package globpath
 
 import (
+	"os"
 	"runtime"
 	"strings"
 	"testing"
@@ -28,7 +29,7 @@ func TestCompileAndMatch(t *testing.T) {
 	require.NoError(t, err)
 
 	matches := g1.Match()
-	assert.Len(t, matches, 3)
+	assert.Len(t, matches, 6)
 	matches = g2.Match()
 	assert.Len(t, matches, 2)
 	matches = g3.Match()
@@ -56,7 +57,34 @@ func TestFindRootDir(t *testing.T) {
 	}
 }
 
+func TestFindNestedTextFile(t *testing.T) {
+	dir := getTestdataDir()
+	// test super asterisk
+	g1, err := Compile(dir + "/**.txt")
+	require.NoError(t, err)
+
+	matches := g1.Match()
+	assert.Len(t, matches, 1)
+}
+
 func getTestdataDir() string {
 	_, filename, _, _ := runtime.Caller(1)
 	return strings.Replace(filename, "globpath_test.go", "testdata", 1)
 }
+
+func TestMatch_ErrPermission(t *testing.T) {
+	tests := []struct {
+		input    string
+		expected map[string]os.FileInfo
+	}{
+		{"/root/foo", map[string]os.FileInfo{}},
+		{"/root/f*", map[string]os.FileInfo{}},
+	}
+
+	for _, test := range tests {
+		glob, err := Compile(test.input)
+		require.NoError(t, err)
+		actual := glob.Match()
+		require.Equal(t, test.expected, actual)
+	}
+}

internal/internal.go

@@ -112,9 +112,10 @@ func RandomString(n int) string {
 	return string(bytes)
 }
 
-// GetTLSConfig gets a tls.Config object from the given certs, key, and CA files.
-// you must give the full path to the files.
-// If all files are blank and InsecureSkipVerify=false, returns a nil pointer.
+// GetTLSConfig gets a tls.Config object from the given certs, key, and CA files
+// for use with a client.
+// The full path to each file must be provided.
+// Returns a nil pointer if all files are blank and InsecureSkipVerify=false.
 func GetTLSConfig(
 	SSLCert, SSLKey, SSLCA string,
 	InsecureSkipVerify bool,
@@ -155,6 +156,50 @@ func GetTLSConfig(
 	return t, nil
 }
 
+// GetServerTLSConfig gets a tls.Config object from the given certs, key, and one or more CA files
+// for use with a server.
+// The full path to each file must be provided.
+// Returns a nil pointer if all files are blank.
+func GetServerTLSConfig(
+	TLSCert, TLSKey string,
+	TLSAllowedCACerts []string,
+) (*tls.Config, error) {
+	if TLSCert == "" && TLSKey == "" && len(TLSAllowedCACerts) == 0 {
+		return nil, nil
+	}
+
+	t := &tls.Config{}
+
+	if len(TLSAllowedCACerts) != 0 {
+		caCertPool := x509.NewCertPool()
+		for _, cert := range TLSAllowedCACerts {
+			c, err := ioutil.ReadFile(cert)
+			if err != nil {
+				return nil, errors.New(fmt.Sprintf("Could not load TLS CA: %s",
+					err))
+			}
+			caCertPool.AppendCertsFromPEM(c)
+		}
+		t.ClientCAs = caCertPool
+		t.ClientAuth = tls.RequireAndVerifyClientCert
+	}
+
+	if TLSCert != "" && TLSKey != "" {
+		cert, err := tls.LoadX509KeyPair(TLSCert, TLSKey)
+		if err != nil {
+			return nil, errors.New(fmt.Sprintf(
+				"Could not load TLS client key/certificate from %s:%s: %s",
+				TLSKey, TLSCert, err))
+		}
+
+		t.Certificates = []tls.Certificate{cert}
+	}
+
+	t.BuildNameToCertificate()
+
+	return t, nil
+}
+
 // SnakeCase converts the given string to snake case following the Golang format:
 // acronyms are converted to lower-case and preceded by an underscore.
 func SnakeCase(in string) string {

internal/internal_test.go

@@ -40,9 +40,13 @@ func TestSnakeCase(t *testing.T) {
 var (
 	sleepbin, _ = exec.LookPath("sleep")
 	echobin, _  = exec.LookPath("echo")
+	shell, _    = exec.LookPath("sh")
 )
 
 func TestRunTimeout(t *testing.T) {
+	if testing.Short() {
+		t.Skip("Skipping test due to random failures.")
+	}
 	if sleepbin == "" {
 		t.Skip("'sleep' binary not available on OS, skipping.")
 	}
@@ -57,6 +61,8 @@ func TestRunTimeout(t *testing.T) {
 }
 
 func TestCombinedOutputTimeout(t *testing.T) {
+	// TODO: Fix this test
+	t.Skip("Test failing too often, skip for now and revisit later.")
 	if sleepbin == "" {
 		t.Skip("'sleep' binary not available on OS, skipping.")
 	}
@@ -84,13 +90,13 @@ func TestCombinedOutput(t *testing.T) {
 // test that CombinedOutputTimeout and exec.Cmd.CombinedOutput return
 // the same output from a failed command.
 func TestCombinedOutputError(t *testing.T) {
-	if sleepbin == "" {
-		t.Skip("'sleep' binary not available on OS, skipping.")
+	if shell == "" {
+		t.Skip("'sh' binary not available on OS, skipping.")
 	}
-	cmd := exec.Command(sleepbin, "foo")
+	cmd := exec.Command(shell, "-c", "false")
 	expected, err := cmd.CombinedOutput()
 
-	cmd2 := exec.Command(sleepbin, "foo")
+	cmd2 := exec.Command(shell, "-c", "false")
 	actual, err := CombinedOutputTimeout(cmd2, time.Second)
 
 	assert.Error(t, err)
@@ -98,16 +104,18 @@ func TestCombinedOutputError(t *testing.T) {
 }
 
 func TestRunError(t *testing.T) {
-	if sleepbin == "" {
-		t.Skip("'sleep' binary not available on OS, skipping.")
+	if shell == "" {
+		t.Skip("'sh' binary not available on OS, skipping.")
 	}
-	cmd := exec.Command(sleepbin, "foo")
+	cmd := exec.Command(shell, "-c", "false")
 	err := RunTimeout(cmd, time.Second)
 
 	assert.Error(t, err)
 }
 
 func TestRandomSleep(t *testing.T) {
+	// TODO: Fix this test
+	t.Skip("Test failing too often, skip for now and revisit later.")
 	// test that zero max returns immediately
 	s := time.Now()
 	RandomSleep(time.Duration(0), make(chan struct{}))

internal/limiter/limiter_test.go

@@ -1,54 +0,0 @@
-package limiter
-
-import (
-	"testing"
-	"time"
-
-	"github.com/stretchr/testify/assert"
-)
-
-func TestRateLimiter(t *testing.T) {
-	r := NewRateLimiter(5, time.Second)
-	ticker := time.NewTicker(time.Millisecond * 75)
-
-	// test that we can only get 5 receives from the rate limiter
-	counter := 0
-outer:
-	for {
-		select {
-		case <-r.C:
-			counter++
-		case <-ticker.C:
-			break outer
-		}
-	}
-
-	assert.Equal(t, 5, counter)
-	r.Stop()
-	// verify that the Stop function closes the channel.
-	_, ok := <-r.C
-	assert.False(t, ok)
-}
-
-func TestRateLimiterMultipleIterations(t *testing.T) {
-	r := NewRateLimiter(5, time.Millisecond*50)
-	ticker := time.NewTicker(time.Millisecond * 250)
-
-	// test that we can get 15 receives from the rate limiter
-	counter := 0
-outer:
-	for {
-		select {
-		case <-ticker.C:
-			break outer
-		case <-r.C:
-			counter++
-		}
-	}
-
-	assert.True(t, counter > 10)
-	r.Stop()
-	// verify that the Stop function closes the channel.
-	_, ok := <-r.C
-	assert.False(t, ok)
-}

internal/models/filter.go

@@ -132,6 +132,7 @@ func (f *Filter) Apply(
 	return true
 }
 
+// IsActive checking if filter is active
 func (f *Filter) IsActive() bool {
 	return f.isActive
 }
@@ -139,43 +140,66 @@ func (f *Filter) IsActive() bool {
 // shouldNamePass returns true if the metric should pass, false if should drop
 // based on the drop/pass filter parameters
 func (f *Filter) shouldNamePass(key string) bool {
-	if f.namePass != nil {
+
+	pass := func(f *Filter) bool {
 		if f.namePass.Match(key) {
 			return true
 		}
 		return false
 	}
 
-	if f.nameDrop != nil {
+	drop := func(f *Filter) bool {
 		if f.nameDrop.Match(key) {
 			return false
 		}
+		return true
 	}
+
+	if f.namePass != nil && f.nameDrop != nil {
+		return pass(f) && drop(f)
+	} else if f.namePass != nil {
+		return pass(f)
+	} else if f.nameDrop != nil {
+		return drop(f)
+	}
+
 	return true
 }
 
 // shouldFieldPass returns true if the metric should pass, false if should drop
 // based on the drop/pass filter parameters
 func (f *Filter) shouldFieldPass(key string) bool {
-	if f.fieldPass != nil {
+
+	pass := func(f *Filter) bool {
 		if f.fieldPass.Match(key) {
 			return true
 		}
 		return false
 	}
 
-	if f.fieldDrop != nil {
+	drop := func(f *Filter) bool {
 		if f.fieldDrop.Match(key) {
 			return false
 		}
+		return true
 	}
+
+	if f.fieldPass != nil && f.fieldDrop != nil {
+		return pass(f) && drop(f)
+	} else if f.fieldPass != nil {
+		return pass(f)
+	} else if f.fieldDrop != nil {
+		return drop(f)
+	}
+
 	return true
 }
 
 // shouldTagsPass returns true if the metric should pass, false if should drop
 // based on the tagdrop/tagpass filter parameters
 func (f *Filter) shouldTagsPass(tags map[string]string) bool {
-	if f.TagPass != nil {
+
+	pass := func(f *Filter) bool {
 		for _, pat := range f.TagPass {
 			if pat.filter == nil {
 				continue
@@ -189,7 +213,7 @@ func (f *Filter) shouldTagsPass(tags map[string]string) bool {
 		return false
 	}
 
-	if f.TagDrop != nil {
+	drop := func(f *Filter) bool {
 		for _, pat := range f.TagDrop {
 			if pat.filter == nil {
 				continue
@@ -203,6 +227,18 @@ func (f *Filter) shouldTagsPass(tags map[string]string) bool {
 		return true
 	}
 
+	// Add additional logic in case where both parameters are set.
+	// see: https://github.com/influxdata/telegraf/issues/2860
+	if f.TagPass != nil && f.TagDrop != nil {
+		// return true only in case when tag pass and won't be dropped (true, true).
+		// in case when the same tag should be passed and dropped it will be dropped (true, false).
+		return pass(f) && drop(f)
+	} else if f.TagPass != nil {
+		return pass(f)
+	} else if f.TagDrop != nil {
+		return drop(f)
+	}
+
 	return true
 }
 

internal/models/filter_test.go

@@ -357,3 +357,88 @@ func TestFilter_FilterTagsMatches(t *testing.T) {
 		"mytag": "foobar",
 	}, pretags)
 }
+
+// TestFilter_FilterNamePassAndDrop used for check case when
+// both parameters were defined
+// see: https://github.com/influxdata/telegraf/issues/2860
+func TestFilter_FilterNamePassAndDrop(t *testing.T) {
+
+	inputData := []string{"name1", "name2", "name3", "name4"}
+	expectedResult := []bool{false, true, false, false}
+
+	f := Filter{
+		NamePass: []string{"name1", "name2"},
+		NameDrop: []string{"name1", "name3"},
+	}
+
+	require.NoError(t, f.Compile())
+
+	for i, name := range inputData {
+		assert.Equal(t, f.shouldNamePass(name), expectedResult[i])
+	}
+}
+
+// TestFilter_FilterFieldPassAndDrop used for check case when
+// both parameters were defined
+// see: https://github.com/influxdata/telegraf/issues/2860
+func TestFilter_FilterFieldPassAndDrop(t *testing.T) {
+
+	inputData := []string{"field1", "field2", "field3", "field4"}
+	expectedResult := []bool{false, true, false, false}
+
+	f := Filter{
+		FieldPass: []string{"field1", "field2"},
+		FieldDrop: []string{"field1", "field3"},
+	}
+
+	require.NoError(t, f.Compile())
+
+	for i, field := range inputData {
+		assert.Equal(t, f.shouldFieldPass(field), expectedResult[i])
+	}
+}
+
+// TestFilter_FilterTagsPassAndDrop used for check case when
+// both parameters were defined
+// see: https://github.com/influxdata/telegraf/issues/2860
+func TestFilter_FilterTagsPassAndDrop(t *testing.T) {
+
+	inputData := []map[string]string{
+		{"tag1": "1", "tag2": "3"},
+		{"tag1": "1", "tag2": "2"},
+		{"tag1": "2", "tag2": "1"},
+		{"tag1": "4", "tag2": "1"},
+	}
+
+	expectedResult := []bool{false, true, false, false}
+
+	filterPass := []TagFilter{
+		TagFilter{
+			Name:   "tag1",
+			Filter: []string{"1", "4"},
+		},
+	}
+
+	filterDrop := []TagFilter{
+		TagFilter{
+			Name:   "tag1",
+			Filter: []string{"4"},
+		},
+		TagFilter{
+			Name:   "tag2",
+			Filter: []string{"3"},
+		},
+	}
+
+	f := Filter{
+		TagDrop: filterDrop,
+		TagPass: filterPass,
+	}
+
+	require.NoError(t, f.Compile())
+
+	for i, tag := range inputData {
+		assert.Equal(t, f.shouldTagsPass(tag), expectedResult[i])
+	}
+
+}

internal/models/makemetric.go

@@ -2,7 +2,6 @@ package models
 
 import (
 	"log"
-	"math"
 	"time"
 
 	"github.com/influxdata/telegraf"
@@ -77,62 +76,6 @@ func makemetric(
 		}
 	}
 
-	for k, v := range fields {
-		// Validate uint64 and float64 fields
-		// convert all int & uint types to int64
-		switch val := v.(type) {
-		case nil:
-			// delete nil fields
-			delete(fields, k)
-		case uint:
-			fields[k] = int64(val)
-			continue
-		case uint8:
-			fields[k] = int64(val)
-			continue
-		case uint16:
-			fields[k] = int64(val)
-			continue
-		case uint32:
-			fields[k] = int64(val)
-			continue
-		case int:
-			fields[k] = int64(val)
-			continue
-		case int8:
-			fields[k] = int64(val)
-			continue
-		case int16:
-			fields[k] = int64(val)
-			continue
-		case int32:
-			fields[k] = int64(val)
-			continue
-		case uint64:
-			// InfluxDB does not support writing uint64
-			if val < uint64(9223372036854775808) {
-				fields[k] = int64(val)
-			} else {
-				fields[k] = int64(9223372036854775807)
-			}
-			continue
-		case float32:
-			fields[k] = float64(val)
-			continue
-		case float64:
-			// NaNs are invalid values in influxdb, skip measurement
-			if math.IsNaN(val) || math.IsInf(val, 0) {
-				log.Printf("D! Measurement [%s] field [%s] has a NaN or Inf "+
-					"field, skipping",
-					measurement, k)
-				delete(fields, k)
-				continue
-			}
-		default:
-			fields[k] = v
-		}
-	}
-
 	m, err := metric.New(measurement, tags, fields, t, mType)
 	if err != nil {
 		log.Printf("Error adding point [%s]: %s\n", measurement, err.Error())

internal/models/running_aggregator_test.go

@@ -1,7 +1,6 @@
 package models
 
 import (
-	"fmt"
 	"sync"
 	"sync/atomic"
 	"testing"
@@ -167,69 +166,6 @@ func TestAddDropOriginal(t *testing.T) {
 	assert.False(t, ra.Add(m2))
 }
 
-// make an untyped, counter, & gauge metric
-func TestMakeMetricA(t *testing.T) {
-	now := time.Now()
-	ra := NewRunningAggregator(&TestAggregator{}, &AggregatorConfig{
-		Name: "TestRunningAggregator",
-	})
-	assert.Equal(t, "aggregators.TestRunningAggregator", ra.Name())
-
-	m := ra.MakeMetric(
-		"RITest",
-		map[string]interface{}{"value": int(101)},
-		map[string]string{},
-		telegraf.Untyped,
-		now,
-	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-	assert.Equal(
-		t,
-		m.Type(),
-		telegraf.Untyped,
-	)
-
-	m = ra.MakeMetric(
-		"RITest",
-		map[string]interface{}{"value": int(101)},
-		map[string]string{},
-		telegraf.Counter,
-		now,
-	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-	assert.Equal(
-		t,
-		m.Type(),
-		telegraf.Counter,
-	)
-
-	m = ra.MakeMetric(
-		"RITest",
-		map[string]interface{}{"value": int(101)},
-		map[string]string{},
-		telegraf.Gauge,
-		now,
-	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-	assert.Equal(
-		t,
-		m.Type(),
-		telegraf.Gauge,
-	)
-}
-
 type TestAggregator struct {
 	sum int64
 }

internal/models/running_input.go

@@ -5,6 +5,7 @@ import (
 	"time"
 
 	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/serializers/influx"
 	"github.com/influxdata/telegraf/selfstat"
 )
 
@@ -75,7 +76,12 @@ func (r *RunningInput) MakeMetric(
 	)
 
 	if r.trace && m != nil {
-		fmt.Print("> " + m.String())
+		s := influx.NewSerializer()
+		s.SetFieldSortOrder(influx.SortFields)
+		octets, err := s.Serialize(m)
+		if err == nil {
+			fmt.Print("> " + string(octets))
+		}
 	}
 
 	r.MetricsGathered.Incr(1)

internal/models/running_input_test.go

@@ -1,14 +1,14 @@
 package models
 
 import (
-	"fmt"
-	"math"
 	"testing"
 	"time"
 
 	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/metric"
 
 	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 )
 
 func TestMakeMetricNoFields(t *testing.T) {
@@ -44,77 +44,17 @@ func TestMakeMetricNilFields(t *testing.T) {
 		telegraf.Untyped,
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-}
 
-// make an untyped, counter, & gauge metric
-func TestMakeMetric(t *testing.T) {
-	now := time.Now()
-	ri := NewRunningInput(&testInput{}, &InputConfig{
-		Name: "TestRunningInput",
-	})
-
-	ri.SetTrace(true)
-	assert.Equal(t, true, ri.Trace())
-	assert.Equal(t, "inputs.TestRunningInput", ri.Name())
-
-	m := ri.MakeMetric(
-		"RITest",
-		map[string]interface{}{"value": int(101)},
+	expected, err := metric.New("RITest",
 		map[string]string{},
-		telegraf.Untyped,
+		map[string]interface{}{
+			"value": int(101),
+		},
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-	assert.Equal(
-		t,
-		m.Type(),
-		telegraf.Untyped,
-	)
+	require.NoError(t, err)
 
-	m = ri.MakeMetric(
-		"RITest",
-		map[string]interface{}{"value": int(101)},
-		map[string]string{},
-		telegraf.Counter,
-		now,
-	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-	assert.Equal(
-		t,
-		m.Type(),
-		telegraf.Counter,
-	)
-
-	m = ri.MakeMetric(
-		"RITest",
-		map[string]interface{}{"value": int(101)},
-		map[string]string{},
-		telegraf.Gauge,
-		now,
-	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-	assert.Equal(
-		t,
-		m.Type(),
-		telegraf.Gauge,
-	)
+	require.Equal(t, expected, m)
 }
 
 func TestMakeMetricWithPluginTags(t *testing.T) {
@@ -136,11 +76,18 @@ func TestMakeMetricWithPluginTags(t *testing.T) {
 		telegraf.Untyped,
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest,foo=bar value=101i %d\n", now.UnixNano()),
-		m.String(),
+
+	expected, err := metric.New("RITest",
+		map[string]string{
+			"foo": "bar",
+		},
+		map[string]interface{}{
+			"value": 101,
+		},
+		now,
 	)
+	require.NoError(t, err)
+	require.Equal(t, expected, m)
 }
 
 func TestMakeMetricFilteredOut(t *testing.T) {
@@ -186,87 +133,17 @@ func TestMakeMetricWithDaemonTags(t *testing.T) {
 		telegraf.Untyped,
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest,foo=bar value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-}
-
-// make an untyped, counter, & gauge metric
-func TestMakeMetricInfFields(t *testing.T) {
-	inf := math.Inf(1)
-	ninf := math.Inf(-1)
-	now := time.Now()
-	ri := NewRunningInput(&testInput{}, &InputConfig{
-		Name: "TestRunningInput",
-	})
-
-	ri.SetTrace(true)
-	assert.Equal(t, true, ri.Trace())
-
-	m := ri.MakeMetric(
-		"RITest",
-		map[string]interface{}{
-			"value": int(101),
-			"inf":   inf,
-			"ninf":  ninf,
+	expected, err := metric.New("RITest",
+		map[string]string{
+			"foo": "bar",
+		},
+		map[string]interface{}{
+			"value": 101,
 		},
-		map[string]string{},
-		telegraf.Untyped,
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
-	)
-}
-
-func TestMakeMetricAllFieldTypes(t *testing.T) {
-	now := time.Now()
-	ri := NewRunningInput(&testInput{}, &InputConfig{
-		Name: "TestRunningInput",
-	})
-
-	ri.SetTrace(true)
-	assert.Equal(t, true, ri.Trace())
-
-	m := ri.MakeMetric(
-		"RITest",
-		map[string]interface{}{
-			"a": int(10),
-			"b": int8(10),
-			"c": int16(10),
-			"d": int32(10),
-			"e": uint(10),
-			"f": uint8(10),
-			"g": uint16(10),
-			"h": uint32(10),
-			"i": uint64(10),
-			"j": float32(10),
-			"k": uint64(9223372036854775810),
-			"l": "foobar",
-			"m": true,
-		},
-		map[string]string{},
-		telegraf.Untyped,
-		now,
-	)
-	assert.Contains(t, m.String(), "a=10i")
-	assert.Contains(t, m.String(), "b=10i")
-	assert.Contains(t, m.String(), "c=10i")
-	assert.Contains(t, m.String(), "d=10i")
-	assert.Contains(t, m.String(), "e=10i")
-	assert.Contains(t, m.String(), "f=10i")
-	assert.Contains(t, m.String(), "g=10i")
-	assert.Contains(t, m.String(), "h=10i")
-	assert.Contains(t, m.String(), "i=10i")
-	assert.Contains(t, m.String(), "j=10")
-	assert.NotContains(t, m.String(), "j=10i")
-	assert.Contains(t, m.String(), "k=9223372036854775807i")
-	assert.Contains(t, m.String(), "l=\"foobar\"")
-	assert.Contains(t, m.String(), "m=true")
+	require.NoError(t, err)
+	require.Equal(t, expected, m)
 }
 
 func TestMakeMetricNameOverride(t *testing.T) {
@@ -283,11 +160,15 @@ func TestMakeMetricNameOverride(t *testing.T) {
 		telegraf.Untyped,
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("foobar value=101i %d\n", now.UnixNano()),
-		m.String(),
+	expected, err := metric.New("foobar",
+		nil,
+		map[string]interface{}{
+			"value": 101,
+		},
+		now,
 	)
+	require.NoError(t, err)
+	require.Equal(t, expected, m)
 }
 
 func TestMakeMetricNamePrefix(t *testing.T) {
@@ -304,11 +185,15 @@ func TestMakeMetricNamePrefix(t *testing.T) {
 		telegraf.Untyped,
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("foobar_RITest value=101i %d\n", now.UnixNano()),
-		m.String(),
+	expected, err := metric.New("foobar_RITest",
+		nil,
+		map[string]interface{}{
+			"value": 101,
+		},
+		now,
 	)
+	require.NoError(t, err)
+	require.Equal(t, expected, m)
 }
 
 func TestMakeMetricNameSuffix(t *testing.T) {
@@ -325,11 +210,15 @@ func TestMakeMetricNameSuffix(t *testing.T) {
 		telegraf.Untyped,
 		now,
 	)
-	assert.Equal(
-		t,
-		fmt.Sprintf("RITest_foobar value=101i %d\n", now.UnixNano()),
-		m.String(),
+	expected, err := metric.New("RITest_foobar",
+		nil,
+		map[string]interface{}{
+			"value": 101,
+		},
+		now,
 	)
+	require.NoError(t, err)
+	require.Equal(t, expected, m)
 }
 
 type testInput struct{}

internal/models/running_output.go

@@ -2,6 +2,7 @@ package models
 
 import (
 	"log"
+	"sync"
 	"time"
 
 	"github.com/influxdata/telegraf"
@@ -34,6 +35,9 @@ type RunningOutput struct {
 
 	metrics     *buffer.Buffer
 	failMetrics *buffer.Buffer
+
+	// Guards against concurrent calls to the Output as described in #3009
+	sync.Mutex
 }
 
 func NewRunningOutput(
@@ -83,13 +87,16 @@ func NewRunningOutput(
 			map[string]string{"output": name},
 		),
 	}
-	ro.BufferLimit.Incr(int64(ro.MetricBufferLimit))
+	ro.BufferLimit.Set(int64(ro.MetricBufferLimit))
 	return ro
 }
 
 // AddMetric adds a metric to the output. This function can also write cached
 // points if FlushBufferWhenFull is true.
 func (ro *RunningOutput) AddMetric(m telegraf.Metric) {
+	if m == nil {
+		return
+	}
 	// Filter any tagexclude/taginclude parameters before adding metric
 	if ro.Config.Filter.IsActive() {
 		// In order to filter out tags, we need to create a new metric, since
@@ -106,6 +113,11 @@ func (ro *RunningOutput) AddMetric(m telegraf.Metric) {
 		m, _ = metric.New(name, tags, fields, t)
 	}
 
+	if output, ok := ro.Output.(telegraf.AggregatingOutput); ok {
+		output.Add(m)
+		return
+	}
+
 	ro.metrics.Add(m)
 	if ro.metrics.Len() == ro.MetricBatchSize {
 		batch := ro.metrics.Batch(ro.MetricBatchSize)
@@ -118,10 +130,16 @@ func (ro *RunningOutput) AddMetric(m telegraf.Metric) {
 
 // Write writes all cached points to this output.
 func (ro *RunningOutput) Write() error {
+	if output, ok := ro.Output.(telegraf.AggregatingOutput); ok {
+		metrics := output.Push()
+		ro.metrics.Add(metrics...)
+		output.Reset()
+	}
+
 	nFails, nMetrics := ro.failMetrics.Len(), ro.metrics.Len()
+	ro.BufferSize.Set(int64(nFails + nMetrics))
 	log.Printf("D! Output [%s] buffer fullness: %d / %d metrics. ",
 		ro.Name, nFails+nMetrics, ro.MetricBufferLimit)
-	ro.BufferSize.Incr(int64(nFails + nMetrics))
 	var err error
 	if !ro.failMetrics.IsEmpty() {
 		// how many batches of failed writes we need to write.
@@ -166,6 +184,8 @@ func (ro *RunningOutput) write(metrics []telegraf.Metric) error {
 	if nMetrics == 0 {
 		return nil
 	}
+	ro.Lock()
+	defer ro.Unlock()
 	start := time.Now()
 	err := ro.Output.Write(metrics)
 	elapsed := time.Since(start)
@@ -173,7 +193,6 @@ func (ro *RunningOutput) write(metrics []telegraf.Metric) error {
 		log.Printf("D! Output [%s] wrote batch of %d metrics in %s\n",
 			ro.Name, nMetrics, elapsed)
 		ro.MetricsWritten.Incr(int64(nMetrics))
-		ro.BufferSize.Incr(-int64(nMetrics))
 		ro.WriteTime.Incr(elapsed.Nanoseconds())
 	}
 	return err

internal/models/running_output_test.go

@@ -75,6 +75,23 @@ func BenchmarkRunningOutputAddFailWrites(b *testing.B) {
 	}
 }
 
+func TestAddingNilMetric(t *testing.T) {
+	conf := &OutputConfig{
+		Filter: Filter{},
+	}
+
+	m := &mockOutput{}
+	ro := NewRunningOutput("test", m, conf, 1000, 10000)
+
+	ro.AddMetric(nil)
+	ro.AddMetric(nil)
+	ro.AddMetric(nil)
+
+	err := ro.Write()
+	assert.NoError(t, err)
+	assert.Len(t, m.Metrics(), 0)
+}
+
 // Test that NameDrop filters ger properly applied.
 func TestRunningOutput_DropFilter(t *testing.T) {
 	conf := &OutputConfig{

internal/models/running_processor.go

@@ -1,11 +1,15 @@
 package models
 
 import (
+	"sync"
+
 	"github.com/influxdata/telegraf"
 )
 
 type RunningProcessor struct {
-	Name      string
+	Name string
+
+	sync.Mutex
 	Processor telegraf.Processor
 	Config    *ProcessorConfig
 }
@@ -24,6 +28,9 @@ type ProcessorConfig struct {
 }
 
 func (rp *RunningProcessor) Apply(in ...telegraf.Metric) []telegraf.Metric {
+	rp.Lock()
+	defer rp.Unlock()
+
 	ret := []telegraf.Metric{}
 
 	for _, metric := range in {

internal/templating/engine.go

@@ -0,0 +1,86 @@
+package templating
+
+import (
+	"sort"
+	"strings"
+)
+
+const (
+	// DefaultSeparator is the default separation character to use when separating template parts.
+	DefaultSeparator = "."
+)
+
+// Engine uses a Matcher to retrieve the appropriate template and applies the template
+// to the input string
+type Engine struct {
+	joiner  string
+	matcher *matcher
+}
+
+// Apply extracts the template fields from the given line and returns the measurement
+// name, tags and field name
+func (e *Engine) Apply(line string) (string, map[string]string, string, error) {
+	return e.matcher.match(line).Apply(line, e.joiner)
+}
+
+// NewEngine creates a new templating engine
+func NewEngine(joiner string, defaultTemplate *Template, templates []string) (*Engine, error) {
+	engine := Engine{
+		joiner:  joiner,
+		matcher: newMatcher(defaultTemplate),
+	}
+	templateSpecs := parseTemplateSpecs(templates)
+
+	for _, templateSpec := range templateSpecs {
+		if err := engine.matcher.addSpec(templateSpec); err != nil {
+			return nil, err
+		}
+	}
+
+	return &engine, nil
+}
+
+func parseTemplateSpecs(templates []string) templateSpecs {
+	tmplts := templateSpecs{}
+	for _, pattern := range templates {
+		tmplt := templateSpec{
+			separator: DefaultSeparator,
+		}
+
+		// Format is [separator] [filter] <template> [tag1=value1,tag2=value2]
+		parts := strings.Fields(pattern)
+		partsLength := len(parts)
+		if partsLength < 1 {
+			// ignore
+			continue
+		}
+		if partsLength == 1 {
+			tmplt.template = pattern
+		} else if partsLength == 4 {
+			tmplt.separator = parts[0]
+			tmplt.filter = parts[1]
+			tmplt.template = parts[2]
+			tmplt.tagstring = parts[3]
+		} else {
+			hasTagstring := strings.Contains(parts[partsLength-1], "=")
+			if hasTagstring {
+				tmplt.tagstring = parts[partsLength-1]
+				tmplt.template = parts[partsLength-2]
+				if partsLength == 3 {
+					tmplt.filter = parts[0]
+				}
+			} else {
+				tmplt.template = parts[partsLength-1]
+				if partsLength == 2 {
+					tmplt.filter = parts[0]
+				} else { // length == 3
+					tmplt.separator = parts[0]
+					tmplt.filter = parts[1]
+				}
+			}
+		}
+		tmplts = append(tmplts, tmplt)
+	}
+	sort.Sort(tmplts)
+	return tmplts
+}

internal/templating/matcher.go

@@ -0,0 +1,58 @@
+package templating
+
+import (
+	"strings"
+)
+
+// matcher determines which template should be applied to a given metric
+// based on a filter tree.
+type matcher struct {
+	root            *node
+	defaultTemplate *Template
+}
+
+// newMatcher creates a new matcher.
+func newMatcher(defaultTemplate *Template) *matcher {
+	return &matcher{
+		root:            &node{},
+		defaultTemplate: defaultTemplate,
+	}
+}
+
+func (m *matcher) addSpec(tmplt templateSpec) error {
+	// Parse out the default tags specific to this template
+	tags := map[string]string{}
+	if tmplt.tagstring != "" {
+		for _, kv := range strings.Split(tmplt.tagstring, ",") {
+			parts := strings.Split(kv, "=")
+			tags[parts[0]] = parts[1]
+		}
+	}
+
+	tmpl, err := NewTemplate(tmplt.separator, tmplt.template, tags)
+	if err != nil {
+		return err
+	}
+	m.add(tmplt.filter, tmpl)
+	return nil
+}
+
+// add inserts the template in the filter tree based the given filter
+func (m *matcher) add(filter string, template *Template) {
+	if filter == "" {
+		m.defaultTemplate = template
+		m.root.separator = template.separator
+		return
+	}
+	m.root.insert(filter, template)
+}
+
+// match returns the template that matches the given measurement line.
+// If no template matches, the default template is returned.
+func (m *matcher) match(line string) *Template {
+	tmpl := m.root.search(line)
+	if tmpl != nil {
+		return tmpl
+	}
+	return m.defaultTemplate
+}

internal/templating/node.go

@@ -0,0 +1,122 @@
+package templating
+
+import (
+	"sort"
+	"strings"
+)
+
+// node is an item in a sorted k-ary tree of filter parts. Each child is sorted by its part value.
+// The special value of "*", is always sorted last.
+type node struct {
+	separator string
+	value     string
+	children  nodes
+	template  *Template
+}
+
+// insert inserts the given string template into the tree.  The filter string is separated
+// on the template separator and each part is used as the path in the tree.
+func (n *node) insert(filter string, template *Template) {
+	n.separator = template.separator
+	n.recursiveInsert(strings.Split(filter, n.separator), template)
+}
+
+// recursiveInsert does the actual recursive insertion
+func (n *node) recursiveInsert(values []string, template *Template) {
+	// Add the end, set the template
+	if len(values) == 0 {
+		n.template = template
+		return
+	}
+
+	// See if the the current element already exists in the tree. If so, insert the
+	// into that sub-tree
+	for _, v := range n.children {
+		if v.value == values[0] {
+			v.recursiveInsert(values[1:], template)
+			return
+		}
+	}
+
+	// New element, add it to the tree and sort the children
+	newNode := &node{value: values[0]}
+	n.children = append(n.children, newNode)
+	sort.Sort(&n.children)
+
+	// Now insert the rest of the tree into the new element
+	newNode.recursiveInsert(values[1:], template)
+}
+
+// search searches for a template matching the input string
+func (n *node) search(line string) *Template {
+	separator := n.separator
+	return n.recursiveSearch(strings.Split(line, separator))
+}
+
+// recursiveSearch performs the actual recursive search
+func (n *node) recursiveSearch(lineParts []string) *Template {
+	// Nothing to search
+	if len(lineParts) == 0 || len(n.children) == 0 {
+		return n.template
+	}
+
+	// If last element is a wildcard, don't include it in this search since it's sorted
+	// to the end but lexicographically it would not always be and sort.Search assumes
+	// the slice is sorted.
+	length := len(n.children)
+	if n.children[length-1].value == "*" {
+		length--
+	}
+
+	// Find the index of child with an exact match
+	i := sort.Search(length, func(i int) bool {
+		return n.children[i].value >= lineParts[0]
+	})
+
+	// Found an exact match, so search that child sub-tree
+	if i < len(n.children) && n.children[i].value == lineParts[0] {
+		return n.children[i].recursiveSearch(lineParts[1:])
+	}
+	// Not an exact match, see if we have a wildcard child to search
+	if n.children[len(n.children)-1].value == "*" {
+		return n.children[len(n.children)-1].recursiveSearch(lineParts[1:])
+	}
+	return n.template
+}
+
+// nodes is simply an array of nodes implementing the sorting interface.
+type nodes []*node
+
+// Less returns a boolean indicating whether the filter at position j
+// is less than the filter at position k. Filters are order by string
+// comparison of each component parts.  A wildcard value "*" is never
+// less than a non-wildcard value.
+//
+// For example, the filters:
+//             "*.*"
+//             "servers.*"
+//             "servers.localhost"
+//             "*.localhost"
+//
+// Would be sorted as:
+//             "servers.localhost"
+//             "servers.*"
+//             "*.localhost"
+//             "*.*"
+func (n *nodes) Less(j, k int) bool {
+	if (*n)[j].value == "*" && (*n)[k].value != "*" {
+		return false
+	}
+
+	if (*n)[j].value != "*" && (*n)[k].value == "*" {
+		return true
+	}
+
+	return (*n)[j].value < (*n)[k].value
+}
+
+// Swap swaps two elements of the array
+func (n *nodes) Swap(i, j int) { (*n)[i], (*n)[j] = (*n)[j], (*n)[i] }
+
+// Len returns the length of the array
+func (n *nodes) Len() int { return len(*n) }

internal/templating/template.go

@@ -0,0 +1,148 @@
+package templating
+
+import (
+	"fmt"
+	"strings"
+)
+
+// Template represents a pattern and tags to map a metric string to a influxdb Point
+type Template struct {
+	separator         string
+	parts             []string
+	defaultTags       map[string]string
+	greedyField       bool
+	greedyMeasurement bool
+}
+
+// apply extracts the template fields from the given line and returns the measurement
+// name, tags and field name
+func (t *Template) Apply(line string, joiner string) (string, map[string]string, string, error) {
+	fields := strings.Split(line, t.separator)
+	var (
+		measurement []string
+		tags        = make(map[string][]string)
+		field       []string
+	)
+
+	// Set any default tags
+	for k, v := range t.defaultTags {
+		tags[k] = append(tags[k], v)
+	}
+
+	// See if an invalid combination has been specified in the template:
+	for _, tag := range t.parts {
+		if tag == "measurement*" {
+			t.greedyMeasurement = true
+		} else if tag == "field*" {
+			t.greedyField = true
+		}
+	}
+	if t.greedyField && t.greedyMeasurement {
+		return "", nil, "",
+			fmt.Errorf("either 'field*' or 'measurement*' can be used in each "+
+				"template (but not both together): %q",
+				strings.Join(t.parts, joiner))
+	}
+
+	for i, tag := range t.parts {
+		if i >= len(fields) {
+			continue
+		}
+		if tag == "" {
+			continue
+		}
+
+		switch tag {
+		case "measurement":
+			measurement = append(measurement, fields[i])
+		case "field":
+			field = append(field, fields[i])
+		case "field*":
+			field = append(field, fields[i:]...)
+			break
+		case "measurement*":
+			measurement = append(measurement, fields[i:]...)
+			break
+		default:
+			tags[tag] = append(tags[tag], fields[i])
+		}
+	}
+
+	// Convert to map of strings.
+	outtags := make(map[string]string)
+	for k, values := range tags {
+		outtags[k] = strings.Join(values, joiner)
+	}
+
+	return strings.Join(measurement, joiner), outtags, strings.Join(field, joiner), nil
+}
+
+func NewDefaultTemplateWithPattern(pattern string) (*Template, error) {
+	return NewTemplate(DefaultSeparator, pattern, nil)
+}
+
+// NewTemplate returns a new template ensuring it has a measurement
+// specified.
+func NewTemplate(separator string, pattern string, defaultTags map[string]string) (*Template, error) {
+	parts := strings.Split(pattern, separator)
+	hasMeasurement := false
+	template := &Template{
+		separator:   separator,
+		parts:       parts,
+		defaultTags: defaultTags,
+	}
+
+	for _, part := range parts {
+		if strings.HasPrefix(part, "measurement") {
+			hasMeasurement = true
+		}
+		if part == "measurement*" {
+			template.greedyMeasurement = true
+		} else if part == "field*" {
+			template.greedyField = true
+		}
+	}
+
+	if !hasMeasurement {
+		return nil, fmt.Errorf("no measurement specified for template. %q", pattern)
+	}
+
+	return template, nil
+}
+
+// templateSpec is a template string split in its constituent parts
+type templateSpec struct {
+	separator string
+	filter    string
+	template  string
+	tagstring string
+}
+
+// templateSpecs is simply an array of template specs implementing the sorting interface
+type templateSpecs []templateSpec
+
+// Less reports whether the element with
+// index j should sort before the element with index k.
+func (e templateSpecs) Less(j, k int) bool {
+	if len(e[j].filter) == 0 && len(e[k].filter) == 0 {
+		jlength := len(strings.Split(e[j].template, e[j].separator))
+		klength := len(strings.Split(e[k].template, e[k].separator))
+		return jlength < klength
+	}
+	if len(e[j].filter) == 0 {
+		return true
+	}
+	if len(e[k].filter) == 0 {
+		return false
+	}
+
+	jlength := len(strings.Split(e[j].template, e[j].separator))
+	klength := len(strings.Split(e[k].template, e[k].separator))
+	return jlength < klength
+}
+
+// Swap swaps the elements with indexes i and j.
+func (e templateSpecs) Swap(i, j int) { e[i], e[j] = e[j], e[i] }
+
+// Len is the number of elements in the collection.
+func (e templateSpecs) Len() int { return len(e) }

logger/logger.go

@@ -4,11 +4,14 @@ import (
 	"io"
 	"log"
 	"os"
+	"regexp"
 	"time"
 
 	"github.com/influxdata/wlog"
 )
 
+var prefixRegex = regexp.MustCompile("^[DIWE]!")
+
 // newTelegrafWriter returns a logging-wrapped writer.
 func newTelegrafWriter(w io.Writer) io.Writer {
 	return &telegrafLog{
@@ -21,7 +24,13 @@ type telegrafLog struct {
 }
 
 func (t *telegrafLog) Write(b []byte) (n int, err error) {
-	return t.writer.Write(append([]byte(time.Now().UTC().Format(time.RFC3339)+" "), b...))
+	var line []byte
+	if !prefixRegex.Match(b) {
+		line = append([]byte(time.Now().UTC().Format(time.RFC3339)+" I! "), b...)
+	} else {
+		line = append([]byte(time.Now().UTC().Format(time.RFC3339)+" "), b...)
+	}
+	return t.writer.Write(line)
 }
 
 // SetupLogging configures the logging output.

logger/logger_test.go

@@ -51,6 +51,19 @@ func TestErrorWriteLogToFile(t *testing.T) {
 	assert.Equal(t, f[19:], []byte("Z E! TEST\n"))
 }
 
+func TestAddDefaultLogLevel(t *testing.T) {
+	tmpfile, err := ioutil.TempFile("", "")
+	assert.NoError(t, err)
+	defer func() { os.Remove(tmpfile.Name()) }()
+
+	SetupLogging(true, false, tmpfile.Name())
+	log.Printf("TEST")
+
+	f, err := ioutil.ReadFile(tmpfile.Name())
+	assert.NoError(t, err)
+	assert.Equal(t, f[19:], []byte("Z I! TEST\n"))
+}
+
 func BenchmarkTelegrafLogWrite(b *testing.B) {
 	var msg = []byte("test")
 	var buf bytes.Buffer

metric.go

@@ -2,9 +2,6 @@ package telegraf
 
 import (
 	"time"
-
-	// TODO remove
-	"github.com/influxdata/influxdb/client/v2"
 )
 
 // ValueType is an enumeration of metric types that represent a simple value.
@@ -16,47 +13,54 @@ const (
 	Counter
 	Gauge
 	Untyped
+	Summary
+	Histogram
 )
 
+type Tag struct {
+	Key   string
+	Value string
+}
+
+type Field struct {
+	Key   string
+	Value interface{}
+}
+
 type Metric interface {
-	Serialize() []byte
-	String() string // convenience function for string(Serialize())
-	Copy() Metric
-	// Split will attempt to return multiple metrics with the same timestamp
-	// whose string representations are no longer than maxSize.
-	// Metrics with a single field may exceed the requested size.
-	Split(maxSize int) []Metric
+	// Getting data structure functions
+	Name() string
+	Tags() map[string]string
+	TagList() []*Tag
+	Fields() map[string]interface{}
+	FieldList() []*Field
+	Time() time.Time
+	Type() ValueType
+
+	// Name functions
+	SetName(name string)
+	AddPrefix(prefix string)
+	AddSuffix(suffix string)
 
 	// Tag functions
+	GetTag(key string) (string, bool)
 	HasTag(key string) bool
 	AddTag(key, value string)
 	RemoveTag(key string)
 
 	// Field functions
+	GetField(key string) (interface{}, bool)
 	HasField(key string) bool
 	AddField(key string, value interface{})
-	RemoveField(key string) error
+	RemoveField(key string)
 
-	// Name functions
-	SetName(name string)
-	SetPrefix(prefix string)
-	SetSuffix(suffix string)
-
-	// Getting data structure functions
-	Name() string
-	Tags() map[string]string
-	Fields() map[string]interface{}
-	Time() time.Time
-	UnixNano() int64
-	Type() ValueType
-	Len() int // returns the length of the serialized metric, including newline
+	// HashID returns an unique identifier for the series.
 	HashID() uint64
 
-	// aggregator things:
+	// Copy returns a deep copy of the Metric.
+	Copy() Metric
+
+	// Mark Metric as an aggregate
 	SetAggregate(bool)
 	IsAggregate() bool
-
-	// Point returns a influxdb client.Point object
-	// TODO remove this function
-	Point() *client.Point
 }

metric/builder.go

@@ -0,0 +1,53 @@
+package metric
+
+import (
+	"time"
+
+	"github.com/influxdata/telegraf"
+)
+
+type TimeFunc func() time.Time
+
+type Builder struct {
+	TimeFunc
+	TimePrecision time.Duration
+
+	*metric
+}
+
+func NewBuilder() *Builder {
+	b := &Builder{
+		TimeFunc:      time.Now,
+		TimePrecision: 1 * time.Nanosecond,
+	}
+	b.Reset()
+	return b
+}
+
+func (b *Builder) SetName(name string) {
+	b.name = name
+}
+
+func (b *Builder) AddTag(key string, value string) {
+	b.metric.AddTag(key, value)
+}
+
+func (b *Builder) AddField(key string, value interface{}) {
+	b.metric.AddField(key, value)
+}
+
+func (b *Builder) SetTime(tm time.Time) {
+	b.tm = tm
+}
+
+func (b *Builder) Reset() {
+	b.metric = &metric{}
+}
+
+func (b *Builder) Metric() (telegraf.Metric, error) {
+	if b.tm.IsZero() {
+		b.tm = b.TimeFunc().Truncate(b.TimePrecision)
+	}
+
+	return b.metric, nil
+}

metric/escape.go

@@ -1,49 +0,0 @@
-package metric
-
-import (
-	"strings"
-)
-
-var (
-	// escaper is for escaping:
-	//   - tag keys
-	//   - tag values
-	//   - field keys
-	// see https://docs.influxdata.com/influxdb/v1.0/write_protocols/line_protocol_tutorial/#special-characters-and-keywords
-	escaper   = strings.NewReplacer(`,`, `\,`, `"`, `\"`, ` `, `\ `, `=`, `\=`)
-	unEscaper = strings.NewReplacer(`\,`, `,`, `\"`, `"`, `\ `, ` `, `\=`, `=`)
-
-	// nameEscaper is for escaping measurement names only.
-	// see https://docs.influxdata.com/influxdb/v1.0/write_protocols/line_protocol_tutorial/#special-characters-and-keywords
-	nameEscaper   = strings.NewReplacer(`,`, `\,`, ` `, `\ `)
-	nameUnEscaper = strings.NewReplacer(`\,`, `,`, `\ `, ` `)
-
-	// stringFieldEscaper is for escaping string field values only.
-	// see https://docs.influxdata.com/influxdb/v1.0/write_protocols/line_protocol_tutorial/#special-characters-and-keywords
-	stringFieldEscaper   = strings.NewReplacer(`"`, `\"`)
-	stringFieldUnEscaper = strings.NewReplacer(`\"`, `"`)
-)
-
-func escape(s string, t string) string {
-	switch t {
-	case "fieldkey", "tagkey", "tagval":
-		return escaper.Replace(s)
-	case "name":
-		return nameEscaper.Replace(s)
-	case "fieldval":
-		return stringFieldEscaper.Replace(s)
-	}
-	return s
-}
-
-func unescape(s string, t string) string {
-	switch t {
-	case "fieldkey", "tagkey", "tagval":
-		return unEscaper.Replace(s)
-	case "name":
-		return nameUnEscaper.Replace(s)
-	case "fieldval":
-		return stringFieldUnEscaper.Replace(s)
-	}
-	return s
-}

metric/inline_strconv_parse.go

@@ -1,38 +0,0 @@
-package metric
-
-import (
-	"reflect"
-	"strconv"
-	"unsafe"
-)
-
-// parseIntBytes is a zero-alloc wrapper around strconv.ParseInt.
-func parseIntBytes(b []byte, base int, bitSize int) (i int64, err error) {
-	s := unsafeBytesToString(b)
-	return strconv.ParseInt(s, base, bitSize)
-}
-
-// parseFloatBytes is a zero-alloc wrapper around strconv.ParseFloat.
-func parseFloatBytes(b []byte, bitSize int) (float64, error) {
-	s := unsafeBytesToString(b)
-	return strconv.ParseFloat(s, bitSize)
-}
-
-// parseBoolBytes is a zero-alloc wrapper around strconv.ParseBool.
-func parseBoolBytes(b []byte) (bool, error) {
-	return strconv.ParseBool(unsafeBytesToString(b))
-}
-
-// unsafeBytesToString converts a []byte to a string without a heap allocation.
-//
-// It is unsafe, and is intended to prepare input to short-lived functions
-// that require strings.
-func unsafeBytesToString(in []byte) string {
-	src := *(*reflect.SliceHeader)(unsafe.Pointer(&in))
-	dst := reflect.StringHeader{
-		Data: src.Data,
-		Len:  src.Len,
-	}
-	s := *(*string)(unsafe.Pointer(&dst))
-	return s
-}

metric/inline_strconv_parse_test.go

@@ -1,103 +0,0 @@
-package metric
-
-import (
-	"strconv"
-	"testing"
-	"testing/quick"
-)
-
-func TestParseIntBytesEquivalenceFuzz(t *testing.T) {
-	f := func(b []byte, base int, bitSize int) bool {
-		exp, expErr := strconv.ParseInt(string(b), base, bitSize)
-		got, gotErr := parseIntBytes(b, base, bitSize)
-
-		return exp == got && checkErrs(expErr, gotErr)
-	}
-
-	cfg := &quick.Config{
-		MaxCount: 10000,
-	}
-
-	if err := quick.Check(f, cfg); err != nil {
-		t.Fatal(err)
-	}
-}
-
-func TestParseIntBytesValid64bitBase10EquivalenceFuzz(t *testing.T) {
-	buf := []byte{}
-	f := func(n int64) bool {
-		buf = strconv.AppendInt(buf[:0], n, 10)
-
-		exp, expErr := strconv.ParseInt(string(buf), 10, 64)
-		got, gotErr := parseIntBytes(buf, 10, 64)
-
-		return exp == got && checkErrs(expErr, gotErr)
-	}
-
-	cfg := &quick.Config{
-		MaxCount: 10000,
-	}
-
-	if err := quick.Check(f, cfg); err != nil {
-		t.Fatal(err)
-	}
-}
-
-func TestParseFloatBytesEquivalenceFuzz(t *testing.T) {
-	f := func(b []byte, bitSize int) bool {
-		exp, expErr := strconv.ParseFloat(string(b), bitSize)
-		got, gotErr := parseFloatBytes(b, bitSize)
-
-		return exp == got && checkErrs(expErr, gotErr)
-	}
-
-	cfg := &quick.Config{
-		MaxCount: 10000,
-	}
-
-	if err := quick.Check(f, cfg); err != nil {
-		t.Fatal(err)
-	}
-}
-
-func TestParseFloatBytesValid64bitEquivalenceFuzz(t *testing.T) {
-	buf := []byte{}
-	f := func(n float64) bool {
-		buf = strconv.AppendFloat(buf[:0], n, 'f', -1, 64)
-
-		exp, expErr := strconv.ParseFloat(string(buf), 64)
-		got, gotErr := parseFloatBytes(buf, 64)
-
-		return exp == got && checkErrs(expErr, gotErr)
-	}
-
-	cfg := &quick.Config{
-		MaxCount: 10000,
-	}
-
-	if err := quick.Check(f, cfg); err != nil {
-		t.Fatal(err)
-	}
-}
-
-func TestParseBoolBytesEquivalence(t *testing.T) {
-	var buf []byte
-	for _, s := range []string{"1", "t", "T", "TRUE", "true", "True", "0", "f", "F", "FALSE", "false", "False", "fail", "TrUe", "FAlSE", "numbers", ""} {
-		buf = append(buf[:0], s...)
-
-		exp, expErr := strconv.ParseBool(s)
-		got, gotErr := parseBoolBytes(buf)
-
-		if got != exp || !checkErrs(expErr, gotErr) {
-			t.Errorf("Failed to parse boolean value %q correctly: wanted (%t, %v), got (%t, %v)", s, exp, expErr, got, gotErr)
-		}
-	}
-}
-
-func checkErrs(a, b error) bool {
-	if (a == nil) != (b == nil) {
-		return false
-	}
-
-	return a == nil || a.Error() == b.Error()
-}

metric/metric.go

@@ -1,546 +1,282 @@
 package metric
 
 import (
-	"bytes"
 	"fmt"
 	"hash/fnv"
 	"sort"
-	"strconv"
 	"time"
 
 	"github.com/influxdata/telegraf"
-
-	// TODO remove
-	"github.com/influxdata/influxdb/client/v2"
 )
 
-const MaxInt = int(^uint(0) >> 1)
+type metric struct {
+	name   string
+	tags   []*telegraf.Tag
+	fields []*telegraf.Field
+	tm     time.Time
+
+	tp        telegraf.ValueType
+	aggregate bool
+}
 
 func New(
 	name string,
 	tags map[string]string,
 	fields map[string]interface{},
-	t time.Time,
-	mType ...telegraf.ValueType,
+	tm time.Time,
+	tp ...telegraf.ValueType,
 ) (telegraf.Metric, error) {
-	if len(fields) == 0 {
-		return nil, fmt.Errorf("Metric cannot be made without any fields")
-	}
-	if len(name) == 0 {
-		return nil, fmt.Errorf("Metric cannot be made with an empty name")
-	}
-
-	var thisType telegraf.ValueType
-	if len(mType) > 0 {
-		thisType = mType[0]
+	var vtype telegraf.ValueType
+	if len(tp) > 0 {
+		vtype = tp[0]
 	} else {
-		thisType = telegraf.Untyped
+		vtype = telegraf.Untyped
 	}
 
 	m := &metric{
-		name:  []byte(escape(name, "name")),
-		t:     []byte(fmt.Sprint(t.UnixNano())),
-		nsec:  t.UnixNano(),
-		mType: thisType,
+		name:   name,
+		tags:   nil,
+		fields: nil,
+		tm:     tm,
+		tp:     vtype,
 	}
 
-	// pre-allocate exact size of the tags slice
-	taglen := 0
-	for k, v := range tags {
-		// TODO check that length of tag key & value are > 0
-		taglen += 2 + len(escape(k, "tagkey")) + len(escape(v, "tagval"))
-	}
-	m.tags = make([]byte, taglen)
-
-	i := 0
-	for k, v := range tags {
-		m.tags[i] = ','
-		i++
-		i += copy(m.tags[i:], escape(k, "tagkey"))
-		m.tags[i] = '='
-		i++
-		i += copy(m.tags[i:], escape(v, "tagval"))
-	}
-
-	// pre-allocate capacity of the fields slice
-	fieldlen := 0
-	for k, _ := range fields {
-		// 10 bytes is completely arbitrary, but will at least prevent some
-		// amount of allocations. There's a small possibility this will create
-		// slightly more allocations for a metric that has many short fields.
-		fieldlen += len(k) + 10
-	}
-	m.fields = make([]byte, 0, fieldlen)
-
-	i = 0
-	for k, v := range fields {
-		if i != 0 {
-			m.fields = append(m.fields, ',')
+	if len(tags) > 0 {
+		m.tags = make([]*telegraf.Tag, 0, len(tags))
+		for k, v := range tags {
+			m.tags = append(m.tags,
+				&telegraf.Tag{Key: k, Value: v})
 		}
-		m.fields = appendField(m.fields, k, v)
-		i++
+		sort.Slice(m.tags, func(i, j int) bool { return m.tags[i].Key < m.tags[j].Key })
+	}
+
+	m.fields = make([]*telegraf.Field, 0, len(fields))
+	for k, v := range fields {
+		v := convertField(v)
+		if v == nil {
+			continue
+		}
+		m.AddField(k, v)
 	}
 
 	return m, nil
 }
 
-// indexUnescapedByte finds the index of the first byte equal to b in buf that
-// is not escaped. Returns -1 if not found.
-func indexUnescapedByte(buf []byte, b byte) int {
-	var keyi int
-	for {
-		i := bytes.IndexByte(buf[keyi:], b)
-		if i == -1 {
-			return -1
-		} else if i == 0 {
-			break
-		}
-		keyi += i
-		if countBackslashes(buf, keyi-1)%2 == 0 {
-			break
-		} else {
-			keyi++
-		}
-	}
-	return keyi
-}
-
-// countBackslashes counts the number of preceding backslashes starting at
-// the 'start' index.
-func countBackslashes(buf []byte, index int) int {
-	var count int
-	for {
-		if index < 0 {
-			return count
-		}
-		if buf[index] == '\\' {
-			count++
-			index--
-		} else {
-			break
-		}
-	}
-	return count
-}
-
-type metric struct {
-	name   []byte
-	tags   []byte
-	fields []byte
-	t      []byte
-
-	mType     telegraf.ValueType
-	aggregate bool
-
-	// cached values for reuse in "get" functions
-	hashID uint64
-	nsec   int64
-}
-
-func (m *metric) Point() *client.Point {
-	c, _ := client.NewPoint(m.Name(), m.Tags(), m.Fields(), m.Time())
-	return c
-}
-
 func (m *metric) String() string {
-	return string(m.name) + string(m.tags) + " " + string(m.fields) + " " + string(m.t) + "\n"
+	return fmt.Sprintf("%s %v %v %d", m.name, m.Tags(), m.Fields(), m.tm.UnixNano())
+}
+
+func (m *metric) Name() string {
+	return m.name
+}
+
+func (m *metric) Tags() map[string]string {
+	tags := make(map[string]string, len(m.tags))
+	for _, tag := range m.tags {
+		tags[tag.Key] = tag.Value
+	}
+	return tags
+}
+
+func (m *metric) TagList() []*telegraf.Tag {
+	return m.tags
+}
+
+func (m *metric) Fields() map[string]interface{} {
+	fields := make(map[string]interface{}, len(m.fields))
+	for _, field := range m.fields {
+		fields[field.Key] = field.Value
+	}
+
+	return fields
+}
+
+func (m *metric) FieldList() []*telegraf.Field {
+	return m.fields
+}
+
+func (m *metric) Time() time.Time {
+	return m.tm
+}
+
+func (m *metric) Type() telegraf.ValueType {
+	return m.tp
+}
+
+func (m *metric) SetName(name string) {
+	m.name = name
+}
+
+func (m *metric) AddPrefix(prefix string) {
+	m.name = prefix + m.name
+}
+
+func (m *metric) AddSuffix(suffix string) {
+	m.name = m.name + suffix
+}
+
+func (m *metric) AddTag(key, value string) {
+	for i, tag := range m.tags {
+		if key > tag.Key {
+			continue
+		}
+
+		if key == tag.Key {
+			tag.Value = value
+			return
+		}
+
+		m.tags = append(m.tags, nil)
+		copy(m.tags[i+1:], m.tags[i:])
+		m.tags[i] = &telegraf.Tag{Key: key, Value: value}
+		return
+	}
+
+	m.tags = append(m.tags, &telegraf.Tag{Key: key, Value: value})
+}
+
+func (m *metric) HasTag(key string) bool {
+	for _, tag := range m.tags {
+		if tag.Key == key {
+			return true
+		}
+	}
+	return false
+}
+
+func (m *metric) GetTag(key string) (string, bool) {
+	for _, tag := range m.tags {
+		if tag.Key == key {
+			return tag.Value, true
+		}
+	}
+	return "", false
+}
+
+func (m *metric) RemoveTag(key string) {
+	for i, tag := range m.tags {
+		if tag.Key == key {
+			copy(m.tags[i:], m.tags[i+1:])
+			m.tags[len(m.tags)-1] = nil
+			m.tags = m.tags[:len(m.tags)-1]
+			return
+		}
+	}
+}
+
+func (m *metric) AddField(key string, value interface{}) {
+	for i, field := range m.fields {
+		if key == field.Key {
+			m.fields[i] = &telegraf.Field{Key: key, Value: convertField(value)}
+		}
+	}
+	m.fields = append(m.fields, &telegraf.Field{Key: key, Value: convertField(value)})
+}
+
+func (m *metric) HasField(key string) bool {
+	for _, field := range m.fields {
+		if field.Key == key {
+			return true
+		}
+	}
+	return false
+}
+
+func (m *metric) GetField(key string) (interface{}, bool) {
+	for _, field := range m.fields {
+		if field.Key == key {
+			return field.Value, true
+		}
+	}
+	return nil, false
+}
+
+func (m *metric) RemoveField(key string) {
+	for i, field := range m.fields {
+		if field.Key == key {
+			copy(m.fields[i:], m.fields[i+1:])
+			m.fields[len(m.fields)-1] = nil
+			m.fields = m.fields[:len(m.fields)-1]
+			return
+		}
+	}
+}
+
+func (m *metric) Copy() telegraf.Metric {
+	m2 := &metric{
+		name:      m.name,
+		tags:      make([]*telegraf.Tag, len(m.tags)),
+		fields:    make([]*telegraf.Field, len(m.fields)),
+		tm:        m.tm,
+		tp:        m.tp,
+		aggregate: m.aggregate,
+	}
+
+	for i, tag := range m.tags {
+		m2.tags[i] = tag
+	}
+
+	for i, field := range m.fields {
+		m2.fields[i] = field
+	}
+	return m2
 }
 
 func (m *metric) SetAggregate(b bool) {
-	m.aggregate = b
+	m.aggregate = true
 }
 
 func (m *metric) IsAggregate() bool {
 	return m.aggregate
 }
 
-func (m *metric) Type() telegraf.ValueType {
-	return m.mType
-}
-
-func (m *metric) Len() int {
-	// 3 is for 2 spaces surrounding the fields array + newline at the end.
-	return len(m.name) + len(m.tags) + len(m.fields) + len(m.t) + 3
-}
-
-func (m *metric) Serialize() []byte {
-	tmp := make([]byte, m.Len())
-	i := 0
-	i += copy(tmp[i:], m.name)
-	i += copy(tmp[i:], m.tags)
-	tmp[i] = ' '
-	i++
-	i += copy(tmp[i:], m.fields)
-	tmp[i] = ' '
-	i++
-	i += copy(tmp[i:], m.t)
-	tmp[i] = '\n'
-	return tmp
-}
-
-func (m *metric) Split(maxSize int) []telegraf.Metric {
-	if m.Len() < maxSize {
-		return []telegraf.Metric{m}
-	}
-	var out []telegraf.Metric
-
-	// constant number of bytes for each metric (in addition to field bytes)
-	constant := len(m.name) + len(m.tags) + len(m.t) + 3
-	// currently selected fields
-	fields := make([]byte, 0, maxSize)
-
-	i := 0
-	for {
-		if i >= len(m.fields) {
-			// hit the end of the field byte slice
-			if len(fields) > 0 {
-				out = append(out, copyWith(m.name, m.tags, fields, m.t))
-			}
-			break
-		}
-
-		// find the end of the next field
-		j := indexUnescapedByte(m.fields[i:], ',')
-		if j == -1 {
-			j = len(m.fields)
-		} else {
-			j += i
-		}
-
-		// if true, then we need to create a metric _not_ including the currently
-		// selected field
-		if len(m.fields[i:j])+len(fields)+constant > maxSize {
-			// if false, then we'll create a metric including the currently
-			// selected field anyways. This means that the given maxSize is too
-			// small for a single field to fit.
-			if len(fields) > 0 {
-				out = append(out, copyWith(m.name, m.tags, fields, m.t))
-			}
-
-			fields = make([]byte, 0, maxSize)
-		}
-		if len(fields) > 0 {
-			fields = append(fields, ',')
-		}
-		fields = append(fields, m.fields[i:j]...)
-
-		i = j + 1
-	}
-	return out
-}
-
-func (m *metric) Fields() map[string]interface{} {
-	fieldMap := map[string]interface{}{}
-	i := 0
-	for {
-		if i >= len(m.fields) {
-			break
-		}
-		// end index of field key
-		i1 := indexUnescapedByte(m.fields[i:], '=')
-		if i1 == -1 {
-			break
-		}
-		// start index of field value
-		i2 := i1 + 1
-
-		// end index of field value
-		var i3 int
-		if m.fields[i:][i2] == '"' {
-			i3 = indexUnescapedByte(m.fields[i:][i2+1:], '"')
-			if i3 == -1 {
-				i3 = len(m.fields[i:])
-			}
-			i3 += i2 + 2 // increment index to the comma
-		} else {
-			i3 = indexUnescapedByte(m.fields[i:], ',')
-			if i3 == -1 {
-				i3 = len(m.fields[i:])
-			}
-		}
-
-		switch m.fields[i:][i2] {
-		case '"':
-			// string field
-			fieldMap[unescape(string(m.fields[i:][0:i1]), "fieldkey")] = unescape(string(m.fields[i:][i2+1:i3-1]), "fieldval")
-		case '0', '1', '2', '3', '4', '5', '6', '7', '8', '9':
-			// number field
-			switch m.fields[i:][i3-1] {
-			case 'i':
-				// integer field
-				n, err := parseIntBytes(m.fields[i:][i2:i3-1], 10, 64)
-				if err == nil {
-					fieldMap[unescape(string(m.fields[i:][0:i1]), "fieldkey")] = n
-				} else {
-					// TODO handle error or just ignore field silently?
-				}
-			default:
-				// float field
-				n, err := parseFloatBytes(m.fields[i:][i2:i3], 64)
-				if err == nil {
-					fieldMap[unescape(string(m.fields[i:][0:i1]), "fieldkey")] = n
-				} else {
-					// TODO handle error or just ignore field silently?
-				}
-			}
-		case 'T', 't':
-			fieldMap[unescape(string(m.fields[i:][0:i1]), "fieldkey")] = true
-		case 'F', 'f':
-			fieldMap[unescape(string(m.fields[i:][0:i1]), "fieldkey")] = false
-		default:
-			// TODO handle unsupported field type
-		}
-
-		i += i3 + 1
-	}
-
-	return fieldMap
-}
-
-func (m *metric) Tags() map[string]string {
-	tagMap := map[string]string{}
-	if len(m.tags) == 0 {
-		return tagMap
-	}
-
-	i := 0
-	for {
-		// start index of tag key
-		i0 := indexUnescapedByte(m.tags[i:], ',') + 1
-		if i0 == 0 {
-			// didn't find a tag start
-			break
-		}
-		// end index of tag key
-		i1 := indexUnescapedByte(m.tags[i:], '=')
-		// start index of tag value
-		i2 := i1 + 1
-		// end index of tag value (starting from i2)
-		i3 := indexUnescapedByte(m.tags[i+i2:], ',')
-		if i3 == -1 {
-			tagMap[unescape(string(m.tags[i:][i0:i1]), "tagkey")] = unescape(string(m.tags[i:][i2:]), "tagval")
-			break
-		}
-		tagMap[unescape(string(m.tags[i:][i0:i1]), "tagkey")] = unescape(string(m.tags[i:][i2:i2+i3]), "tagval")
-		// increment start index for the next tag
-		i += i2 + i3
-	}
-
-	return tagMap
-}
-
-func (m *metric) Name() string {
-	return unescape(string(m.name), "name")
-}
-
-func (m *metric) Time() time.Time {
-	// assume metric has been verified already and ignore error:
-	if m.nsec == 0 {
-		m.nsec, _ = parseIntBytes(m.t, 10, 64)
-	}
-	return time.Unix(0, m.nsec)
-}
-
-func (m *metric) UnixNano() int64 {
-	// assume metric has been verified already and ignore error:
-	if m.nsec == 0 {
-		m.nsec, _ = parseIntBytes(m.t, 10, 64)
-	}
-	return m.nsec
-}
-
-func (m *metric) SetName(name string) {
-	m.hashID = 0
-	m.name = []byte(nameEscaper.Replace(name))
-}
-
-func (m *metric) SetPrefix(prefix string) {
-	m.hashID = 0
-	m.name = append([]byte(nameEscaper.Replace(prefix)), m.name...)
-}
-
-func (m *metric) SetSuffix(suffix string) {
-	m.hashID = 0
-	m.name = append(m.name, []byte(nameEscaper.Replace(suffix))...)
-}
-
-func (m *metric) AddTag(key, value string) {
-	m.RemoveTag(key)
-	m.tags = append(m.tags, []byte(","+escape(key, "tagkey")+"="+escape(value, "tagval"))...)
-}
-
-func (m *metric) HasTag(key string) bool {
-	i := bytes.Index(m.tags, []byte(escape(key, "tagkey")+"="))
-	if i == -1 {
-		return false
-	}
-	return true
-}
-
-func (m *metric) RemoveTag(key string) {
-	m.hashID = 0
-
-	i := bytes.Index(m.tags, []byte(escape(key, "tagkey")+"="))
-	if i == -1 {
-		return
-	}
-
-	tmp := m.tags[0 : i-1]
-	j := indexUnescapedByte(m.tags[i:], ',')
-	if j != -1 {
-		tmp = append(tmp, m.tags[i+j:]...)
-	}
-	m.tags = tmp
-	return
-}
-
-func (m *metric) AddField(key string, value interface{}) {
-	m.fields = append(m.fields, ',')
-	m.fields = appendField(m.fields, key, value)
-}
-
-func (m *metric) HasField(key string) bool {
-	i := bytes.Index(m.fields, []byte(escape(key, "tagkey")+"="))
-	if i == -1 {
-		return false
-	}
-	return true
-}
-
-func (m *metric) RemoveField(key string) error {
-	i := bytes.Index(m.fields, []byte(escape(key, "tagkey")+"="))
-	if i == -1 {
-		return nil
-	}
-
-	var tmp []byte
-	if i != 0 {
-		tmp = m.fields[0 : i-1]
-	}
-	j := indexUnescapedByte(m.fields[i:], ',')
-	if j != -1 {
-		tmp = append(tmp, m.fields[i+j:]...)
-	}
-
-	if len(tmp) == 0 {
-		return fmt.Errorf("Metric cannot remove final field: %s", m.fields)
-	}
-
-	m.fields = tmp
-	return nil
-}
-
-func (m *metric) Copy() telegraf.Metric {
-	return copyWith(m.name, m.tags, m.fields, m.t)
-}
-
-func copyWith(name, tags, fields, t []byte) telegraf.Metric {
-	out := metric{
-		name:   make([]byte, len(name)),
-		tags:   make([]byte, len(tags)),
-		fields: make([]byte, len(fields)),
-		t:      make([]byte, len(t)),
-	}
-	copy(out.name, name)
-	copy(out.tags, tags)
-	copy(out.fields, fields)
-	copy(out.t, t)
-	return &out
-}
-
 func (m *metric) HashID() uint64 {
-	if m.hashID == 0 {
-		h := fnv.New64a()
-		h.Write(m.name)
-
-		tags := m.Tags()
-		tmp := make([]string, len(tags))
-		i := 0
-		for k, v := range tags {
-			tmp[i] = k + v
-			i++
-		}
-		sort.Strings(tmp)
-
-		for _, s := range tmp {
-			h.Write([]byte(s))
-		}
-
-		m.hashID = h.Sum64()
+	h := fnv.New64a()
+	h.Write([]byte(m.name))
+	h.Write([]byte("\n"))
+	for _, tag := range m.tags {
+		h.Write([]byte(tag.Key))
+		h.Write([]byte("\n"))
+		h.Write([]byte(tag.Value))
+		h.Write([]byte("\n"))
 	}
-	return m.hashID
+	return h.Sum64()
 }
 
-func appendField(b []byte, k string, v interface{}) []byte {
-	if v == nil {
-		return b
-	}
-	b = append(b, []byte(escape(k, "tagkey")+"=")...)
-
-	// check popular types first
+// Convert field to a supported type or nil if unconvertible
+func convertField(v interface{}) interface{} {
 	switch v := v.(type) {
 	case float64:
-		b = strconv.AppendFloat(b, v, 'f', -1, 64)
+		return v
 	case int64:
-		b = strconv.AppendInt(b, v, 10)
-		b = append(b, 'i')
+		return v
 	case string:
-		b = append(b, '"')
-		b = append(b, []byte(escape(v, "fieldval"))...)
-		b = append(b, '"')
+		return v
 	case bool:
-		b = strconv.AppendBool(b, v)
-	case int32:
-		b = strconv.AppendInt(b, int64(v), 10)
-		b = append(b, 'i')
-	case int16:
-		b = strconv.AppendInt(b, int64(v), 10)
-		b = append(b, 'i')
-	case int8:
-		b = strconv.AppendInt(b, int64(v), 10)
-		b = append(b, 'i')
+		return v
 	case int:
-		b = strconv.AppendInt(b, int64(v), 10)
-		b = append(b, 'i')
-	case uint64:
-		// Cap uints above the maximum int value
-		var intv int64
-		if v <= uint64(MaxInt) {
-			intv = int64(v)
-		} else {
-			intv = int64(MaxInt)
-		}
-		b = strconv.AppendInt(b, intv, 10)
-		b = append(b, 'i')
-	case uint32:
-		b = strconv.AppendInt(b, int64(v), 10)
-		b = append(b, 'i')
-	case uint16:
-		b = strconv.AppendInt(b, int64(v), 10)
-		b = append(b, 'i')
-	case uint8:
-		b = strconv.AppendInt(b, int64(v), 10)
-		b = append(b, 'i')
+		return int64(v)
 	case uint:
-		// Cap uints above the maximum int value
-		var intv int64
-		if v <= uint(MaxInt) {
-			intv = int64(v)
-		} else {
-			intv = int64(MaxInt)
-		}
-		b = strconv.AppendInt(b, intv, 10)
-		b = append(b, 'i')
-	case float32:
-		b = strconv.AppendFloat(b, float64(v), 'f', -1, 32)
+		return uint64(v)
+	case uint64:
+		return uint64(v)
 	case []byte:
-		b = append(b, v...)
+		return string(v)
+	case int32:
+		return int64(v)
+	case int16:
+		return int64(v)
+	case int8:
+		return int64(v)
+	case uint32:
+		return uint64(v)
+	case uint16:
+		return uint64(v)
+	case uint8:
+		return uint64(v)
+	case float32:
+		return float64(v)
 	default:
-		// Can't determine the type, so convert to string
-		b = append(b, '"')
-		b = append(b, []byte(escape(fmt.Sprintf("%v", v), "fieldval"))...)
-		b = append(b, '"')
+		return nil
 	}
-
-	return b
 }

metric/metric_benchmark_test.go

@@ -1,148 +0,0 @@
-package metric
-
-import (
-	"fmt"
-	"testing"
-	"time"
-
-	"github.com/influxdata/telegraf"
-)
-
-// vars for making sure that the compiler doesnt optimize out the benchmarks:
-var (
-	s      string
-	I      interface{}
-	tags   map[string]string
-	fields map[string]interface{}
-)
-
-func BenchmarkNewMetric(b *testing.B) {
-	var mt telegraf.Metric
-	for n := 0; n < b.N; n++ {
-		mt, _ = New("test_metric",
-			map[string]string{
-				"test_tag_1": "tag_value_1",
-				"test_tag_2": "tag_value_2",
-				"test_tag_3": "tag_value_3",
-			},
-			map[string]interface{}{
-				"string_field": "string",
-				"int_field":    int64(1000),
-				"float_field":  float64(2.1),
-			},
-			time.Now(),
-		)
-	}
-	s = string(mt.String())
-}
-
-func BenchmarkAddTag(b *testing.B) {
-	var mt telegraf.Metric
-	mt = &metric{
-		name:   []byte("cpu"),
-		tags:   []byte(",host=localhost"),
-		fields: []byte("a=101"),
-		t:      []byte("1480614053000000000"),
-	}
-	for n := 0; n < b.N; n++ {
-		mt.AddTag("foo", "bar")
-	}
-	s = string(mt.String())
-}
-
-func BenchmarkSplit(b *testing.B) {
-	var mt telegraf.Metric
-	mt = &metric{
-		name:   []byte("cpu"),
-		tags:   []byte(",host=localhost"),
-		fields: []byte("a=101,b=10i,c=10101,d=101010,e=42"),
-		t:      []byte("1480614053000000000"),
-	}
-	var metrics []telegraf.Metric
-	for n := 0; n < b.N; n++ {
-		metrics = mt.Split(60)
-	}
-	s = string(metrics[0].String())
-}
-
-func BenchmarkTags(b *testing.B) {
-	for n := 0; n < b.N; n++ {
-		var mt, _ = New("test_metric",
-			map[string]string{
-				"test_tag_1": "tag_value_1",
-				"test_tag_2": "tag_value_2",
-				"test_tag_3": "tag_value_3",
-			},
-			map[string]interface{}{
-				"string_field": "string",
-				"int_field":    int64(1000),
-				"float_field":  float64(2.1),
-			},
-			time.Now(),
-		)
-		tags = mt.Tags()
-	}
-	s = fmt.Sprint(tags)
-}
-
-func BenchmarkFields(b *testing.B) {
-	for n := 0; n < b.N; n++ {
-		var mt, _ = New("test_metric",
-			map[string]string{
-				"test_tag_1": "tag_value_1",
-				"test_tag_2": "tag_value_2",
-				"test_tag_3": "tag_value_3",
-			},
-			map[string]interface{}{
-				"string_field": "string",
-				"int_field":    int64(1000),
-				"float_field":  float64(2.1),
-			},
-			time.Now(),
-		)
-		fields = mt.Fields()
-	}
-	s = fmt.Sprint(fields)
-}
-
-func BenchmarkString(b *testing.B) {
-	mt, _ := New("test_metric",
-		map[string]string{
-			"test_tag_1": "tag_value_1",
-			"test_tag_2": "tag_value_2",
-			"test_tag_3": "tag_value_3",
-		},
-		map[string]interface{}{
-			"string_field": "string",
-			"int_field":    int64(1000),
-			"float_field":  float64(2.1),
-		},
-		time.Now(),
-	)
-	var S string
-	for n := 0; n < b.N; n++ {
-		S = mt.String()
-	}
-	s = S
-}
-
-func BenchmarkSerialize(b *testing.B) {
-	mt, _ := New("test_metric",
-		map[string]string{
-			"test_tag_1": "tag_value_1",
-			"test_tag_2": "tag_value_2",
-			"test_tag_3": "tag_value_3",
-		},
-		map[string]interface{}{
-			"string_field": "string",
-			"int_field":    int64(1000),
-			"float_field":  float64(2.1),
-		},
-		time.Now(),
-	)
-	var B []byte
-	for n := 0; n < b.N; n++ {
-		B = mt.Serialize()
-	}
-	s = string(B)
-}

metric/metric_test.go

@@ -1,15 +1,12 @@
 package metric
 
 import (
-	"fmt"
-	"math"
-	"regexp"
 	"testing"
 	"time"
 
 	"github.com/influxdata/telegraf"
-
 	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 )
 
 func TestNewMetric(t *testing.T) {
@@ -24,102 +21,185 @@ func TestNewMetric(t *testing.T) {
 		"usage_busy": float64(1),
 	}
 	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
+	require.NoError(t, err)
 
-	assert.Equal(t, telegraf.Untyped, m.Type())
-	assert.Equal(t, tags, m.Tags())
-	assert.Equal(t, fields, m.Fields())
-	assert.Equal(t, "cpu", m.Name())
-	assert.Equal(t, now, m.Time())
-	assert.Equal(t, now.UnixNano(), m.UnixNano())
+	require.Equal(t, "cpu", m.Name())
+	require.Equal(t, tags, m.Tags())
+	require.Equal(t, fields, m.Fields())
+	require.Equal(t, 2, len(m.FieldList()))
+	require.Equal(t, now, m.Time())
 }
 
-func TestNewErrors(t *testing.T) {
-	// creating a metric with an empty name produces an error:
-	m, err := New(
-		"",
-		map[string]string{
-			"datacenter": "us-east-1",
-			"mytag":      "foo",
-			"another":    "tag",
-		},
-		map[string]interface{}{
-			"value": float64(1),
-		},
-		time.Now(),
-	)
-	assert.Error(t, err)
-	assert.Nil(t, m)
-
-	// creating a metric with empty fields produces an error:
-	m, err = New(
-		"foobar",
-		map[string]string{
-			"datacenter": "us-east-1",
-			"mytag":      "foo",
-			"another":    "tag",
-		},
-		map[string]interface{}{},
-		time.Now(),
-	)
-	assert.Error(t, err)
-	assert.Nil(t, m)
-}
-
-func TestNewMetric_Tags(t *testing.T) {
-	now := time.Now()
-	tags := map[string]string{
-		"host":       "localhost",
-		"datacenter": "us-east-1",
-	}
+func baseMetric() telegraf.Metric {
+	tags := map[string]string{}
 	fields := map[string]interface{}{
 		"value": float64(1),
 	}
+	now := time.Now()
+
 	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
+	if err != nil {
+		panic(err)
+	}
+	return m
+}
 
-	assert.True(t, m.HasTag("host"))
-	assert.True(t, m.HasTag("datacenter"))
+func TestHasTag(t *testing.T) {
+	m := baseMetric()
 
-	m.AddTag("newtag", "foo")
-	assert.True(t, m.HasTag("newtag"))
+	require.False(t, m.HasTag("host"))
+	m.AddTag("host", "localhost")
+	require.True(t, m.HasTag("host"))
+	m.RemoveTag("host")
+	require.False(t, m.HasTag("host"))
+}
+
+func TestAddTagOverwrites(t *testing.T) {
+	m := baseMetric()
+
+	m.AddTag("host", "localhost")
+	m.AddTag("host", "example.org")
+
+	value, ok := m.GetTag("host")
+	require.True(t, ok)
+	require.Equal(t, "example.org", value)
+	require.Equal(t, 1, len(m.TagList()))
+}
+
+func TestRemoveTagNoEffectOnMissingTags(t *testing.T) {
+	m := baseMetric()
+
+	m.RemoveTag("foo")
+	m.AddTag("a", "x")
+	m.RemoveTag("foo")
+	m.RemoveTag("bar")
+	value, ok := m.GetTag("a")
+	require.True(t, ok)
+	require.Equal(t, "x", value)
+}
+
+func TestGetTag(t *testing.T) {
+	m := baseMetric()
+
+	value, ok := m.GetTag("host")
+	require.False(t, ok)
+
+	m.AddTag("host", "localhost")
+
+	value, ok = m.GetTag("host")
+	require.True(t, ok)
+	require.Equal(t, "localhost", value)
 
 	m.RemoveTag("host")
-	assert.False(t, m.HasTag("host"))
-	assert.True(t, m.HasTag("newtag"))
-	assert.True(t, m.HasTag("datacenter"))
-
-	m.RemoveTag("datacenter")
-	assert.False(t, m.HasTag("datacenter"))
-	assert.True(t, m.HasTag("newtag"))
-	assert.Equal(t, map[string]string{"newtag": "foo"}, m.Tags())
-
-	m.RemoveTag("newtag")
-	assert.False(t, m.HasTag("newtag"))
-	assert.Equal(t, map[string]string{}, m.Tags())
-
-	assert.Equal(t, "cpu value=1 "+fmt.Sprint(now.UnixNano())+"\n", m.String())
+	value, ok = m.GetTag("host")
+	require.False(t, ok)
 }
 
-func TestSerialize(t *testing.T) {
+func TestHasField(t *testing.T) {
+	m := baseMetric()
+
+	require.False(t, m.HasField("x"))
+	m.AddField("x", 42.0)
+	require.True(t, m.HasField("x"))
+	m.RemoveTag("x")
+	require.False(t, m.HasTag("x"))
+}
+
+func TestAddFieldOverwrites(t *testing.T) {
+	m := baseMetric()
+
+	m.AddField("value", 1.0)
+	m.AddField("value", 42.0)
+
+	value, ok := m.GetField("value")
+	require.True(t, ok)
+	require.Equal(t, 42.0, value)
+}
+
+func TestAddFieldChangesType(t *testing.T) {
+	m := baseMetric()
+
+	m.AddField("value", 1.0)
+	m.AddField("value", "xyzzy")
+
+	value, ok := m.GetField("value")
+	require.True(t, ok)
+	require.Equal(t, "xyzzy", value)
+}
+
+func TestRemoveFieldNoEffectOnMissingFields(t *testing.T) {
+	m := baseMetric()
+
+	m.RemoveField("foo")
+	m.AddField("a", "x")
+	m.RemoveField("foo")
+	m.RemoveField("bar")
+	value, ok := m.GetField("a")
+	require.True(t, ok)
+	require.Equal(t, "x", value)
+}
+
+func TestGetField(t *testing.T) {
+	m := baseMetric()
+
+	value, ok := m.GetField("foo")
+	require.False(t, ok)
+
+	m.AddField("foo", "bar")
+
+	value, ok = m.GetField("foo")
+	require.True(t, ok)
+	require.Equal(t, "bar", value)
+
+	m.RemoveTag("foo")
+	value, ok = m.GetTag("foo")
+	require.False(t, ok)
+}
+
+func TestTagList_Sorted(t *testing.T) {
+	m := baseMetric()
+
+	m.AddTag("b", "y")
+	m.AddTag("c", "z")
+	m.AddTag("a", "x")
+
+	taglist := m.TagList()
+	require.Equal(t, "a", taglist[0].Key)
+	require.Equal(t, "b", taglist[1].Key)
+	require.Equal(t, "c", taglist[2].Key)
+}
+
+func TestEquals(t *testing.T) {
 	now := time.Now()
-	tags := map[string]string{
-		"datacenter": "us-east-1",
-	}
-	fields := map[string]interface{}{
-		"value": float64(1),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
+	m1, err := New("cpu",
+		map[string]string{
+			"host": "localhost",
+		},
+		map[string]interface{}{
+			"value": 42.0,
+		},
+		now,
+	)
+	require.NoError(t, err)
 
-	assert.Equal(t,
-		[]byte("cpu,datacenter=us-east-1 value=1 "+fmt.Sprint(now.UnixNano())+"\n"),
-		m.Serialize())
+	m2, err := New("cpu",
+		map[string]string{
+			"host": "localhost",
+		},
+		map[string]interface{}{
+			"value": 42.0,
+		},
+		now,
+	)
+	require.NoError(t, err)
 
-	m.RemoveTag("datacenter")
-	assert.Equal(t,
-		[]byte("cpu value=1 "+fmt.Sprint(now.UnixNano())+"\n"),
-		m.Serialize())
+	lhs := m1.(*metric)
+	require.Equal(t, lhs, m2)
+
+	m3 := m2.Copy()
+	require.Equal(t, lhs, m3)
+	m3.AddTag("a", "x")
+	require.NotEqual(t, lhs, m3)
 }
 
 func TestHashID(t *testing.T) {
@@ -170,477 +250,88 @@ func TestHashID_Consistency(t *testing.T) {
 	)
 	hash := m.HashID()
 
-	for i := 0; i < 1000; i++ {
-		m2, _ := New(
-			"cpu",
-			map[string]string{
-				"datacenter": "us-east-1",
-				"mytag":      "foo",
-				"another":    "tag",
-			},
-			map[string]interface{}{
-				"value": float64(1),
-			},
-			time.Now(),
-		)
-		assert.Equal(t, hash, m2.HashID())
-	}
+	m2, _ := New(
+		"cpu",
+		map[string]string{
+			"datacenter": "us-east-1",
+			"mytag":      "foo",
+			"another":    "tag",
+		},
+		map[string]interface{}{
+			"value": float64(1),
+		},
+		time.Now(),
+	)
+	assert.Equal(t, hash, m2.HashID())
+
+	m3 := m.Copy()
+	assert.Equal(t, m2.HashID(), m3.HashID())
 }
 
-func TestNewMetric_NameModifiers(t *testing.T) {
+func TestHashID_Delimiting(t *testing.T) {
+	m1, _ := New(
+		"cpu",
+		map[string]string{
+			"a": "x",
+			"b": "y",
+			"c": "z",
+		},
+		map[string]interface{}{
+			"value": float64(1),
+		},
+		time.Now(),
+	)
+	m2, _ := New(
+		"cpu",
+		map[string]string{
+			"a": "xbycz",
+		},
+		map[string]interface{}{
+			"value": float64(1),
+		},
+		time.Now(),
+	)
+	assert.NotEqual(t, m1.HashID(), m2.HashID())
+}
+
+func TestSetName(t *testing.T) {
+	m := baseMetric()
+	m.SetName("foo")
+	require.Equal(t, "foo", m.Name())
+}
+
+func TestAddPrefix(t *testing.T) {
+	m := baseMetric()
+	m.AddPrefix("foo_")
+	require.Equal(t, "foo_cpu", m.Name())
+	m.AddPrefix("foo_")
+	require.Equal(t, "foo_foo_cpu", m.Name())
+}
+
+func TestAddSuffix(t *testing.T) {
+	m := baseMetric()
+	m.AddSuffix("_foo")
+	require.Equal(t, "cpu_foo", m.Name())
+	m.AddSuffix("_foo")
+	require.Equal(t, "cpu_foo_foo", m.Name())
+}
+
+func TestValueType(t *testing.T) {
 	now := time.Now()
+
 	tags := map[string]string{}
 	fields := map[string]interface{}{
-		"value": float64(1),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	hash := m.HashID()
-	suffix := fmt.Sprintf(" value=1 %d\n", now.UnixNano())
-	assert.Equal(t, "cpu"+suffix, m.String())
-
-	m.SetPrefix("pre_")
-	assert.NotEqual(t, hash, m.HashID())
-	hash = m.HashID()
-	assert.Equal(t, "pre_cpu"+suffix, m.String())
-
-	m.SetSuffix("_post")
-	assert.NotEqual(t, hash, m.HashID())
-	hash = m.HashID()
-	assert.Equal(t, "pre_cpu_post"+suffix, m.String())
-
-	m.SetName("mem")
-	assert.NotEqual(t, hash, m.HashID())
-	assert.Equal(t, "mem"+suffix, m.String())
-}
-
-func TestNewMetric_FieldModifiers(t *testing.T) {
-	now := time.Now()
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"value": float64(1),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	assert.True(t, m.HasField("value"))
-	assert.False(t, m.HasField("foo"))
-
-	m.AddField("newfield", "foo")
-	assert.True(t, m.HasField("newfield"))
-
-	assert.NoError(t, m.RemoveField("newfield"))
-	assert.False(t, m.HasField("newfield"))
-
-	// don't allow user to remove all fields:
-	assert.Error(t, m.RemoveField("value"))
-
-	m.AddField("value2", int64(101))
-	assert.NoError(t, m.RemoveField("value"))
-	assert.False(t, m.HasField("value"))
-}
-
-func TestNewMetric_Fields(t *testing.T) {
-	now := time.Now()
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"float":  float64(1),
-		"int":    int64(1),
-		"bool":   true,
-		"false":  false,
-		"string": "test",
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	assert.Equal(t, fields, m.Fields())
-}
-
-func TestNewMetric_Time(t *testing.T) {
-	now := time.Now()
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"float":  float64(1),
-		"int":    int64(1),
-		"bool":   true,
-		"false":  false,
-		"string": "test",
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-	m = m.Copy()
-	m2 := m.Copy()
-
-	assert.Equal(t, now.UnixNano(), m.Time().UnixNano())
-	assert.Equal(t, now.UnixNano(), m2.UnixNano())
-}
-
-func TestNewMetric_Copy(t *testing.T) {
-	now := time.Now()
-	tags := map[string]string{}
-	fields := map[string]interface{}{
-		"float": float64(1),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-	m2 := m.Copy()
-
-	assert.Equal(t,
-		fmt.Sprintf("cpu float=1 %d\n", now.UnixNano()),
-		m.String())
-	m.AddTag("host", "localhost")
-	assert.Equal(t,
-		fmt.Sprintf("cpu,host=localhost float=1 %d\n", now.UnixNano()),
-		m.String())
-
-	assert.Equal(t,
-		fmt.Sprintf("cpu float=1 %d\n", now.UnixNano()),
-		m2.String())
-}
-
-func TestNewMetric_AllTypes(t *testing.T) {
-	now := time.Now()
-	tags := map[string]string{}
-	fields := map[string]interface{}{
-		"float64":     float64(1),
-		"float32":     float32(1),
-		"int64":       int64(1),
-		"int32":       int32(1),
-		"int16":       int16(1),
-		"int8":        int8(1),
-		"int":         int(1),
-		"uint64":      uint64(1),
-		"uint32":      uint32(1),
-		"uint16":      uint16(1),
-		"uint8":       uint8(1),
-		"uint":        uint(1),
-		"bytes":       []byte("foo"),
-		"nil":         nil,
-		"maxuint64":   uint64(MaxInt) + 10,
-		"maxuint":     uint(MaxInt) + 10,
-		"unsupported": []int{1, 2},
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	assert.Contains(t, m.String(), "float64=1")
-	assert.Contains(t, m.String(), "float32=1")
-	assert.Contains(t, m.String(), "int64=1i")
-	assert.Contains(t, m.String(), "int32=1i")
-	assert.Contains(t, m.String(), "int16=1i")
-	assert.Contains(t, m.String(), "int8=1i")
-	assert.Contains(t, m.String(), "int=1i")
-	assert.Contains(t, m.String(), "uint64=1i")
-	assert.Contains(t, m.String(), "uint32=1i")
-	assert.Contains(t, m.String(), "uint16=1i")
-	assert.Contains(t, m.String(), "uint8=1i")
-	assert.Contains(t, m.String(), "uint=1i")
-	assert.NotContains(t, m.String(), "nil")
-	assert.Contains(t, m.String(), fmt.Sprintf("maxuint64=%di", MaxInt))
-	assert.Contains(t, m.String(), fmt.Sprintf("maxuint=%di", MaxInt))
-}
-
-func TestIndexUnescapedByte(t *testing.T) {
-	tests := []struct {
-		in       []byte
-		b        byte
-		expected int
-	}{
-		{
-			in:       []byte(`foobar`),
-			b:        'b',
-			expected: 3,
-		},
-		{
-			in:       []byte(`foo\bar`),
-			b:        'b',
-			expected: -1,
-		},
-		{
-			in:       []byte(`foo\\bar`),
-			b:        'b',
-			expected: 5,
-		},
-		{
-			in:       []byte(`foobar`),
-			b:        'f',
-			expected: 0,
-		},
-		{
-			in:       []byte(`foobar`),
-			b:        'r',
-			expected: 5,
-		},
-		{
-			in:       []byte(`\foobar`),
-			b:        'f',
-			expected: -1,
-		},
-	}
-
-	for _, test := range tests {
-		got := indexUnescapedByte(test.in, test.b)
-		assert.Equal(t, test.expected, got)
-	}
-}
-
-func TestNewGaugeMetric(t *testing.T) {
-	now := time.Now()
-
-	tags := map[string]string{
-		"host":       "localhost",
-		"datacenter": "us-east-1",
-	}
-	fields := map[string]interface{}{
-		"usage_idle": float64(99),
-		"usage_busy": float64(1),
+		"value": float64(42),
 	}
 	m, err := New("cpu", tags, fields, now, telegraf.Gauge)
 	assert.NoError(t, err)
 
 	assert.Equal(t, telegraf.Gauge, m.Type())
-	assert.Equal(t, tags, m.Tags())
-	assert.Equal(t, fields, m.Fields())
-	assert.Equal(t, "cpu", m.Name())
-	assert.Equal(t, now, m.Time())
-	assert.Equal(t, now.UnixNano(), m.UnixNano())
 }
 
-func TestNewCounterMetric(t *testing.T) {
-	now := time.Now()
-
-	tags := map[string]string{
-		"host":       "localhost",
-		"datacenter": "us-east-1",
-	}
-	fields := map[string]interface{}{
-		"usage_idle": float64(99),
-		"usage_busy": float64(1),
-	}
-	m, err := New("cpu", tags, fields, now, telegraf.Counter)
-	assert.NoError(t, err)
-
-	assert.Equal(t, telegraf.Counter, m.Type())
-	assert.Equal(t, tags, m.Tags())
-	assert.Equal(t, fields, m.Fields())
-	assert.Equal(t, "cpu", m.Name())
-	assert.Equal(t, now, m.Time())
-	assert.Equal(t, now.UnixNano(), m.UnixNano())
-}
-
-// test splitting metric into various max lengths
-func TestSplitMetric(t *testing.T) {
-	now := time.Unix(0, 1480940990034083306)
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"float":  float64(100001),
-		"int":    int64(100001),
-		"bool":   true,
-		"false":  false,
-		"string": "test",
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	split80 := m.Split(80)
-	assert.Len(t, split80, 2)
-
-	split70 := m.Split(70)
-	assert.Len(t, split70, 3)
-
-	split60 := m.Split(60)
-	assert.Len(t, split60, 4)
-}
-
-// test splitting metric into various max lengths
-// use a simple regex check to verify that the split metrics are valid
-func TestSplitMetric_RegexVerify(t *testing.T) {
-	now := time.Unix(0, 1480940990034083306)
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"foo":     float64(98934259085),
-		"bar":     float64(19385292),
-		"number":  float64(19385292),
-		"another": float64(19385292),
-		"n":       float64(19385292),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	// verification regex
-	re := regexp.MustCompile(`cpu,host=localhost \w+=\d+(,\w+=\d+)* 1480940990034083306`)
-
-	split90 := m.Split(90)
-	assert.Len(t, split90, 2)
-	for _, splitM := range split90 {
-		assert.True(t, re.Match(splitM.Serialize()), splitM.String())
-	}
-
-	split70 := m.Split(70)
-	assert.Len(t, split70, 3)
-	for _, splitM := range split70 {
-		assert.True(t, re.Match(splitM.Serialize()), splitM.String())
-	}
-
-	split20 := m.Split(20)
-	assert.Len(t, split20, 5)
-	for _, splitM := range split20 {
-		assert.True(t, re.Match(splitM.Serialize()), splitM.String())
-	}
-}
-
-// test splitting metric even when given length is shorter than
-// shortest possible length
-// Split should split metric as short as possible, ie, 1 field per metric
-func TestSplitMetric_TooShort(t *testing.T) {
-	now := time.Unix(0, 1480940990034083306)
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"float":  float64(100001),
-		"int":    int64(100001),
-		"bool":   true,
-		"false":  false,
-		"string": "test",
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	split := m.Split(10)
-	assert.Len(t, split, 5)
-	strings := make([]string, 5)
-	for i, splitM := range split {
-		strings[i] = splitM.String()
-	}
-
-	assert.Contains(t, strings, "cpu,host=localhost float=100001 1480940990034083306\n")
-	assert.Contains(t, strings, "cpu,host=localhost int=100001i 1480940990034083306\n")
-	assert.Contains(t, strings, "cpu,host=localhost bool=true 1480940990034083306\n")
-	assert.Contains(t, strings, "cpu,host=localhost false=false 1480940990034083306\n")
-	assert.Contains(t, strings, "cpu,host=localhost string=\"test\" 1480940990034083306\n")
-}
-
-func TestSplitMetric_NoOp(t *testing.T) {
-	now := time.Unix(0, 1480940990034083306)
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"float":  float64(100001),
-		"int":    int64(100001),
-		"bool":   true,
-		"false":  false,
-		"string": "test",
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	split := m.Split(1000)
-	assert.Len(t, split, 1)
-	assert.Equal(t, m, split[0])
-}
-
-func TestSplitMetric_OneField(t *testing.T) {
-	now := time.Unix(0, 1480940990034083306)
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"float": float64(100001),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	assert.Equal(t, "cpu,host=localhost float=100001 1480940990034083306\n", m.String())
-
-	split := m.Split(1000)
-	assert.Len(t, split, 1)
-	assert.Equal(t, "cpu,host=localhost float=100001 1480940990034083306\n", split[0].String())
-
-	split = m.Split(1)
-	assert.Len(t, split, 1)
-	assert.Equal(t, "cpu,host=localhost float=100001 1480940990034083306\n", split[0].String())
-
-	split = m.Split(40)
-	assert.Len(t, split, 1)
-	assert.Equal(t, "cpu,host=localhost float=100001 1480940990034083306\n", split[0].String())
-}
-
-func TestNewMetricAggregate(t *testing.T) {
-	now := time.Now()
-
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"usage_idle": float64(99),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	assert.False(t, m.IsAggregate())
-	m.SetAggregate(true)
-	assert.True(t, m.IsAggregate())
-}
-
-func TestNewMetricPoint(t *testing.T) {
-	now := time.Now()
-
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"usage_idle": float64(99),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	p := m.Point()
-
-	assert.Equal(t, fields, m.Fields())
-	assert.Equal(t, fields, p.Fields())
-	assert.Equal(t, "cpu", p.Name())
-}
-
-func TestNewMetricString(t *testing.T) {
-	now := time.Now()
-
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"usage_idle": float64(99),
-	}
-	m, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
-
-	lineProto := fmt.Sprintf("cpu,host=localhost usage_idle=99 %d\n",
-		now.UnixNano())
-	assert.Equal(t, lineProto, m.String())
-}
-
-func TestNewMetricFailNaN(t *testing.T) {
-	now := time.Now()
-
-	tags := map[string]string{
-		"host": "localhost",
-	}
-	fields := map[string]interface{}{
-		"usage_idle": math.NaN(),
-	}
-
-	_, err := New("cpu", tags, fields, now)
-	assert.NoError(t, err)
+func TestCopyAggreate(t *testing.T) {
+	m1 := baseMetric()
+	m1.SetAggregate(true)
+	m2 := m1.Copy()
+	assert.True(t, m2.IsAggregate())
 }

metric/parse.go

@@ -1,627 +0,0 @@
-package metric
-
-import (
-	"bytes"
-	"errors"
-	"fmt"
-	"time"
-
-	"github.com/influxdata/telegraf"
-)
-
-var (
-	ErrInvalidNumber = errors.New("invalid number")
-)
-
-const (
-	// the number of characters for the largest possible int64 (9223372036854775807)
-	maxInt64Digits = 19
-
-	// the number of characters for the smallest possible int64 (-9223372036854775808)
-	minInt64Digits = 20
-
-	// the number of characters required for the largest float64 before a range check
-	// would occur during parsing
-	maxFloat64Digits = 25
-
-	// the number of characters required for smallest float64 before a range check occur
-	// would occur during parsing
-	minFloat64Digits = 27
-
-	MaxKeyLength = 65535
-)
-
-// The following constants allow us to specify which state to move to
-// next, when scanning sections of a Point.
-const (
-	tagKeyState = iota
-	tagValueState
-	fieldsState
-)
-
-func Parse(buf []byte) ([]telegraf.Metric, error) {
-	return ParseWithDefaultTime(buf, time.Now())
-}
-
-func ParseWithDefaultTime(buf []byte, t time.Time) ([]telegraf.Metric, error) {
-	if len(buf) <= 6 {
-		return []telegraf.Metric{}, makeError("buffer too short", buf, 0)
-	}
-	metrics := make([]telegraf.Metric, 0, bytes.Count(buf, []byte("\n"))+1)
-	var errStr string
-	i := 0
-	for {
-		j := bytes.IndexByte(buf[i:], '\n')
-		if j == -1 {
-			break
-		}
-		if len(buf[i:i+j]) < 2 {
-			i += j + 1 // increment i past the previous newline
-			continue
-		}
-
-		m, err := parseMetric(buf[i:i+j], t)
-		if err != nil {
-			i += j + 1 // increment i past the previous newline
-			errStr += " " + err.Error()
-			continue
-		}
-		i += j + 1 // increment i past the previous newline
-
-		metrics = append(metrics, m)
-	}
-
-	if len(errStr) > 0 {
-		return metrics, fmt.Errorf(errStr)
-	}
-	return metrics, nil
-}
-
-func parseMetric(buf []byte, defaultTime time.Time) (telegraf.Metric, error) {
-	var dTime string
-	// scan the first block which is measurement[,tag1=value1,tag2=value=2...]
-	pos, key, err := scanKey(buf, 0)
-	if err != nil {
-		return nil, err
-	}
-
-	// measurement name is required
-	if len(key) == 0 {
-		return nil, fmt.Errorf("missing measurement")
-	}
-
-	if len(key) > MaxKeyLength {
-		return nil, fmt.Errorf("max key length exceeded: %v > %v", len(key), MaxKeyLength)
-	}
-
-	// scan the second block is which is field1=value1[,field2=value2,...]
-	pos, fields, err := scanFields(buf, pos)
-	if err != nil {
-		return nil, err
-	}
-
-	// at least one field is required
-	if len(fields) == 0 {
-		return nil, fmt.Errorf("missing fields")
-	}
-
-	// scan the last block which is an optional integer timestamp
-	pos, ts, err := scanTime(buf, pos)
-	if err != nil {
-		return nil, err
-	}
-
-	m := &metric{
-		fields: fields,
-		t:      ts,
-	}
-
-	// parse out the measurement name
-	// namei is the index at which the "name" ends
-	namei := indexUnescapedByte(key, ',')
-	if namei < 1 {
-		// no tags
-		m.name = key
-	} else {
-		m.name = key[0:namei]
-		m.tags = key[namei:]
-	}
-
-	if len(m.t) == 0 {
-		if len(dTime) == 0 {
-			dTime = fmt.Sprint(defaultTime.UnixNano())
-		}
-		// use default time
-		m.t = []byte(dTime)
-	}
-
-	// here we copy on return because this allows us to later call
-	// AddTag, AddField, RemoveTag, RemoveField, etc. without worrying about
-	// modifying 'tag' bytes having an affect on 'field' bytes, for example.
-	return m.Copy(), nil
-}
-
-// scanKey scans buf starting at i for the measurement and tag portion of the point.
-// It returns the ending position and the byte slice of key within buf.  If there
-// are tags, they will be sorted if they are not already.
-func scanKey(buf []byte, i int) (int, []byte, error) {
-	start := skipWhitespace(buf, i)
-	i = start
-
-	// First scan the Point's measurement.
-	state, i, err := scanMeasurement(buf, i)
-	if err != nil {
-		return i, buf[start:i], err
-	}
-
-	// Optionally scan tags if needed.
-	if state == tagKeyState {
-		i, err = scanTags(buf, i)
-		if err != nil {
-			return i, buf[start:i], err
-		}
-	}
-
-	return i, buf[start:i], nil
-}
-
-// scanMeasurement examines the measurement part of a Point, returning
-// the next state to move to, and the current location in the buffer.
-func scanMeasurement(buf []byte, i int) (int, int, error) {
-	// Check first byte of measurement, anything except a comma is fine.
-	// It can't be a space, since whitespace is stripped prior to this
-	// function call.
-	if i >= len(buf) || buf[i] == ',' {
-		return -1, i, makeError("missing measurement", buf, i)
-	}
-
-	for {
-		i++
-		if i >= len(buf) {
-			// cpu
-			return -1, i, makeError("missing fields", buf, i)
-		}
-
-		if buf[i-1] == '\\' {
-			// Skip character (it's escaped).
-			continue
-		}
-
-		// Unescaped comma; move onto scanning the tags.
-		if buf[i] == ',' {
-			return tagKeyState, i + 1, nil
-		}
-
-		// Unescaped space; move onto scanning the fields.
-		if buf[i] == ' ' {
-			// cpu value=1.0
-			return fieldsState, i, nil
-		}
-	}
-}
-
-// scanTags examines all the tags in a Point, keeping track of and
-// returning the updated indices slice, number of commas and location
-// in buf where to start examining the Point fields.
-func scanTags(buf []byte, i int) (int, error) {
-	var (
-		err   error
-		state = tagKeyState
-	)
-
-	for {
-		switch state {
-		case tagKeyState:
-			i, err = scanTagsKey(buf, i)
-			state = tagValueState // tag value always follows a tag key
-		case tagValueState:
-			state, i, err = scanTagsValue(buf, i)
-		case fieldsState:
-			return i, nil
-		}
-
-		if err != nil {
-			return i, err
-		}
-	}
-}
-
-// scanTagsKey scans each character in a tag key.
-func scanTagsKey(buf []byte, i int) (int, error) {
-	// First character of the key.
-	if i >= len(buf) || buf[i] == ' ' || buf[i] == ',' || buf[i] == '=' {
-		// cpu,{'', ' ', ',', '='}
-		return i, makeError("missing tag key", buf, i)
-	}
-
-	// Examine each character in the tag key until we hit an unescaped
-	// equals (the tag value), or we hit an error (i.e., unescaped
-	// space or comma).
-	for {
-		i++
-
-		// Either we reached the end of the buffer or we hit an
-		// unescaped comma or space.
-		if i >= len(buf) ||
-			((buf[i] == ' ' || buf[i] == ',') && buf[i-1] != '\\') {
-			// cpu,tag{'', ' ', ','}
-			return i, makeError("missing tag value", buf, i)
-		}
-
-		if buf[i] == '=' && buf[i-1] != '\\' {
-			// cpu,tag=
-			return i + 1, nil
-		}
-	}
-}
-
-// scanTagsValue scans each character in a tag value.
-func scanTagsValue(buf []byte, i int) (int, int, error) {
-	// Tag value cannot be empty.
-	if i >= len(buf) || buf[i] == ',' || buf[i] == ' ' {
-		// cpu,tag={',', ' '}
-		return -1, i, makeError("missing tag value", buf, i)
-	}
-
-	// Examine each character in the tag value until we hit an unescaped
-	// comma (move onto next tag key), an unescaped space (move onto
-	// fields), or we error out.
-	for {
-		i++
-		if i >= len(buf) {
-			// cpu,tag=value
-			return -1, i, makeError("missing fields", buf, i)
-		}
-
-		// An unescaped equals sign is an invalid tag value.
-		if buf[i] == '=' && buf[i-1] != '\\' {
-			// cpu,tag={'=', 'fo=o'}
-			return -1, i, makeError("invalid tag format", buf, i)
-		}
-
-		if buf[i] == ',' && buf[i-1] != '\\' {
-			// cpu,tag=foo,
-			return tagKeyState, i + 1, nil
-		}
-
-		// cpu,tag=foo value=1.0
-		// cpu, tag=foo\= value=1.0
-		if buf[i] == ' ' && buf[i-1] != '\\' {
-			return fieldsState, i, nil
-		}
-	}
-}
-
-// scanFields scans buf, starting at i for the fields section of a point.  It returns
-// the ending position and the byte slice of the fields within buf
-func scanFields(buf []byte, i int) (int, []byte, error) {
-	start := skipWhitespace(buf, i)
-	i = start
-	quoted := false
-
-	// tracks how many '=' we've seen
-	equals := 0
-
-	// tracks how many commas we've seen
-	commas := 0
-
-	for {
-		// reached the end of buf?
-		if i >= len(buf) {
-			break
-		}
-
-		// escaped characters?
-		if buf[i] == '\\' && i+1 < len(buf) {
-			i += 2
-			continue
-		}
-
-		// If the value is quoted, scan until we get to the end quote
-		// Only quote values in the field value since quotes are not significant
-		// in the field key
-		if buf[i] == '"' && equals > commas {
-			quoted = !quoted
-			i++
-			continue
-		}
-
-		// If we see an =, ensure that there is at least on char before and after it
-		if buf[i] == '=' && !quoted {
-			equals++
-
-			// check for "... =123" but allow "a\ =123"
-			if buf[i-1] == ' ' && buf[i-2] != '\\' {
-				return i, buf[start:i], makeError("missing field key", buf, i)
-			}
-
-			// check for "...a=123,=456" but allow "a=123,a\,=456"
-			if buf[i-1] == ',' && buf[i-2] != '\\' {
-				return i, buf[start:i], makeError("missing field key", buf, i)
-			}
-
-			// check for "... value="
-			if i+1 >= len(buf) {
-				return i, buf[start:i], makeError("missing field value", buf, i)
-			}
-
-			// check for "... value=,value2=..."
-			if buf[i+1] == ',' || buf[i+1] == ' ' {
-				return i, buf[start:i], makeError("missing field value", buf, i)
-			}
-
-			if isNumeric(buf[i+1]) || buf[i+1] == '-' || buf[i+1] == 'N' || buf[i+1] == 'n' {
-				var err error
-				i, err = scanNumber(buf, i+1)
-				if err != nil {
-					return i, buf[start:i], err
-				}
-				continue
-			}
-			// If next byte is not a double-quote, the value must be a boolean
-			if buf[i+1] != '"' {
-				var err error
-				i, _, err = scanBoolean(buf, i+1)
-				if err != nil {
-					return i, buf[start:i], err
-				}
-				continue
-			}
-		}
-
-		if buf[i] == ',' && !quoted {
-			commas++
-		}
-
-		// reached end of block?
-		if buf[i] == ' ' && !quoted {
-			break
-		}
-		i++
-	}
-
-	if quoted {
-		return i, buf[start:i], makeError("unbalanced quotes", buf, i)
-	}
-
-	// check that all field sections had key and values (e.g. prevent "a=1,b"
-	if equals == 0 || commas != equals-1 {
-		return i, buf[start:i], makeError("invalid field format", buf, i)
-	}
-
-	return i, buf[start:i], nil
-}
-
-// scanTime scans buf, starting at i for the time section of a point. It
-// returns the ending position and the byte slice of the timestamp within buf
-// and and error if the timestamp is not in the correct numeric format.
-func scanTime(buf []byte, i int) (int, []byte, error) {
-	start := skipWhitespace(buf, i)
-	i = start
-
-	for {
-		// reached the end of buf?
-		if i >= len(buf) {
-			break
-		}
-
-		// Reached end of block or trailing whitespace?
-		if buf[i] == '\n' || buf[i] == ' ' {
-			break
-		}
-
-		// Handle negative timestamps
-		if i == start && buf[i] == '-' {
-			i++
-			continue
-		}
-
-		// Timestamps should be integers, make sure they are so we don't need
-		// to actually  parse the timestamp until needed.
-		if buf[i] < '0' || buf[i] > '9' {
-			return i, buf[start:i], makeError("invalid timestamp", buf, i)
-		}
-		i++
-	}
-	return i, buf[start:i], nil
-}
-
-func isNumeric(b byte) bool {
-	return (b >= '0' && b <= '9') || b == '.'
-}
-
-// scanNumber returns the end position within buf, start at i after
-// scanning over buf for an integer, or float.  It returns an
-// error if a invalid number is scanned.
-func scanNumber(buf []byte, i int) (int, error) {
-	start := i
-	var isInt bool
-
-	// Is negative number?
-	if i < len(buf) && buf[i] == '-' {
-		i++
-		// There must be more characters now, as just '-' is illegal.
-		if i == len(buf) {
-			return i, ErrInvalidNumber
-		}
-	}
-
-	// how many decimal points we've see
-	decimal := false
-
-	// indicates the number is float in scientific notation
-	scientific := false
-
-	for {
-		if i >= len(buf) {
-			break
-		}
-
-		if buf[i] == ',' || buf[i] == ' ' {
-			break
-		}
-
-		if buf[i] == 'i' && i > start && !isInt {
-			isInt = true
-			i++
-			continue
-		}
-
-		if buf[i] == '.' {
-			// Can't have more than 1 decimal (e.g. 1.1.1 should fail)
-			if decimal {
-				return i, ErrInvalidNumber
-			}
-			decimal = true
-		}
-
-		// `e` is valid for floats but not as the first char
-		if i > start && (buf[i] == 'e' || buf[i] == 'E') {
-			scientific = true
-			i++
-			continue
-		}
-
-		// + and - are only valid at this point if they follow an e (scientific notation)
-		if (buf[i] == '+' || buf[i] == '-') && (buf[i-1] == 'e' || buf[i-1] == 'E') {
-			i++
-			continue
-		}
-
-		// NaN is an unsupported value
-		if i+2 < len(buf) && (buf[i] == 'N' || buf[i] == 'n') {
-			return i, ErrInvalidNumber
-		}
-
-		if !isNumeric(buf[i]) {
-			return i, ErrInvalidNumber
-		}
-		i++
-	}
-
-	if isInt && (decimal || scientific) {
-		return i, ErrInvalidNumber
-	}
-
-	numericDigits := i - start
-	if isInt {
-		numericDigits--
-	}
-	if decimal {
-		numericDigits--
-	}
-	if buf[start] == '-' {
-		numericDigits--
-	}
-
-	if numericDigits == 0 {
-		return i, ErrInvalidNumber
-	}
-
-	// It's more common that numbers will be within min/max range for their type but we need to prevent
-	// out or range numbers from being parsed successfully.  This uses some simple heuristics to decide
-	// if we should parse the number to the actual type.  It does not do it all the time because it incurs
-	// extra allocations and we end up converting the type again when writing points to disk.
-	if isInt {
-		// Make sure the last char is an 'i' for integers (e.g. 9i10 is not valid)
-		if buf[i-1] != 'i' {
-			return i, ErrInvalidNumber
-		}
-		// Parse the int to check bounds the number of digits could be larger than the max range
-		// We subtract 1 from the index to remove the `i` from our tests
-		if len(buf[start:i-1]) >= maxInt64Digits || len(buf[start:i-1]) >= minInt64Digits {
-			if _, err := parseIntBytes(buf[start:i-1], 10, 64); err != nil {
-				return i, makeError(fmt.Sprintf("unable to parse integer %s: %s", buf[start:i-1], err), buf, i)
-			}
-		}
-	} else {
-		// Parse the float to check bounds if it's scientific or the number of digits could be larger than the max range
-		if scientific || len(buf[start:i]) >= maxFloat64Digits || len(buf[start:i]) >= minFloat64Digits {
-			if _, err := parseFloatBytes(buf[start:i], 10); err != nil {
-				return i, makeError("invalid float", buf, i)
-			}
-		}
-	}
-
-	return i, nil
-}
-
-// scanBoolean returns the end position within buf, start at i after
-// scanning over buf for boolean. Valid values for a boolean are
-// t, T, true, TRUE, f, F, false, FALSE. It returns an error if a invalid boolean
-// is scanned.
-func scanBoolean(buf []byte, i int) (int, []byte, error) {
-	start := i
-
-	if i < len(buf) && (buf[i] != 't' && buf[i] != 'f' && buf[i] != 'T' && buf[i] != 'F') {
-		return i, buf[start:i], makeError("invalid value", buf, i)
-	}
-
-	i++
-	for {
-		if i >= len(buf) {
-			break
-		}
-
-		if buf[i] == ',' || buf[i] == ' ' {
-			break
-		}
-		i++
-	}
-
-	// Single char bool (t, T, f, F) is ok
-	if i-start == 1 {
-		return i, buf[start:i], nil
-	}
-
-	// length must be 4 for true or TRUE
-	if (buf[start] == 't' || buf[start] == 'T') && i-start != 4 {
-		return i, buf[start:i], makeError("invalid boolean", buf, i)
-	}
-
-	// length must be 5 for false or FALSE
-	if (buf[start] == 'f' || buf[start] == 'F') && i-start != 5 {
-		return i, buf[start:i], makeError("invalid boolean", buf, i)
-	}
-
-	// Otherwise
-	valid := false
-	switch buf[start] {
-	case 't':
-		valid = bytes.Equal(buf[start:i], []byte("true"))
-	case 'f':
-		valid = bytes.Equal(buf[start:i], []byte("false"))
-	case 'T':
-		valid = bytes.Equal(buf[start:i], []byte("TRUE")) || bytes.Equal(buf[start:i], []byte("True"))
-	case 'F':
-		valid = bytes.Equal(buf[start:i], []byte("FALSE")) || bytes.Equal(buf[start:i], []byte("False"))
-	}
-
-	if !valid {
-		return i, buf[start:i], makeError("invalid boolean", buf, i)
-	}
-
-	return i, buf[start:i], nil
-
-}
-
-// skipWhitespace returns the end position within buf, starting at i after
-// scanning over spaces in tags
-func skipWhitespace(buf []byte, i int) int {
-	for i < len(buf) {
-		if buf[i] != ' ' && buf[i] != '\t' && buf[i] != 0 {
-			break
-		}
-		i++
-	}
-	return i
-}
-
-// makeError is a helper function for making a metric parsing error.
-//   reason is the reason that the error occured.
-//   buf should be the current buffer we are parsing.
-//   i is the current index, to give some context on where in the buffer we are.
-func makeError(reason string, buf []byte, i int) error {
-	return fmt.Errorf("metric parsing error, reason: [%s], buffer: [%s], index: [%d]",
-		reason, buf, i)
-}

metric/parse_test.go

@@ -1,355 +0,0 @@
-package metric
-
-import (
-	"testing"
-	"time"
-
-	"github.com/stretchr/testify/assert"
-)
-
-const trues = `booltest b=T
-booltest b=t
-booltest b=True
-booltest b=TRUE
-booltest b=true
-`
-
-const falses = `booltest b=F
-booltest b=f
-booltest b=False
-booltest b=FALSE
-booltest b=false
-`
-
-const withEscapes = `w\,\ eather,host=local temp=99 1465839830100400200
-w\,eather,host=local temp=99 1465839830100400200
-weather,location=us\,midwest temperature=82 1465839830100400200
-weather,location=us-midwest temp\=rature=82 1465839830100400200
-weather,location\ place=us-midwest temperature=82 1465839830100400200
-weather,location=us-midwest temperature="too\"hot\"" 1465839830100400200
-`
-
-const withTimestamps = `cpu usage=99 1480595849000000000
-cpu usage=99 1480595850000000000
-cpu usage=99 1480595851700030000
-cpu usage=99 1480595852000000300
-`
-
-const sevenMetrics = `cpu,host=foo,datacenter=us-east idle=99,busy=1i,b=true,s="string"
-cpu,host=foo,datacenter=us-east idle=99,busy=1i,b=true,s="string"
-cpu,host=foo,datacenter=us-east idle=99,busy=1i,b=true,s="string"
-cpu,host=foo,datacenter=us-east idle=99,busy=1i,b=true,s="string"
-cpu,host=foo,datacenter=us-east idle=99,busy=1i,b=true,s="string"
-cpu,host=foo,datacenter=us-east idle=99,busy=1i,b=true,s="string"
-cpu,host=foo,datacenter=us-east idle=99,busy=1i,b=true,s="string"
-`
-
-// some metrics are invalid
-const someInvalid = `cpu,host=foo,datacenter=us-east usage_idle=99,usage_busy=1
-cpu,host=foo,datacenter=us-east usage_idle=99,usage_busy=1
-cpu,host=foo,datacenter=us-east usage_idle=99,usage_busy=1
-cpu,cpu=cpu3, host=foo,datacenter=us-east usage_idle=99,usage_busy=1
-cpu,cpu=cpu4 , usage_idle=99,usage_busy=1
-cpu 1480595852000000300
-cpu usage=99 1480595852foobar300
-cpu,host=foo,datacenter=us-east usage_idle=99,usage_busy=1
-`
-
-func TestParse(t *testing.T) {
-	start := time.Now()
-	metrics, err := Parse([]byte(sevenMetrics))
-	assert.NoError(t, err)
-	assert.Len(t, metrics, 7)
-
-	// all metrics parsed together w/o a timestamp should have the same time.
-	firstTime := metrics[0].Time()
-	for _, m := range metrics {
-		assert.Equal(t,
-			map[string]interface{}{
-				"idle": float64(99),
-				"busy": int64(1),
-				"b":    true,
-				"s":    "string",
-			},
-			m.Fields(),
-		)
-		assert.Equal(t,
-			map[string]string{
-				"host":       "foo",
-				"datacenter": "us-east",
-			},
-			m.Tags(),
-		)
-		assert.True(t, m.Time().After(start))
-		assert.True(t, m.Time().Equal(firstTime))
-	}
-}
-
-func TestParseErrors(t *testing.T) {
-	start := time.Now()
-	metrics, err := Parse([]byte(someInvalid))
-	assert.Error(t, err)
-	assert.Len(t, metrics, 4)
-
-	// all metrics parsed together w/o a timestamp should have the same time.
-	firstTime := metrics[0].Time()
-	for _, m := range metrics {
-		assert.Equal(t,
-			map[string]interface{}{
-				"usage_idle": float64(99),
-				"usage_busy": float64(1),
-			},
-			m.Fields(),
-		)
-		assert.Equal(t,
-			map[string]string{
-				"host":       "foo",
-				"datacenter": "us-east",
-			},
-			m.Tags(),
-		)
-		assert.True(t, m.Time().After(start))
-		assert.True(t, m.Time().Equal(firstTime))
-	}
-}
-
-func TestParseWithTimestamps(t *testing.T) {
-	metrics, err := Parse([]byte(withTimestamps))
-	assert.NoError(t, err)
-	assert.Len(t, metrics, 4)
-
-	expectedTimestamps := []time.Time{
-		time.Unix(0, 1480595849000000000),
-		time.Unix(0, 1480595850000000000),
-		time.Unix(0, 1480595851700030000),
-		time.Unix(0, 1480595852000000300),
-	}
-
-	// all metrics parsed together w/o a timestamp should have the same time.
-	for i, m := range metrics {
-		assert.Equal(t,
-			map[string]interface{}{
-				"usage": float64(99),
-			},
-			m.Fields(),
-		)
-		assert.True(t, m.Time().Equal(expectedTimestamps[i]))
-	}
-}
-
-func TestParseEscapes(t *testing.T) {
-	metrics, err := Parse([]byte(withEscapes))
-	assert.NoError(t, err)
-	assert.Len(t, metrics, 6)
-
-	tests := []struct {
-		name   string
-		fields map[string]interface{}
-		tags   map[string]string
-	}{
-		{
-			name:   `w, eather`,
-			fields: map[string]interface{}{"temp": float64(99)},
-			tags:   map[string]string{"host": "local"},
-		},
-		{
-			name:   `w,eather`,
-			fields: map[string]interface{}{"temp": float64(99)},
-			tags:   map[string]string{"host": "local"},
-		},
-		{
-			name:   `weather`,
-			fields: map[string]interface{}{"temperature": float64(82)},
-			tags:   map[string]string{"location": `us,midwest`},
-		},
-		{
-			name:   `weather`,
-			fields: map[string]interface{}{`temp=rature`: float64(82)},
-			tags:   map[string]string{"location": `us-midwest`},
-		},
-		{
-			name:   `weather`,
-			fields: map[string]interface{}{"temperature": float64(82)},
-			tags:   map[string]string{`location place`: `us-midwest`},
-		},
-		{
-			name:   `weather`,
-			fields: map[string]interface{}{`temperature`: `too"hot"`},
-			tags:   map[string]string{"location": `us-midwest`},
-		},
-	}
-
-	for i, test := range tests {
-		assert.Equal(t, test.name, metrics[i].Name())
-		assert.Equal(t, test.fields, metrics[i].Fields())
-		assert.Equal(t, test.tags, metrics[i].Tags())
-	}
-}
-
-func TestParseTrueBooleans(t *testing.T) {
-	metrics, err := Parse([]byte(trues))
-	assert.NoError(t, err)
-	assert.Len(t, metrics, 5)
-
-	for _, metric := range metrics {
-		assert.Equal(t, "booltest", metric.Name())
-		assert.Equal(t, true, metric.Fields()["b"])
-	}
-}
-
-func TestParseFalseBooleans(t *testing.T) {
-	metrics, err := Parse([]byte(falses))
-	assert.NoError(t, err)
-	assert.Len(t, metrics, 5)
-
-	for _, metric := range metrics {
-		assert.Equal(t, "booltest", metric.Name())
-		assert.Equal(t, false, metric.Fields()["b"])
-	}
-}
-
-func TestParsePointBadNumber(t *testing.T) {
-	for _, tt := range []string{
-		"cpu v=- ",
-		"cpu v=-i ",
-		"cpu v=-. ",
-		"cpu v=. ",
-		"cpu v=1.0i ",
-		"cpu v=1ii ",
-		"cpu v=1a ",
-		"cpu v=-e-e-e ",
-		"cpu v=42+3 ",
-		"cpu v= ",
-	} {
-		_, err := Parse([]byte(tt + "\n"))
-		assert.Error(t, err, tt)
-	}
-}
-
-func TestParseTagsMissingParts(t *testing.T) {
-	for _, tt := range []string{
-		`cpu,host`,
-		`cpu,host,`,
-		`cpu,host=`,
-		`cpu,f=oo=bar value=1`,
-		`cpu,host value=1i`,
-		`cpu,host=serverA,region value=1i`,
-		`cpu,host=serverA,region= value=1i`,
-		`cpu,host=serverA,region=,zone=us-west value=1i`,
-		`cpu, value=1`,
-		`cpu, ,,`,
-		`cpu,,,`,
-		`cpu,host=serverA,=us-east value=1i`,
-		`cpu,host=serverAa\,,=us-east value=1i`,
-		`cpu,host=serverA\,,=us-east value=1i`,
-		`cpu, =serverA value=1i`,
-	} {
-		_, err := Parse([]byte(tt + "\n"))
-		assert.Error(t, err, tt)
-	}
-}
-
-func TestParsePointWhitespace(t *testing.T) {
-	for _, tt := range []string{
-		`cpu    value=1.0 1257894000000000000`,
-		`cpu value=1.0     1257894000000000000`,
-		`cpu      value=1.0     1257894000000000000`,
-		`cpu value=1.0 1257894000000000000   `,
-	} {
-		m, err := Parse([]byte(tt + "\n"))
-		assert.NoError(t, err, tt)
-		assert.Equal(t, "cpu", m[0].Name())
-		assert.Equal(t, map[string]interface{}{"value": float64(1)}, m[0].Fields())
-	}
-}
-
-func TestParsePointInvalidFields(t *testing.T) {
-	for _, tt := range []string{
-		"test,foo=bar a=101,=value",
-		"test,foo=bar =value",
-		"test,foo=bar a=101,key=",
-		"test,foo=bar key=",
-		`test,foo=bar a=101,b="foo`,
-	} {
-		_, err := Parse([]byte(tt + "\n"))
-		assert.Error(t, err, tt)
-	}
-}
-
-func TestParsePointNoFields(t *testing.T) {
-	for _, tt := range []string{
-		"cpu_load_short,host=server01,region=us-west",
-		"very_long_measurement_name",
-		"cpu,host==",
-		"============",
-		"cpu",
-		"cpu\n\n\n\n\n\n\n",
-		"                ",
-	} {
-		_, err := Parse([]byte(tt + "\n"))
-		assert.Error(t, err, tt)
-	}
-}
-
-// a b=1 << this is the shortest possible metric
-// any shorter is just ignored
-func TestParseBufTooShort(t *testing.T) {
-	for _, tt := range []string{
-		"",
-		"a",
-		"a ",
-		"a b=",
-	} {
-		_, err := Parse([]byte(tt + "\n"))
-		assert.Error(t, err, tt)
-	}
-}
-
-func TestParseInvalidBooleans(t *testing.T) {
-	for _, tt := range []string{
-		"test b=tru",
-		"test b=fals",
-		"test b=faLse",
-		"test q=foo",
-		"test b=lambchops",
-	} {
-		_, err := Parse([]byte(tt + "\n"))
-		assert.Error(t, err, tt)
-	}
-}
-
-func TestParseInvalidNumbers(t *testing.T) {
-	for _, tt := range []string{
-		"test b=-",
-		"test b=1.1.1",
-		"test b=nan",
-		"test b=9i10",
-		"test b=9999999999999999999i",
-	} {
-		_, err := Parse([]byte(tt + "\n"))
-		assert.Error(t, err, tt)
-	}
-}
-
-func TestParseNegativeTimestamps(t *testing.T) {
-	for _, tt := range []string{
-		"test foo=101 -1257894000000000000",
-	} {
-		metrics, err := Parse([]byte(tt + "\n"))
-		assert.NoError(t, err, tt)
-		assert.True(t, metrics[0].Time().Equal(time.Unix(0, -1257894000000000000)))
-	}
-}
-
-func TestParseMaxKeyLength(t *testing.T) {
-	key := ""
-	for {
-		if len(key) > MaxKeyLength {
-			break
-		}
-		key += "test"
-	}
-
-	_, err := Parse([]byte(key + " value=1\n"))
-	assert.Error(t, err)
-}

metric/uint_support.go

@@ -0,0 +1,7 @@
+// +build uint64
+
+package metric
+
+func init() {
+	EnableUintSupport()
+}

output.go

@@ -13,6 +13,12 @@ type Output interface {
 	Write(metrics []Metric) error
 }
 
+type AggregatingOutput interface {
+	Add(in Metric)
+	Push() []Metric
+	Reset()
+}
+
 type ServiceOutput interface {
 	// Connect to the Output
 	Connect() error

plugins/aggregators/all/all.go

@@ -1,5 +1,7 @@
 package all
 
 import (
+	_ "github.com/influxdata/telegraf/plugins/aggregators/basicstats"
+	_ "github.com/influxdata/telegraf/plugins/aggregators/histogram"
 	_ "github.com/influxdata/telegraf/plugins/aggregators/minmax"
 )

plugins/aggregators/basicstats/README.md

@@ -0,0 +1,56 @@
+# BasicStats Aggregator Plugin
+
+The BasicStats aggregator plugin give us count,max,min,mean,sum,s2(variance), stdev for a set of values,
+emitting the aggregate every `period` seconds.
+
+### Configuration:
+
+```toml
+# Keep the aggregate basicstats of each metric passing through.
+[[aggregators.basicstats]]
+
+  ## General Aggregator Arguments:
+
+  ## The period on which to flush & clear the aggregator.
+  period = "30s"
+
+  ## If true, the original metric will be dropped by the
+  ## aggregator and will not get sent to the output plugins.
+  drop_original = false
+
+  ## BasicStats Arguments:
+
+  ## Configures which basic stats to push as fields
+  stats = ["count","min","max","mean","stdev","s2","sum"]
+```
+
+- stats
+    - If not specified, then `count`, `min`, `max`, `mean`, `stdev`, and `s2` are aggregated and pushed as fields.  `sum` is not aggregated by default to maintain backwards compatibility.
+    - If empty array, no stats are aggregated
+
+### Measurements & Fields:
+
+- measurement1
+    - field1_count
+    - field1_max
+    - field1_min
+    - field1_mean
+    - field1_sum
+    - field1_s2 (variance)
+    - field1_stdev (standard deviation)
+
+### Tags:
+
+No tags are applied by this aggregator.
+
+### Example Output:
+
+```
+$ telegraf --config telegraf.conf --quiet
+system,host=tars load1=1 1475583980000000000
+system,host=tars load1=1 1475583990000000000
+system,host=tars load1_count=2,load1_max=1,load1_min=1,load1_mean=1,load1_sum=2,load1_s2=0,load1_stdev=0 1475584010000000000
+system,host=tars load1=1 1475584020000000000
+system,host=tars load1=3 1475584030000000000
+system,host=tars load1_count=2,load1_max=3,load1_min=1,load1_mean=2,load1_sum=4,load1_s2=2,load1_stdev=1.414162 1475584010000000000
+```

plugins/aggregators/basicstats/basicstats.go

@@ -0,0 +1,258 @@
+package basicstats
+
+import (
+	"log"
+	"math"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/aggregators"
+)
+
+type BasicStats struct {
+	Stats []string `toml:"stats"`
+
+	cache       map[uint64]aggregate
+	statsConfig *configuredStats
+}
+
+type configuredStats struct {
+	count    bool
+	min      bool
+	max      bool
+	mean     bool
+	variance bool
+	stdev    bool
+	sum      bool
+}
+
+func NewBasicStats() *BasicStats {
+	mm := &BasicStats{}
+	mm.Reset()
+	return mm
+}
+
+type aggregate struct {
+	fields map[string]basicstats
+	name   string
+	tags   map[string]string
+}
+
+type basicstats struct {
+	count float64
+	min   float64
+	max   float64
+	sum   float64
+	mean  float64
+	M2    float64 //intermedia value for variance/stdev
+}
+
+var sampleConfig = `
+  ## General Aggregator Arguments:
+  ## The period on which to flush & clear the aggregator.
+  period = "30s"
+  ## If true, the original metric will be dropped by the
+  ## aggregator and will not get sent to the output plugins.
+  drop_original = false
+`
+
+func (m *BasicStats) SampleConfig() string {
+	return sampleConfig
+}
+
+func (m *BasicStats) Description() string {
+	return "Keep the aggregate basicstats of each metric passing through."
+}
+
+func (m *BasicStats) Add(in telegraf.Metric) {
+	id := in.HashID()
+	if _, ok := m.cache[id]; !ok {
+		// hit an uncached metric, create caches for first time:
+		a := aggregate{
+			name:   in.Name(),
+			tags:   in.Tags(),
+			fields: make(map[string]basicstats),
+		}
+		for k, v := range in.Fields() {
+			if fv, ok := convert(v); ok {
+				a.fields[k] = basicstats{
+					count: 1,
+					min:   fv,
+					max:   fv,
+					mean:  fv,
+					sum:   fv,
+					M2:    0.0,
+				}
+			}
+		}
+		m.cache[id] = a
+	} else {
+		for k, v := range in.Fields() {
+			if fv, ok := convert(v); ok {
+				if _, ok := m.cache[id].fields[k]; !ok {
+					// hit an uncached field of a cached metric
+					m.cache[id].fields[k] = basicstats{
+						count: 1,
+						min:   fv,
+						max:   fv,
+						mean:  fv,
+						sum:   fv,
+						M2:    0.0,
+					}
+					continue
+				}
+
+				tmp := m.cache[id].fields[k]
+				//https://en.m.wikipedia.org/wiki/Algorithms_for_calculating_variance
+				//variable initialization
+				x := fv
+				mean := tmp.mean
+				M2 := tmp.M2
+				//counter compute
+				n := tmp.count + 1
+				tmp.count = n
+				//mean compute
+				delta := x - mean
+				mean = mean + delta/n
+				tmp.mean = mean
+				//variance/stdev compute
+				M2 = M2 + delta*(x-mean)
+				tmp.M2 = M2
+				//max/min compute
+				if fv < tmp.min {
+					tmp.min = fv
+				} else if fv > tmp.max {
+					tmp.max = fv
+				}
+				//sum compute
+				tmp.sum += fv
+				//store final data
+				m.cache[id].fields[k] = tmp
+			}
+		}
+	}
+}
+
+func (m *BasicStats) Push(acc telegraf.Accumulator) {
+
+	config := getConfiguredStats(m)
+
+	for _, aggregate := range m.cache {
+		fields := map[string]interface{}{}
+		for k, v := range aggregate.fields {
+
+			if config.count {
+				fields[k+"_count"] = v.count
+			}
+			if config.min {
+				fields[k+"_min"] = v.min
+			}
+			if config.max {
+				fields[k+"_max"] = v.max
+			}
+			if config.mean {
+				fields[k+"_mean"] = v.mean
+			}
+			if config.sum {
+				fields[k+"_sum"] = v.sum
+			}
+
+			//v.count always >=1
+			if v.count > 1 {
+				variance := v.M2 / (v.count - 1)
+
+				if config.variance {
+					fields[k+"_s2"] = variance
+				}
+				if config.stdev {
+					fields[k+"_stdev"] = math.Sqrt(variance)
+				}
+			}
+			//if count == 1 StdDev = infinite => so I won't send data
+		}
+
+		if len(fields) > 0 {
+			acc.AddFields(aggregate.name, fields, aggregate.tags)
+		}
+	}
+}
+
+func parseStats(names []string) *configuredStats {
+
+	parsed := &configuredStats{}
+
+	for _, name := range names {
+
+		switch name {
+
+		case "count":
+			parsed.count = true
+		case "min":
+			parsed.min = true
+		case "max":
+			parsed.max = true
+		case "mean":
+			parsed.mean = true
+		case "s2":
+			parsed.variance = true
+		case "stdev":
+			parsed.stdev = true
+		case "sum":
+			parsed.sum = true
+
+		default:
+			log.Printf("W! Unrecognized basic stat '%s', ignoring", name)
+		}
+	}
+
+	return parsed
+}
+
+func defaultStats() *configuredStats {
+
+	defaults := &configuredStats{}
+
+	defaults.count = true
+	defaults.min = true
+	defaults.max = true
+	defaults.mean = true
+	defaults.variance = true
+	defaults.stdev = true
+	defaults.sum = false
+
+	return defaults
+}
+
+func getConfiguredStats(m *BasicStats) *configuredStats {
+
+	if m.statsConfig == nil {
+
+		if m.Stats == nil {
+			m.statsConfig = defaultStats()
+		} else {
+			m.statsConfig = parseStats(m.Stats)
+		}
+	}
+
+	return m.statsConfig
+}
+
+func (m *BasicStats) Reset() {
+	m.cache = make(map[uint64]aggregate)
+}
+
+func convert(in interface{}) (float64, bool) {
+	switch v := in.(type) {
+	case float64:
+		return v, true
+	case int64:
+		return float64(v), true
+	default:
+		return 0, false
+	}
+}
+
+func init() {
+	aggregators.Add("basicstats", func() telegraf.Aggregator {
+		return NewBasicStats()
+	})
+}

plugins/aggregators/basicstats/basicstats_test.go

@@ -0,0 +1,511 @@
+package basicstats
+
+import (
+	"math"
+	"testing"
+	"time"
+
+	"github.com/influxdata/telegraf/metric"
+	"github.com/influxdata/telegraf/testutil"
+	"github.com/stretchr/testify/assert"
+)
+
+var m1, _ = metric.New("m1",
+	map[string]string{"foo": "bar"},
+	map[string]interface{}{
+		"a": int64(1),
+		"b": int64(1),
+		"c": float64(2),
+		"d": float64(2),
+	},
+	time.Now(),
+)
+var m2, _ = metric.New("m1",
+	map[string]string{"foo": "bar"},
+	map[string]interface{}{
+		"a":        int64(1),
+		"b":        int64(3),
+		"c":        float64(4),
+		"d":        float64(6),
+		"e":        float64(200),
+		"ignoreme": "string",
+		"andme":    true,
+	},
+	time.Now(),
+)
+
+func BenchmarkApply(b *testing.B) {
+	minmax := NewBasicStats()
+
+	for n := 0; n < b.N; n++ {
+		minmax.Add(m1)
+		minmax.Add(m2)
+	}
+}
+
+// Test two metrics getting added.
+func TestBasicStatsWithPeriod(t *testing.T) {
+	acc := testutil.Accumulator{}
+	minmax := NewBasicStats()
+
+	minmax.Add(m1)
+	minmax.Add(m2)
+	minmax.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_count": float64(2), //a
+		"a_max":   float64(1),
+		"a_min":   float64(1),
+		"a_mean":  float64(1),
+		"a_stdev": float64(0),
+		"a_s2":    float64(0),
+		"b_count": float64(2), //b
+		"b_max":   float64(3),
+		"b_min":   float64(1),
+		"b_mean":  float64(2),
+		"b_s2":    float64(2),
+		"b_stdev": math.Sqrt(2),
+		"c_count": float64(2), //c
+		"c_max":   float64(4),
+		"c_min":   float64(2),
+		"c_mean":  float64(3),
+		"c_s2":    float64(2),
+		"c_stdev": math.Sqrt(2),
+		"d_count": float64(2), //d
+		"d_max":   float64(6),
+		"d_min":   float64(2),
+		"d_mean":  float64(4),
+		"d_s2":    float64(8),
+		"d_stdev": math.Sqrt(8),
+		"e_count": float64(1), //e
+		"e_max":   float64(200),
+		"e_min":   float64(200),
+		"e_mean":  float64(200),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test two metrics getting added with a push/reset in between (simulates
+// getting added in different periods.)
+func TestBasicStatsDifferentPeriods(t *testing.T) {
+	acc := testutil.Accumulator{}
+	minmax := NewBasicStats()
+
+	minmax.Add(m1)
+	minmax.Push(&acc)
+	expectedFields := map[string]interface{}{
+		"a_count": float64(1), //a
+		"a_max":   float64(1),
+		"a_min":   float64(1),
+		"a_mean":  float64(1),
+		"b_count": float64(1), //b
+		"b_max":   float64(1),
+		"b_min":   float64(1),
+		"b_mean":  float64(1),
+		"c_count": float64(1), //c
+		"c_max":   float64(2),
+		"c_min":   float64(2),
+		"c_mean":  float64(2),
+		"d_count": float64(1), //d
+		"d_max":   float64(2),
+		"d_min":   float64(2),
+		"d_mean":  float64(2),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+
+	acc.ClearMetrics()
+	minmax.Reset()
+	minmax.Add(m2)
+	minmax.Push(&acc)
+	expectedFields = map[string]interface{}{
+		"a_count": float64(1), //a
+		"a_max":   float64(1),
+		"a_min":   float64(1),
+		"a_mean":  float64(1),
+		"b_count": float64(1), //b
+		"b_max":   float64(3),
+		"b_min":   float64(3),
+		"b_mean":  float64(3),
+		"c_count": float64(1), //c
+		"c_max":   float64(4),
+		"c_min":   float64(4),
+		"c_mean":  float64(4),
+		"d_count": float64(1), //d
+		"d_max":   float64(6),
+		"d_min":   float64(6),
+		"d_mean":  float64(6),
+		"e_count": float64(1), //e
+		"e_max":   float64(200),
+		"e_min":   float64(200),
+		"e_mean":  float64(200),
+	}
+	expectedTags = map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating count
+func TestBasicStatsWithOnlyCount(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"count"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_count": float64(2),
+		"b_count": float64(2),
+		"c_count": float64(2),
+		"d_count": float64(2),
+		"e_count": float64(1),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating minimum
+func TestBasicStatsWithOnlyMin(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"min"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_min": float64(1),
+		"b_min": float64(1),
+		"c_min": float64(2),
+		"d_min": float64(2),
+		"e_min": float64(200),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating maximum
+func TestBasicStatsWithOnlyMax(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"max"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_max": float64(1),
+		"b_max": float64(3),
+		"c_max": float64(4),
+		"d_max": float64(6),
+		"e_max": float64(200),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating mean
+func TestBasicStatsWithOnlyMean(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"mean"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_mean": float64(1),
+		"b_mean": float64(2),
+		"c_mean": float64(3),
+		"d_mean": float64(4),
+		"e_mean": float64(200),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating sum
+func TestBasicStatsWithOnlySum(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"sum"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_sum": float64(2),
+		"b_sum": float64(4),
+		"c_sum": float64(6),
+		"d_sum": float64(8),
+		"e_sum": float64(200),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Verify that sum doesn't suffer from floating point errors.  Early
+// implementations of sum were calulated from mean and count, which
+// e.g. summed "1, 1, 5, 1" as "7.999999..." instead of 8.
+func TestBasicStatsWithOnlySumFloatingPointErrata(t *testing.T) {
+
+	var sum1, _ = metric.New("m1",
+		map[string]string{},
+		map[string]interface{}{
+			"a": int64(1),
+		},
+		time.Now(),
+	)
+	var sum2, _ = metric.New("m1",
+		map[string]string{},
+		map[string]interface{}{
+			"a": int64(1),
+		},
+		time.Now(),
+	)
+	var sum3, _ = metric.New("m1",
+		map[string]string{},
+		map[string]interface{}{
+			"a": int64(5),
+		},
+		time.Now(),
+	)
+	var sum4, _ = metric.New("m1",
+		map[string]string{},
+		map[string]interface{}{
+			"a": int64(1),
+		},
+		time.Now(),
+	)
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"sum"}
+
+	aggregator.Add(sum1)
+	aggregator.Add(sum2)
+	aggregator.Add(sum3)
+	aggregator.Add(sum4)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_sum": float64(8),
+	}
+	expectedTags := map[string]string{}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating variance
+func TestBasicStatsWithOnlyVariance(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"s2"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_s2": float64(0),
+		"b_s2": float64(2),
+		"c_s2": float64(2),
+		"d_s2": float64(8),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating standard deviation
+func TestBasicStatsWithOnlyStandardDeviation(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"stdev"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_stdev": float64(0),
+		"b_stdev": math.Sqrt(2),
+		"c_stdev": math.Sqrt(2),
+		"d_stdev": math.Sqrt(8),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test only aggregating minimum and maximum
+func TestBasicStatsWithMinAndMax(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"min", "max"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_max": float64(1), //a
+		"a_min": float64(1),
+		"b_max": float64(3), //b
+		"b_min": float64(1),
+		"c_max": float64(4), //c
+		"c_min": float64(2),
+		"d_max": float64(6), //d
+		"d_min": float64(2),
+		"e_max": float64(200), //e
+		"e_min": float64(200),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test aggregating with all stats
+func TestBasicStatsWithAllStats(t *testing.T) {
+	acc := testutil.Accumulator{}
+	minmax := NewBasicStats()
+	minmax.Stats = []string{"count", "min", "max", "mean", "stdev", "s2", "sum"}
+
+	minmax.Add(m1)
+	minmax.Add(m2)
+	minmax.Push(&acc)
+
+	expectedFields := map[string]interface{}{
+		"a_count": float64(2), //a
+		"a_max":   float64(1),
+		"a_min":   float64(1),
+		"a_mean":  float64(1),
+		"a_stdev": float64(0),
+		"a_s2":    float64(0),
+		"a_sum":   float64(2),
+		"b_count": float64(2), //b
+		"b_max":   float64(3),
+		"b_min":   float64(1),
+		"b_mean":  float64(2),
+		"b_s2":    float64(2),
+		"b_sum":   float64(4),
+		"b_stdev": math.Sqrt(2),
+		"c_count": float64(2), //c
+		"c_max":   float64(4),
+		"c_min":   float64(2),
+		"c_mean":  float64(3),
+		"c_s2":    float64(2),
+		"c_stdev": math.Sqrt(2),
+		"c_sum":   float64(6),
+		"d_count": float64(2), //d
+		"d_max":   float64(6),
+		"d_min":   float64(2),
+		"d_mean":  float64(4),
+		"d_s2":    float64(8),
+		"d_stdev": math.Sqrt(8),
+		"d_sum":   float64(8),
+		"e_count": float64(1), //e
+		"e_max":   float64(200),
+		"e_min":   float64(200),
+		"e_mean":  float64(200),
+		"e_sum":   float64(200),
+	}
+	expectedTags := map[string]string{
+		"foo": "bar",
+	}
+	acc.AssertContainsTaggedFields(t, "m1", expectedFields, expectedTags)
+}
+
+// Test that if an empty array is passed, no points are pushed
+func TestBasicStatsWithNoStats(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	acc.AssertDoesNotContainMeasurement(t, "m1")
+}
+
+// Test that if an unknown stat is configured, it doesn't explode
+func TestBasicStatsWithUnknownStat(t *testing.T) {
+
+	aggregator := NewBasicStats()
+	aggregator.Stats = []string{"crazy"}
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	acc.AssertDoesNotContainMeasurement(t, "m1")
+}
+
+// Test that if Stats isn't supplied, then we only do count, min, max, mean,
+// stdev, and s2.  We purposely exclude sum for backwards compatability,
+// otherwise user's working systems will suddenly (and surprisingly) start
+// capturing sum without their input.
+func TestBasicStatsWithDefaultStats(t *testing.T) {
+
+	aggregator := NewBasicStats()
+
+	aggregator.Add(m1)
+	aggregator.Add(m2)
+
+	acc := testutil.Accumulator{}
+	aggregator.Push(&acc)
+
+	assert.True(t, acc.HasField("m1", "a_count"))
+	assert.True(t, acc.HasField("m1", "a_min"))
+	assert.True(t, acc.HasField("m1", "a_max"))
+	assert.True(t, acc.HasField("m1", "a_mean"))
+	assert.True(t, acc.HasField("m1", "a_stdev"))
+	assert.True(t, acc.HasField("m1", "a_s2"))
+	assert.False(t, acc.HasField("m1", "a_sum"))
+}

plugins/aggregators/histogram/README.md

@@ -0,0 +1,97 @@
+# Histogram Aggregator Plugin
+
+The histogram aggregator plugin creates histograms containing the counts of
+field values within a range.
+
+Values added to a bucket are also added to the larger buckets in the
+distribution.  This creates a [cumulative histogram](https://en.wikipedia.org/wiki/Histogram#/media/File:Cumulative_vs_normal_histogram.svg).
+
+Like other Telegraf aggregators, the metric is emitted every `period` seconds.
+Bucket counts however are not reset between periods and will be non-strictly
+increasing while Telegraf is running.
+
+#### Design
+
+Each metric is passed to the aggregator and this aggregator searches
+histogram buckets for those fields, which have been specified in the
+config. If buckets are found, the aggregator will increment +1 to the appropriate
+bucket otherwise it will be added to the `+Inf` bucket.  Every `period`
+seconds this data will be forwarded to the outputs.
+
+The algorithm of hit counting to buckets was implemented on the base
+of the algorithm which is implemented in the Prometheus
+[client](https://github.com/prometheus/client_golang/blob/master/prometheus/histogram.go).
+
+### Configuration
+
+```toml
+# Configuration for aggregate histogram metrics
+[[aggregators.histogram]]
+  ## The period in which to flush the aggregator.
+  period = "30s"
+
+  ## If true, the original metric will be dropped by the
+  ## aggregator and will not get sent to the output plugins.
+  drop_original = false
+
+  ## Example config that aggregates all fields of the metric.
+  # [[aggregators.histogram.config]]
+  #   ## The set of buckets.
+  #   buckets = [0.0, 15.6, 34.5, 49.1, 71.5, 80.5, 94.5, 100.0]
+  #   ## The name of metric.
+  #   measurement_name = "cpu"
+
+  ## Example config that aggregates only specific fields of the metric.
+  # [[aggregators.histogram.config]]
+  #   ## The set of buckets.
+  #   buckets = [0.0, 10.0, 20.0, 30.0, 40.0, 50.0, 60.0, 70.0, 80.0, 90.0, 100.0]
+  #   ## The name of metric.
+  #   measurement_name = "diskio"
+  #   ## The concrete fields of metric
+  #   fields = ["io_time", "read_time", "write_time"]
+```
+
+The user is responsible for defining the bounds of the histogram bucket as
+well as the measurement name and fields to aggregate.
+
+Each histogram config section must contain a `buckets` and `measurement_name`
+option.  Optionally, if `fields` is set only the fields listed will be
+aggregated.  If `fields` is not set all fields are aggregated.
+
+The `buckets` option contains a list of floats which specify the bucket
+boundaries.  Each float value defines the inclusive upper bound of the bucket.
+The `+Inf` bucket is added automatically and does not need to be defined.
+
+### Measurements & Fields:
+
+The postfix `bucket` will be added to each field key.
+
+- measurement1
+    - field1_bucket
+    - field2_bucket
+
+### Tags:
+
+All measurements are given the tag `le`. This tag has the border value of
+bucket. It means that the metric value is less than or equal to the value of
+this tag.  For example, let assume that we have the metric value 10 and the
+following buckets: [5, 10, 30, 70, 100]. Then the tag `le` will have the value
+10, because the metrics value is passed into bucket with right border value
+`10`.
+
+### Example Output:
+
+```
+cpu,cpu=cpu1,host=localhost,le=0.0 usage_idle_bucket=0i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=10.0 usage_idle_bucket=0i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=20.0 usage_idle_bucket=1i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=30.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=40.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=50.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=60.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=70.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=80.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=90.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=100.0 usage_idle_bucket=2i 1486998330000000000
+cpu,cpu=cpu1,host=localhost,le=+Inf usage_idle_bucket=2i 1486998330000000000
+```

plugins/aggregators/histogram/histogram.go

@@ -0,0 +1,315 @@
+package histogram
+
+import (
+	"sort"
+	"strconv"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/aggregators"
+)
+
+// bucketTag is the tag, which contains right bucket border
+const bucketTag = "le"
+
+// bucketInf is the right bucket border for infinite values
+const bucketInf = "+Inf"
+
+// HistogramAggregator is aggregator with histogram configs and particular histograms for defined metrics
+type HistogramAggregator struct {
+	Configs []config `toml:"config"`
+
+	buckets bucketsByMetrics
+	cache   map[uint64]metricHistogramCollection
+}
+
+// config is the config, which contains name, field of metric and histogram buckets.
+type config struct {
+	Metric  string   `toml:"measurement_name"`
+	Fields  []string `toml:"fields"`
+	Buckets buckets  `toml:"buckets"`
+}
+
+// bucketsByMetrics contains the buckets grouped by metric and field name
+type bucketsByMetrics map[string]bucketsByFields
+
+// bucketsByFields contains the buckets grouped by field name
+type bucketsByFields map[string]buckets
+
+// buckets contains the right borders buckets
+type buckets []float64
+
+// metricHistogramCollection aggregates the histogram data
+type metricHistogramCollection struct {
+	histogramCollection map[string]counts
+	name                string
+	tags                map[string]string
+}
+
+// counts is the number of hits in the bucket
+type counts []int64
+
+// groupedByCountFields contains grouped fields by their count and fields values
+type groupedByCountFields struct {
+	name            string
+	tags            map[string]string
+	fieldsWithCount map[string]int64
+}
+
+// NewHistogramAggregator creates new histogram aggregator
+func NewHistogramAggregator() telegraf.Aggregator {
+	h := &HistogramAggregator{}
+	h.buckets = make(bucketsByMetrics)
+	h.resetCache()
+
+	return h
+}
+
+var sampleConfig = `
+  ## The period in which to flush the aggregator.
+  period = "30s"
+
+  ## If true, the original metric will be dropped by the
+  ## aggregator and will not get sent to the output plugins.
+  drop_original = false
+
+  ## Example config that aggregates all fields of the metric.
+  # [[aggregators.histogram.config]]
+  #   ## The set of buckets.
+  #   buckets = [0.0, 15.6, 34.5, 49.1, 71.5, 80.5, 94.5, 100.0]
+  #   ## The name of metric.
+  #   measurement_name = "cpu"
+
+  ## Example config that aggregates only specific fields of the metric.
+  # [[aggregators.histogram.config]]
+  #   ## The set of buckets.
+  #   buckets = [0.0, 10.0, 20.0, 30.0, 40.0, 50.0, 60.0, 70.0, 80.0, 90.0, 100.0]
+  #   ## The name of metric.
+  #   measurement_name = "diskio"
+  #   ## The concrete fields of metric
+  #   fields = ["io_time", "read_time", "write_time"]
+`
+
+// SampleConfig returns sample of config
+func (h *HistogramAggregator) SampleConfig() string {
+	return sampleConfig
+}
+
+// Description returns description of aggregator plugin
+func (h *HistogramAggregator) Description() string {
+	return "Create aggregate histograms."
+}
+
+// Add adds new hit to the buckets
+func (h *HistogramAggregator) Add(in telegraf.Metric) {
+	bucketsByField := make(map[string][]float64)
+	for field := range in.Fields() {
+		buckets := h.getBuckets(in.Name(), field)
+		if buckets != nil {
+			bucketsByField[field] = buckets
+		}
+	}
+
+	if len(bucketsByField) == 0 {
+		return
+	}
+
+	id := in.HashID()
+	agr, ok := h.cache[id]
+	if !ok {
+		agr = metricHistogramCollection{
+			name:                in.Name(),
+			tags:                in.Tags(),
+			histogramCollection: make(map[string]counts),
+		}
+	}
+
+	for field, value := range in.Fields() {
+		if buckets, ok := bucketsByField[field]; ok {
+			if agr.histogramCollection[field] == nil {
+				agr.histogramCollection[field] = make(counts, len(buckets)+1)
+			}
+
+			if value, ok := convert(value); ok {
+				index := sort.SearchFloat64s(buckets, value)
+				agr.histogramCollection[field][index]++
+			}
+		}
+	}
+
+	h.cache[id] = agr
+}
+
+// Push returns histogram values for metrics
+func (h *HistogramAggregator) Push(acc telegraf.Accumulator) {
+	metricsWithGroupedFields := []groupedByCountFields{}
+
+	for _, aggregate := range h.cache {
+		for field, counts := range aggregate.histogramCollection {
+			h.groupFieldsByBuckets(&metricsWithGroupedFields, aggregate.name, field, copyTags(aggregate.tags), counts)
+		}
+	}
+
+	for _, metric := range metricsWithGroupedFields {
+		acc.AddFields(metric.name, makeFieldsWithCount(metric.fieldsWithCount), metric.tags)
+	}
+}
+
+// groupFieldsByBuckets groups fields by metric buckets which are represented as tags
+func (h *HistogramAggregator) groupFieldsByBuckets(
+	metricsWithGroupedFields *[]groupedByCountFields,
+	name string,
+	field string,
+	tags map[string]string,
+	counts []int64,
+) {
+	count := int64(0)
+	for index, bucket := range h.getBuckets(name, field) {
+		count += counts[index]
+
+		tags[bucketTag] = strconv.FormatFloat(bucket, 'f', -1, 64)
+		h.groupField(metricsWithGroupedFields, name, field, count, copyTags(tags))
+	}
+
+	count += counts[len(counts)-1]
+	tags[bucketTag] = bucketInf
+
+	h.groupField(metricsWithGroupedFields, name, field, count, tags)
+}
+
+// groupField groups field by count value
+func (h *HistogramAggregator) groupField(
+	metricsWithGroupedFields *[]groupedByCountFields,
+	name string,
+	field string,
+	count int64,
+	tags map[string]string,
+) {
+	for key, metric := range *metricsWithGroupedFields {
+		if name == metric.name && isTagsIdentical(tags, metric.tags) {
+			(*metricsWithGroupedFields)[key].fieldsWithCount[field] = count
+			return
+		}
+	}
+
+	fieldsWithCount := map[string]int64{
+		field: count,
+	}
+
+	*metricsWithGroupedFields = append(
+		*metricsWithGroupedFields,
+		groupedByCountFields{name: name, tags: tags, fieldsWithCount: fieldsWithCount},
+	)
+}
+
+// Reset does nothing, because we need to collect counts for a long time, otherwise if config parameter 'reset' has
+// small value, we will get a histogram with a small amount of the distribution.
+func (h *HistogramAggregator) Reset() {}
+
+// resetCache resets cached counts(hits) in the buckets
+func (h *HistogramAggregator) resetCache() {
+	h.cache = make(map[uint64]metricHistogramCollection)
+}
+
+// getBuckets finds buckets and returns them
+func (h *HistogramAggregator) getBuckets(metric string, field string) []float64 {
+	if buckets, ok := h.buckets[metric][field]; ok {
+		return buckets
+	}
+
+	for _, config := range h.Configs {
+		if config.Metric == metric {
+			if !isBucketExists(field, config) {
+				continue
+			}
+
+			if _, ok := h.buckets[metric]; !ok {
+				h.buckets[metric] = make(bucketsByFields)
+			}
+
+			h.buckets[metric][field] = sortBuckets(config.Buckets)
+		}
+	}
+
+	return h.buckets[metric][field]
+}
+
+// isBucketExists checks if buckets exists for the passed field
+func isBucketExists(field string, cfg config) bool {
+	if len(cfg.Fields) == 0 {
+		return true
+	}
+
+	for _, fl := range cfg.Fields {
+		if fl == field {
+			return true
+		}
+	}
+
+	return false
+}
+
+// sortBuckets sorts the buckets if it is needed
+func sortBuckets(buckets []float64) []float64 {
+	for i, bucket := range buckets {
+		if i < len(buckets)-1 && bucket >= buckets[i+1] {
+			sort.Float64s(buckets)
+			break
+		}
+	}
+
+	return buckets
+}
+
+// convert converts interface to concrete type
+func convert(in interface{}) (float64, bool) {
+	switch v := in.(type) {
+	case float64:
+		return v, true
+	case int64:
+		return float64(v), true
+	default:
+		return 0, false
+	}
+}
+
+// copyTags copies tags
+func copyTags(tags map[string]string) map[string]string {
+	copiedTags := map[string]string{}
+	for key, val := range tags {
+		copiedTags[key] = val
+	}
+
+	return copiedTags
+}
+
+// isTagsIdentical checks the identity of two list of tags
+func isTagsIdentical(originalTags, checkedTags map[string]string) bool {
+	if len(originalTags) != len(checkedTags) {
+		return false
+	}
+
+	for tagName, tagValue := range originalTags {
+		if tagValue != checkedTags[tagName] {
+			return false
+		}
+	}
+
+	return true
+}
+
+// makeFieldsWithCount assigns count value to all metric fields
+func makeFieldsWithCount(fieldsWithCountIn map[string]int64) map[string]interface{} {
+	fieldsWithCountOut := map[string]interface{}{}
+	for field, count := range fieldsWithCountIn {
+		fieldsWithCountOut[field+"_bucket"] = count
+	}
+
+	return fieldsWithCountOut
+}
+
+// init initializes histogram aggregator plugin
+func init() {
+	aggregators.Add("histogram", func() telegraf.Aggregator {
+		return NewHistogramAggregator()
+	})
+}

plugins/aggregators/histogram/histogram_test.go

@@ -0,0 +1,210 @@
+package histogram
+
+import (
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/metric"
+	"github.com/influxdata/telegraf/testutil"
+	"github.com/stretchr/testify/assert"
+)
+
+// NewTestHistogram creates new test histogram aggregation with specified config
+func NewTestHistogram(cfg []config) telegraf.Aggregator {
+	htm := &HistogramAggregator{Configs: cfg}
+	htm.buckets = make(bucketsByMetrics)
+	htm.resetCache()
+
+	return htm
+}
+
+// firstMetric1 is the first test metric
+var firstMetric1, _ = metric.New(
+	"first_metric_name",
+	map[string]string{"tag_name": "tag_value"},
+	map[string]interface{}{
+		"a": float64(15.3),
+		"b": float64(40),
+	},
+	time.Now(),
+)
+
+// firstMetric1 is the first test metric with other value
+var firstMetric2, _ = metric.New(
+	"first_metric_name",
+	map[string]string{"tag_name": "tag_value"},
+	map[string]interface{}{
+		"a": float64(15.9),
+		"c": float64(40),
+	},
+	time.Now(),
+)
+
+// secondMetric is the second metric
+var secondMetric, _ = metric.New(
+	"second_metric_name",
+	map[string]string{"tag_name": "tag_value"},
+	map[string]interface{}{
+		"a":        float64(105),
+		"ignoreme": "string",
+		"andme":    true,
+	},
+	time.Now(),
+)
+
+// BenchmarkApply runs benchmarks
+func BenchmarkApply(b *testing.B) {
+	histogram := NewHistogramAggregator()
+
+	for n := 0; n < b.N; n++ {
+		histogram.Add(firstMetric1)
+		histogram.Add(firstMetric2)
+		histogram.Add(secondMetric)
+	}
+}
+
+// TestHistogramWithPeriodAndOneField tests metrics for one period and for one field
+func TestHistogramWithPeriodAndOneField(t *testing.T) {
+	var cfg []config
+	cfg = append(cfg, config{Metric: "first_metric_name", Fields: []string{"a"}, Buckets: []float64{0.0, 10.0, 20.0, 30.0, 40.0}})
+	histogram := NewTestHistogram(cfg)
+
+	acc := &testutil.Accumulator{}
+
+	histogram.Add(firstMetric1)
+	histogram.Add(firstMetric2)
+	histogram.Push(acc)
+
+	if len(acc.Metrics) != 6 {
+		assert.Fail(t, "Incorrect number of metrics")
+	}
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(0)}, "0")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(0)}, "10")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2)}, "20")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2)}, "30")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2)}, "40")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2)}, bucketInf)
+}
+
+// TestHistogramWithPeriodAndAllFields tests two metrics for one period and for all fields
+func TestHistogramWithPeriodAndAllFields(t *testing.T) {
+	var cfg []config
+	cfg = append(cfg, config{Metric: "first_metric_name", Buckets: []float64{0.0, 15.5, 20.0, 30.0, 40.0}})
+	cfg = append(cfg, config{Metric: "second_metric_name", Buckets: []float64{0.0, 4.0, 10.0, 23.0, 30.0}})
+	histogram := NewTestHistogram(cfg)
+
+	acc := &testutil.Accumulator{}
+
+	histogram.Add(firstMetric1)
+	histogram.Add(firstMetric2)
+	histogram.Add(secondMetric)
+	histogram.Push(acc)
+
+	if len(acc.Metrics) != 12 {
+		assert.Fail(t, "Incorrect number of metrics")
+	}
+
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(0), "b_bucket": int64(0), "c_bucket": int64(0)}, "0")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(1), "b_bucket": int64(0), "c_bucket": int64(0)}, "15.5")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(0), "c_bucket": int64(0)}, "20")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(0), "c_bucket": int64(0)}, "30")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(1), "c_bucket": int64(1)}, "40")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(1), "c_bucket": int64(1)}, bucketInf)
+
+	assertContainsTaggedField(t, acc, "second_metric_name", map[string]interface{}{"a_bucket": int64(0), "ignoreme_bucket": int64(0), "andme_bucket": int64(0)}, "0")
+	assertContainsTaggedField(t, acc, "second_metric_name", map[string]interface{}{"a_bucket": int64(0), "ignoreme_bucket": int64(0), "andme_bucket": int64(0)}, "4")
+	assertContainsTaggedField(t, acc, "second_metric_name", map[string]interface{}{"a_bucket": int64(0), "ignoreme_bucket": int64(0), "andme_bucket": int64(0)}, "10")
+	assertContainsTaggedField(t, acc, "second_metric_name", map[string]interface{}{"a_bucket": int64(0), "ignoreme_bucket": int64(0), "andme_bucket": int64(0)}, "23")
+	assertContainsTaggedField(t, acc, "second_metric_name", map[string]interface{}{"a_bucket": int64(0), "ignoreme_bucket": int64(0), "andme_bucket": int64(0)}, "30")
+	assertContainsTaggedField(t, acc, "second_metric_name", map[string]interface{}{"a_bucket": int64(1), "ignoreme_bucket": int64(0), "andme_bucket": int64(0)}, bucketInf)
+}
+
+// TestHistogramDifferentPeriodsAndAllFields tests two metrics getting added with a push/reset in between (simulates
+// getting added in different periods) for all fields
+func TestHistogramDifferentPeriodsAndAllFields(t *testing.T) {
+
+	var cfg []config
+	cfg = append(cfg, config{Metric: "first_metric_name", Buckets: []float64{0.0, 10.0, 20.0, 30.0, 40.0}})
+	histogram := NewTestHistogram(cfg)
+
+	acc := &testutil.Accumulator{}
+	histogram.Add(firstMetric1)
+	histogram.Push(acc)
+
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(0), "b_bucket": int64(0)}, "0")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(0), "b_bucket": int64(0)}, "10")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(1), "b_bucket": int64(0)}, "20")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(1), "b_bucket": int64(0)}, "30")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(1), "b_bucket": int64(1)}, "40")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(1), "b_bucket": int64(1)}, bucketInf)
+
+	acc.ClearMetrics()
+	histogram.Add(firstMetric2)
+	histogram.Push(acc)
+
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(0), "b_bucket": int64(0), "c_bucket": int64(0)}, "0")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(0), "b_bucket": int64(0), "c_bucket": int64(0)}, "10")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(0), "c_bucket": int64(0)}, "20")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(0), "c_bucket": int64(0)}, "30")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(1), "c_bucket": int64(1)}, "40")
+	assertContainsTaggedField(t, acc, "first_metric_name", map[string]interface{}{"a_bucket": int64(2), "b_bucket": int64(1), "c_bucket": int64(1)}, bucketInf)
+}
+
+// TestWrongBucketsOrder tests the calling panic with incorrect order of buckets
+func TestWrongBucketsOrder(t *testing.T) {
+	defer func() {
+		if r := recover(); r != nil {
+			assert.Equal(
+				t,
+				"histogram buckets must be in increasing order: 90.00 >= 20.00, metrics: first_metric_name, field: a",
+				fmt.Sprint(r),
+			)
+		}
+	}()
+
+	var cfg []config
+	cfg = append(cfg, config{Metric: "first_metric_name", Buckets: []float64{0.0, 90.0, 20.0, 30.0, 40.0}})
+	histogram := NewTestHistogram(cfg)
+	histogram.Add(firstMetric2)
+}
+
+// assertContainsTaggedField is help functions to test histogram data
+func assertContainsTaggedField(t *testing.T, acc *testutil.Accumulator, metricName string, fields map[string]interface{}, le string) {
+	acc.Lock()
+	defer acc.Unlock()
+
+	for _, checkedMetric := range acc.Metrics {
+		// check metric name
+		if checkedMetric.Measurement != metricName {
+			continue
+		}
+
+		// check "le" tag
+		if checkedMetric.Tags[bucketTag] != le {
+			continue
+		}
+
+		// check fields
+		isFieldsIdentical := true
+		for field := range fields {
+			if _, ok := checkedMetric.Fields[field]; !ok {
+				isFieldsIdentical = false
+				break
+			}
+		}
+		if !isFieldsIdentical {
+			continue
+		}
+
+		// check fields with their counts
+		if assert.Equal(t, fields, checkedMetric.Fields) {
+			return
+		}
+
+		assert.Fail(t, fmt.Sprintf("incorrect fields %v of metric %s", fields, metricName))
+	}
+
+	assert.Fail(t, fmt.Sprintf("unknown measurement '%s' with tags: %v, fields: %v", metricName, map[string]string{"le": le}, fields))
+}

plugins/inputs/EXAMPLE_README.md

@@ -1,45 +1,61 @@
 # Example Input Plugin
 
-The example plugin gathers metrics about example things
+The example plugin gathers metrics about example things.  This description
+explains at a high level what the plugin does and provides links to where
+additional information can be found.
 
 ### Configuration:
 
+This section contains the default TOML to configure the plugin.  You can
+generate it using `telegraf --usage <plugin-name>`.
+
 ```toml
 # Description
 [[inputs.example]]
-  # SampleConfig
+  example_option = "example_value"
 ```
 
-### Measurements & Fields:
+### Metrics:
 
-<optional description>
+Here you should add an optional description and links to where the user can
+get more information about the measurements.
+
+If the output is determined dynamically based on the input source, or there
+are more metrics than can reasonably be listed, describe how the input is
+mapped to the output.
 
 - measurement1
-    - field1 (type, unit)
-    - field2 (float, percent)
-- measurement2
-    - field3 (integer, bytes)
-
-### Tags:
-
-- All measurements have the following tags:
+  - tags:
     - tag1 (optional description)
     - tag2
-- measurement2 has the following tags:
+  - fields:
+    - field1 (type, unit)
+    - field2 (float, percent)
+
+- measurement2
+  - tags:
     - tag3
-    
+  - fields:
+    - field3 (integer, bytes)
+
 ### Sample Queries:
 
-These are some useful queries (to generate dashboards or other) to run against data from this plugin:
+This section should contain some useful InfluxDB queries that can be used to
+get started with the plugin or to generate dashboards.  For each query listed,
+describe at a high level what data is returned.
 
+Get the max, mean, and min for the measurement in the last hour:
 ```
 SELECT max(field1), mean(field1), min(field1) FROM measurement1 WHERE tag1=bar AND time > now() - 1h GROUP BY tag
 ```
 
 ### Example Output:
 
+This section shows example output in Line Protocol format.  You can often use
+`telegraf --input-filter <plugin-name> --test` or use the `file` output to get
+this information.
+
 ```
-$ ./telegraf -config telegraf.conf -input-filter example -test
 measurement1,tag1=foo,tag2=bar field1=1i,field2=2.1 1453831884664956455
 measurement2,tag1=foo,tag2=bar,tag3=baz field3=1i 1453831884664956455
 ```

plugins/inputs/aerospike/aerospike.go

@@ -10,7 +10,6 @@ import (
 	"time"
 
 	"github.com/influxdata/telegraf"
-	"github.com/influxdata/telegraf/internal/errchan"
 	"github.com/influxdata/telegraf/plugins/inputs"
 
 	as "github.com/aerospike/aerospike-client-go"
@@ -41,17 +40,16 @@ func (a *Aerospike) Gather(acc telegraf.Accumulator) error {
 	}
 
 	var wg sync.WaitGroup
-	errChan := errchan.New(len(a.Servers))
 	wg.Add(len(a.Servers))
 	for _, server := range a.Servers {
 		go func(serv string) {
 			defer wg.Done()
-			errChan.C <- a.gatherServer(serv, acc)
+			acc.AddError(a.gatherServer(serv, acc))
 		}(server)
 	}
 
 	wg.Wait()
-	return errChan.Error()
+	return nil
 }
 
 func (a *Aerospike) gatherServer(hostport string, acc telegraf.Accumulator) error {
@@ -75,10 +73,9 @@ func (a *Aerospike) gatherServer(hostport string, acc telegraf.Accumulator) erro
 	for _, n := range nodes {
 		tags := map[string]string{
 			"aerospike_host": hostport,
+			"node_name":      n.GetName(),
 		}
-		fields := map[string]interface{}{
-			"node_name": n.GetName(),
-		}
+		fields := make(map[string]interface{})
 		stats, err := as.RequestNodeStats(n)
 		if err != nil {
 			return err
@@ -88,7 +85,7 @@ func (a *Aerospike) gatherServer(hostport string, acc telegraf.Accumulator) erro
 			if err == nil {
 				fields[strings.Replace(k, "-", "_", -1)] = val
 			} else {
-				log.Printf("I! skipping aerospike field %v with int64 overflow", k)
+				log.Printf("I! skipping aerospike field %v with int64 overflow: %q", k, v)
 			}
 		}
 		acc.AddFields("aerospike_node", fields, tags, time.Now())
@@ -102,11 +99,10 @@ func (a *Aerospike) gatherServer(hostport string, acc telegraf.Accumulator) erro
 		for _, namespace := range namespaces {
 			nTags := map[string]string{
 				"aerospike_host": hostport,
+				"node_name":      n.GetName(),
 			}
 			nTags["namespace"] = namespace
-			nFields := map[string]interface{}{
-				"node_name": n.GetName(),
-			}
+			nFields := make(map[string]interface{})
 			info, err := as.RequestNodeInfo(n, "namespace/"+namespace)
 			if err != nil {
 				continue
@@ -121,7 +117,7 @@ func (a *Aerospike) gatherServer(hostport string, acc telegraf.Accumulator) erro
 				if err == nil {
 					nFields[strings.Replace(parts[0], "-", "_", -1)] = val
 				} else {
-					log.Printf("I! skipping aerospike field %v with int64 overflow", parts[0])
+					log.Printf("I! skipping aerospike field %v with int64 overflow: %q", parts[0], parts[1])
 				}
 			}
 			acc.AddFields("aerospike_namespace", nFields, nTags, time.Now())

plugins/inputs/aerospike/aerospike_test.go

@@ -19,12 +19,14 @@ func TestAerospikeStatistics(t *testing.T) {
 
 	var acc testutil.Accumulator
 
-	err := a.Gather(&acc)
+	err := acc.GatherError(a.Gather)
 	require.NoError(t, err)
 
 	assert.True(t, acc.HasMeasurement("aerospike_node"))
+	assert.True(t, acc.HasTag("aerospike_node", "node_name"))
 	assert.True(t, acc.HasMeasurement("aerospike_namespace"))
-	assert.True(t, acc.HasIntField("aerospike_node", "batch_error"))
+	assert.True(t, acc.HasTag("aerospike_namespace", "node_name"))
+	assert.True(t, acc.HasInt64Field("aerospike_node", "batch_error"))
 }
 
 func TestAerospikeStatisticsPartialErr(t *testing.T) {
@@ -41,12 +43,11 @@ func TestAerospikeStatisticsPartialErr(t *testing.T) {
 
 	var acc testutil.Accumulator
 
-	err := a.Gather(&acc)
-	require.Error(t, err)
+	require.Error(t, acc.GatherError(a.Gather))
 
 	assert.True(t, acc.HasMeasurement("aerospike_node"))
 	assert.True(t, acc.HasMeasurement("aerospike_namespace"))
-	assert.True(t, acc.HasIntField("aerospike_node", "batch_error"))
+	assert.True(t, acc.HasInt64Field("aerospike_node", "batch_error"))
 }
 
 func TestAerospikeParseValue(t *testing.T) {

plugins/inputs/all/all.go

@@ -2,8 +2,10 @@ package all
 
 import (
 	_ "github.com/influxdata/telegraf/plugins/inputs/aerospike"
+	_ "github.com/influxdata/telegraf/plugins/inputs/amqp_consumer"
 	_ "github.com/influxdata/telegraf/plugins/inputs/apache"
 	_ "github.com/influxdata/telegraf/plugins/inputs/bcache"
+	_ "github.com/influxdata/telegraf/plugins/inputs/bond"
 	_ "github.com/influxdata/telegraf/plugins/inputs/cassandra"
 	_ "github.com/influxdata/telegraf/plugins/inputs/ceph"
 	_ "github.com/influxdata/telegraf/plugins/inputs/cgroup"
@@ -13,45 +15,65 @@ import (
 	_ "github.com/influxdata/telegraf/plugins/inputs/consul"
 	_ "github.com/influxdata/telegraf/plugins/inputs/couchbase"
 	_ "github.com/influxdata/telegraf/plugins/inputs/couchdb"
+	_ "github.com/influxdata/telegraf/plugins/inputs/dcos"
 	_ "github.com/influxdata/telegraf/plugins/inputs/disque"
+	_ "github.com/influxdata/telegraf/plugins/inputs/dmcache"
 	_ "github.com/influxdata/telegraf/plugins/inputs/dns_query"
 	_ "github.com/influxdata/telegraf/plugins/inputs/docker"
 	_ "github.com/influxdata/telegraf/plugins/inputs/dovecot"
 	_ "github.com/influxdata/telegraf/plugins/inputs/elasticsearch"
 	_ "github.com/influxdata/telegraf/plugins/inputs/exec"
+	_ "github.com/influxdata/telegraf/plugins/inputs/fail2ban"
+	_ "github.com/influxdata/telegraf/plugins/inputs/fibaro"
 	_ "github.com/influxdata/telegraf/plugins/inputs/filestat"
+	_ "github.com/influxdata/telegraf/plugins/inputs/fluentd"
 	_ "github.com/influxdata/telegraf/plugins/inputs/graylog"
 	_ "github.com/influxdata/telegraf/plugins/inputs/haproxy"
 	_ "github.com/influxdata/telegraf/plugins/inputs/hddtemp"
+	_ "github.com/influxdata/telegraf/plugins/inputs/http"
 	_ "github.com/influxdata/telegraf/plugins/inputs/http_listener"
 	_ "github.com/influxdata/telegraf/plugins/inputs/http_response"
 	_ "github.com/influxdata/telegraf/plugins/inputs/httpjson"
 	_ "github.com/influxdata/telegraf/plugins/inputs/influxdb"
 	_ "github.com/influxdata/telegraf/plugins/inputs/internal"
+	_ "github.com/influxdata/telegraf/plugins/inputs/interrupts"
 	_ "github.com/influxdata/telegraf/plugins/inputs/ipmi_sensor"
+	_ "github.com/influxdata/telegraf/plugins/inputs/ipset"
 	_ "github.com/influxdata/telegraf/plugins/inputs/iptables"
 	_ "github.com/influxdata/telegraf/plugins/inputs/jolokia"
+	_ "github.com/influxdata/telegraf/plugins/inputs/jolokia2"
 	_ "github.com/influxdata/telegraf/plugins/inputs/kafka_consumer"
+	_ "github.com/influxdata/telegraf/plugins/inputs/kafka_consumer_legacy"
+	_ "github.com/influxdata/telegraf/plugins/inputs/kapacitor"
 	_ "github.com/influxdata/telegraf/plugins/inputs/kubernetes"
 	_ "github.com/influxdata/telegraf/plugins/inputs/leofs"
 	_ "github.com/influxdata/telegraf/plugins/inputs/logparser"
 	_ "github.com/influxdata/telegraf/plugins/inputs/lustre2"
 	_ "github.com/influxdata/telegraf/plugins/inputs/mailchimp"
+	_ "github.com/influxdata/telegraf/plugins/inputs/mcrouter"
 	_ "github.com/influxdata/telegraf/plugins/inputs/memcached"
 	_ "github.com/influxdata/telegraf/plugins/inputs/mesos"
+	_ "github.com/influxdata/telegraf/plugins/inputs/minecraft"
 	_ "github.com/influxdata/telegraf/plugins/inputs/mongodb"
 	_ "github.com/influxdata/telegraf/plugins/inputs/mqtt_consumer"
 	_ "github.com/influxdata/telegraf/plugins/inputs/mysql"
+	_ "github.com/influxdata/telegraf/plugins/inputs/nats"
 	_ "github.com/influxdata/telegraf/plugins/inputs/nats_consumer"
 	_ "github.com/influxdata/telegraf/plugins/inputs/net_response"
 	_ "github.com/influxdata/telegraf/plugins/inputs/nginx"
+	_ "github.com/influxdata/telegraf/plugins/inputs/nginx_plus"
 	_ "github.com/influxdata/telegraf/plugins/inputs/nsq"
 	_ "github.com/influxdata/telegraf/plugins/inputs/nsq_consumer"
 	_ "github.com/influxdata/telegraf/plugins/inputs/nstat"
 	_ "github.com/influxdata/telegraf/plugins/inputs/ntpq"
+	_ "github.com/influxdata/telegraf/plugins/inputs/nvidia_smi"
+	_ "github.com/influxdata/telegraf/plugins/inputs/openldap"
+	_ "github.com/influxdata/telegraf/plugins/inputs/opensmtpd"
 	_ "github.com/influxdata/telegraf/plugins/inputs/passenger"
+	_ "github.com/influxdata/telegraf/plugins/inputs/pf"
 	_ "github.com/influxdata/telegraf/plugins/inputs/phpfpm"
 	_ "github.com/influxdata/telegraf/plugins/inputs/ping"
+	_ "github.com/influxdata/telegraf/plugins/inputs/postfix"
 	_ "github.com/influxdata/telegraf/plugins/inputs/postgresql"
 	_ "github.com/influxdata/telegraf/plugins/inputs/postgresql_extensible"
 	_ "github.com/influxdata/telegraf/plugins/inputs/powerdns"
@@ -63,21 +85,30 @@ import (
 	_ "github.com/influxdata/telegraf/plugins/inputs/redis"
 	_ "github.com/influxdata/telegraf/plugins/inputs/rethinkdb"
 	_ "github.com/influxdata/telegraf/plugins/inputs/riak"
+	_ "github.com/influxdata/telegraf/plugins/inputs/salesforce"
 	_ "github.com/influxdata/telegraf/plugins/inputs/sensors"
+	_ "github.com/influxdata/telegraf/plugins/inputs/smart"
 	_ "github.com/influxdata/telegraf/plugins/inputs/snmp"
 	_ "github.com/influxdata/telegraf/plugins/inputs/snmp_legacy"
+	_ "github.com/influxdata/telegraf/plugins/inputs/socket_listener"
+	_ "github.com/influxdata/telegraf/plugins/inputs/solr"
 	_ "github.com/influxdata/telegraf/plugins/inputs/sqlserver"
 	_ "github.com/influxdata/telegraf/plugins/inputs/statsd"
 	_ "github.com/influxdata/telegraf/plugins/inputs/sysstat"
 	_ "github.com/influxdata/telegraf/plugins/inputs/system"
 	_ "github.com/influxdata/telegraf/plugins/inputs/tail"
 	_ "github.com/influxdata/telegraf/plugins/inputs/tcp_listener"
+	_ "github.com/influxdata/telegraf/plugins/inputs/teamspeak"
+	_ "github.com/influxdata/telegraf/plugins/inputs/tomcat"
 	_ "github.com/influxdata/telegraf/plugins/inputs/trig"
 	_ "github.com/influxdata/telegraf/plugins/inputs/twemproxy"
 	_ "github.com/influxdata/telegraf/plugins/inputs/udp_listener"
+	_ "github.com/influxdata/telegraf/plugins/inputs/unbound"
 	_ "github.com/influxdata/telegraf/plugins/inputs/varnish"
 	_ "github.com/influxdata/telegraf/plugins/inputs/webhooks"
 	_ "github.com/influxdata/telegraf/plugins/inputs/win_perf_counters"
+	_ "github.com/influxdata/telegraf/plugins/inputs/win_services"
 	_ "github.com/influxdata/telegraf/plugins/inputs/zfs"
+	_ "github.com/influxdata/telegraf/plugins/inputs/zipkin"
 	_ "github.com/influxdata/telegraf/plugins/inputs/zookeeper"
 )

plugins/inputs/amqp_consumer/README.md

@@ -0,0 +1,47 @@
+# AMQP Consumer Input Plugin
+
+This plugin provides a consumer for use with AMQP 0-9-1, a promenent implementation of this protocol being [RabbitMQ](https://www.rabbitmq.com/).
+
+Metrics are read from a topic exchange using the configured queue and binding_key.
+
+Message payload should be formatted in one of the [Telegraf Data Formats](https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md).
+
+For an introduction to AMQP see:
+- https://www.rabbitmq.com/tutorials/amqp-concepts.html
+- https://www.rabbitmq.com/getstarted.html
+
+The following defaults are known to work with RabbitMQ:
+
+```toml
+# AMQP consumer plugin
+[[inputs.amqp_consumer]]
+  ## AMQP url
+  url = "amqp://localhost:5672/influxdb"
+  ## AMQP exchange
+  exchange = "telegraf"
+  ## AMQP queue name
+  queue = "telegraf"
+  ## Binding Key
+  binding_key = "#"
+
+  ## Controls how many messages the server will try to keep on the network
+  ## for consumers before receiving delivery acks.
+  #prefetch_count = 50
+
+  ## Auth method. PLAIN and EXTERNAL are supported.
+  ## Using EXTERNAL requires enabling the rabbitmq_auth_mechanism_ssl plugin as
+  ## described here: https://www.rabbitmq.com/plugins.html
+  # auth_method = "PLAIN"
+  ## Optional SSL Config
+  # ssl_ca = "/etc/telegraf/ca.pem"
+  # ssl_cert = "/etc/telegraf/cert.pem"
+  # ssl_key = "/etc/telegraf/key.pem"
+  ## Use SSL but skip chain & host verification
+  # insecure_skip_verify = false
+
+  ## Data format to consume.
+  ## Each data format has its own unique set of configuration options, read
+  ## more about them here:
+  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
+  data_format = "influx"
+```

plugins/inputs/amqp_consumer/amqp_consumer.go

@@ -0,0 +1,282 @@
+package amqp_consumer
+
+import (
+	"fmt"
+	"log"
+	"strings"
+	"sync"
+	"time"
+
+	"github.com/streadway/amqp"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/internal"
+	"github.com/influxdata/telegraf/plugins/inputs"
+	"github.com/influxdata/telegraf/plugins/parsers"
+)
+
+// AMQPConsumer is the top level struct for this plugin
+type AMQPConsumer struct {
+	URL string
+	// AMQP exchange
+	Exchange string
+	// Queue Name
+	Queue string
+	// Binding Key
+	BindingKey string `toml:"binding_key"`
+
+	// Controls how many messages the server will try to keep on the network
+	// for consumers before receiving delivery acks.
+	PrefetchCount int
+
+	// AMQP Auth method
+	AuthMethod string
+	// Path to CA file
+	SSLCA string `toml:"ssl_ca"`
+	// Path to host cert file
+	SSLCert string `toml:"ssl_cert"`
+	// Path to cert key file
+	SSLKey string `toml:"ssl_key"`
+	// Use SSL but skip chain & host verification
+	InsecureSkipVerify bool
+
+	parser parsers.Parser
+	conn   *amqp.Connection
+	wg     *sync.WaitGroup
+}
+
+type externalAuth struct{}
+
+func (a *externalAuth) Mechanism() string {
+	return "EXTERNAL"
+}
+func (a *externalAuth) Response() string {
+	return fmt.Sprintf("\000")
+}
+
+const (
+	DefaultAuthMethod    = "PLAIN"
+	DefaultPrefetchCount = 50
+)
+
+func (a *AMQPConsumer) SampleConfig() string {
+	return `
+  ## AMQP url
+  url = "amqp://localhost:5672/influxdb"
+  ## AMQP exchange
+  exchange = "telegraf"
+  ## AMQP queue name
+  queue = "telegraf"
+  ## Binding Key
+  binding_key = "#"
+
+  ## Maximum number of messages server should give to the worker.
+  prefetch_count = 50
+
+  ## Auth method. PLAIN and EXTERNAL are supported
+  ## Using EXTERNAL requires enabling the rabbitmq_auth_mechanism_ssl plugin as
+  ## described here: https://www.rabbitmq.com/plugins.html
+  # auth_method = "PLAIN"
+
+  ## Optional SSL Config
+  # ssl_ca = "/etc/telegraf/ca.pem"
+  # ssl_cert = "/etc/telegraf/cert.pem"
+  # ssl_key = "/etc/telegraf/key.pem"
+  ## Use SSL but skip chain & host verification
+  # insecure_skip_verify = false
+
+  ## Data format to consume.
+  ## Each data format has its own unique set of configuration options, read
+  ## more about them here:
+  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_INPUT.md
+  data_format = "influx"
+`
+}
+
+func (a *AMQPConsumer) Description() string {
+	return "AMQP consumer plugin"
+}
+
+func (a *AMQPConsumer) SetParser(parser parsers.Parser) {
+	a.parser = parser
+}
+
+// All gathering is done in the Start function
+func (a *AMQPConsumer) Gather(_ telegraf.Accumulator) error {
+	return nil
+}
+
+func (a *AMQPConsumer) createConfig() (*amqp.Config, error) {
+	// make new tls config
+	tls, err := internal.GetTLSConfig(
+		a.SSLCert, a.SSLKey, a.SSLCA, a.InsecureSkipVerify)
+	if err != nil {
+		return nil, err
+	}
+
+	// parse auth method
+	var sasl []amqp.Authentication // nil by default
+
+	if strings.ToUpper(a.AuthMethod) == "EXTERNAL" {
+		sasl = []amqp.Authentication{&externalAuth{}}
+	}
+
+	config := amqp.Config{
+		TLSClientConfig: tls,
+		SASL:            sasl, // if nil, it will be PLAIN
+	}
+	return &config, nil
+}
+
+// Start satisfies the telegraf.ServiceInput interface
+func (a *AMQPConsumer) Start(acc telegraf.Accumulator) error {
+	amqpConf, err := a.createConfig()
+	if err != nil {
+		return err
+	}
+
+	msgs, err := a.connect(amqpConf)
+	if err != nil {
+		return err
+	}
+
+	a.wg = &sync.WaitGroup{}
+	a.wg.Add(1)
+	go a.process(msgs, acc)
+
+	go func() {
+		for {
+			err := <-a.conn.NotifyClose(make(chan *amqp.Error))
+			if err == nil {
+				break
+			}
+
+			log.Printf("I! AMQP consumer connection closed: %s; trying to reconnect", err)
+			for {
+				msgs, err := a.connect(amqpConf)
+				if err != nil {
+					log.Printf("E! AMQP connection failed: %s", err)
+					time.Sleep(10 * time.Second)
+					continue
+				}
+
+				a.wg.Add(1)
+				go a.process(msgs, acc)
+				break
+			}
+		}
+	}()
+
+	return nil
+}
+
+func (a *AMQPConsumer) connect(amqpConf *amqp.Config) (<-chan amqp.Delivery, error) {
+	conn, err := amqp.DialConfig(a.URL, *amqpConf)
+	if err != nil {
+		return nil, err
+	}
+	a.conn = conn
+
+	ch, err := conn.Channel()
+	if err != nil {
+		return nil, fmt.Errorf("Failed to open a channel: %s", err)
+	}
+
+	err = ch.ExchangeDeclare(
+		a.Exchange, // name
+		"topic",    // type
+		true,       // durable
+		false,      // auto-deleted
+		false,      // internal
+		false,      // no-wait
+		nil,        // arguments
+	)
+	if err != nil {
+		return nil, fmt.Errorf("Failed to declare an exchange: %s", err)
+	}
+
+	q, err := ch.QueueDeclare(
+		a.Queue, // queue
+		true,    // durable
+		false,   // delete when unused
+		false,   // exclusive
+		false,   // no-wait
+		nil,     // arguments
+	)
+	if err != nil {
+		return nil, fmt.Errorf("Failed to declare a queue: %s", err)
+	}
+
+	err = ch.QueueBind(
+		q.Name,       // queue
+		a.BindingKey, // binding-key
+		a.Exchange,   // exchange
+		false,
+		nil,
+	)
+	if err != nil {
+		return nil, fmt.Errorf("Failed to bind a queue: %s", err)
+	}
+
+	err = ch.Qos(
+		a.PrefetchCount,
+		0,     // prefetch-size
+		false, // global
+	)
+	if err != nil {
+		return nil, fmt.Errorf("Failed to set QoS: %s", err)
+	}
+
+	msgs, err := ch.Consume(
+		q.Name, // queue
+		"",     // consumer
+		false,  // auto-ack
+		false,  // exclusive
+		false,  // no-local
+		false,  // no-wait
+		nil,    // arguments
+	)
+	if err != nil {
+		return nil, fmt.Errorf("Failed establishing connection to queue: %s", err)
+	}
+
+	log.Println("I! Started AMQP consumer")
+	return msgs, err
+}
+
+// Read messages from queue and add them to the Accumulator
+func (a *AMQPConsumer) process(msgs <-chan amqp.Delivery, acc telegraf.Accumulator) {
+	defer a.wg.Done()
+	for d := range msgs {
+		metrics, err := a.parser.Parse(d.Body)
+		if err != nil {
+			log.Printf("E! %v: error parsing metric - %v", err, string(d.Body))
+		} else {
+			for _, m := range metrics {
+				acc.AddFields(m.Name(), m.Fields(), m.Tags(), m.Time())
+			}
+		}
+
+		d.Ack(false)
+	}
+	log.Printf("I! AMQP consumer queue closed")
+}
+
+func (a *AMQPConsumer) Stop() {
+	err := a.conn.Close()
+	if err != nil && err != amqp.ErrClosed {
+		log.Printf("E! Error closing AMQP connection: %s", err)
+		return
+	}
+	a.wg.Wait()
+	log.Println("I! Stopped AMQP service")
+}
+
+func init() {
+	inputs.Add("amqp_consumer", func() telegraf.Input {
+		return &AMQPConsumer{
+			AuthMethod:    DefaultAuthMethod,
+			PrefetchCount: DefaultPrefetchCount,
+		}
+	})
+}

plugins/inputs/apache/README.md

@@ -1,55 +1,84 @@
-# Telegraf plugin: Apache
+# Apache Input Plugin
 
-#### Plugin arguments:
-- **urls** []string: List of apache-status URLs to collect from. Default is "http://localhost/server-status?auto".
-- **username** string: Username for HTTP basic authentication
-- **password** string: Password for HTTP basic authentication
-- **timeout** duration: time that the HTTP connection will remain waiting for response. Defalt 4 seconds ("4s")
+The Apache plugin collects server performance information using the [`mod_status`](https://httpd.apache.org/docs/2.4/mod/mod_status.html) module of the [Apache HTTP Server](https://httpd.apache.org/).
 
-##### Optional SSL Config
+Typically, the `mod_status` module is configured to expose a page at the `/server-status?auto` location of the Apache server.  The [ExtendedStatus](https://httpd.apache.org/docs/2.4/mod/core.html#extendedstatus) option must be enabled in order to collect all available fields.  For information about how to configure your server reference the [module documentation](https://httpd.apache.org/docs/2.4/mod/mod_status.html#enable).
 
-- **ssl_ca** string: the full path for the SSL CA certicate
-- **ssl_cert** string: the full path for the SSL certificate
-- **ssl_key** string: the full path for the key file
-- **insecure_skip_verify** bool: if true HTTP client will skip all SSL verifications related to peer and host. Default to false
+### Configuration:
 
-#### Description
+```toml
+# Read Apache status information (mod_status)
+[[inputs.apache]]
+  ## An array of URLs to gather from, must be directed at the machine
+  ## readable version of the mod_status page including the auto query string.
+  ## Default is "http://localhost/server-status?auto".
+  urls = ["http://localhost/server-status?auto"]
 
-The Apache plugin collects from the /server-status?auto URL. See
-[apache.org/server-status?auto](http://www.apache.org/server-status?auto) for an
-example. And
-[here](http://httpd.apache.org/docs/2.2/mod/mod_status.html) for the apache
-mod_status documentation.
+  ## Credentials for basic HTTP authentication.
+  # username = "myuser"
+  # password = "mypassword"
 
-# Measurements:
+  ## Maximum time to receive response.
+  # response_timeout = "5s"
 
-Meta:
-- tags: `port=<port>`, `server=url`
+  ## Optional SSL Config
+  # ssl_ca = "/etc/telegraf/ca.pem"
+  # ssl_cert = "/etc/telegraf/cert.pem"
+  # ssl_key = "/etc/telegraf/key.pem"
+  ## Use SSL but skip chain & host verification
+  # insecure_skip_verify = false
+```
 
-- apache_TotalAccesses
-- apache_TotalkBytes
-- apache_CPULoad
-- apache_Uptime
-- apache_ReqPerSec
-- apache_BytesPerSec
-- apache_BytesPerReq
-- apache_BusyWorkers
-- apache_IdleWorkers
-- apache_ConnsTotal
-- apache_ConnsAsyncWriting
-- apache_ConnsAsyncKeepAlive
-- apache_ConnsAsyncClosing
+### Measurements & Fields:
 
-### Scoreboard measurements
+- apache
+  - BusyWorkers (float)
+  - BytesPerReq (float)
+  - BytesPerSec (float)
+  - ConnsAsyncClosing (float)
+  - ConnsAsyncKeepAlive (float)
+  - ConnsAsyncWriting (float)
+  - ConnsTotal (float)
+  - CPUChildrenSystem (float)
+  - CPUChildrenUser (float)
+  - CPULoad (float)
+  - CPUSystem (float)
+  - CPUUser (float)
+  - IdleWorkers (float)
+  - Load1 (float)
+  - Load5 (float)
+  - Load15 (float)
+  - ParentServerConfigGeneration (float)
+  - ParentServerMPMGeneration (float)
+  - ReqPerSec (float)
+  - ServerUptimeSeconds (float)
+  - TotalAccesses (float)
+  - TotalkBytes (float)
+  - Uptime (float)
 
-- apache_scboard_waiting
-- apache_scboard_starting
-- apache_scboard_reading
-- apache_scboard_sending
-- apache_scboard_keepalive
-- apache_scboard_dnslookup
-- apache_scboard_closing
-- apache_scboard_logging
-- apache_scboard_finishing
-- apache_scboard_idle_cleanup
-- apache_scboard_open
+The following fields are collected from the `Scoreboard`, and represent the number of requests in the given state:
+
+- apache
+  - scboard_closing (float)
+  - scboard_dnslookup (float)
+  - scboard_finishing (float)
+  - scboard_idle_cleanup (float)
+  - scboard_keepalive (float)
+  - scboard_logging (float)
+  - scboard_open (float)
+  - scboard_reading (float)
+  - scboard_sending (float)
+  - scboard_starting (float)
+  - scboard_waiting (float)
+
+### Tags:
+
+- All measurements have the following tags:
+    - port
+    - server
+
+### Example Output:
+
+```
+apache,port=80,server=debian-stretch-apache BusyWorkers=1,BytesPerReq=0,BytesPerSec=0,CPUChildrenSystem=0,CPUChildrenUser=0,CPULoad=0.00995025,CPUSystem=0.01,CPUUser=0.01,ConnsAsyncClosing=0,ConnsAsyncKeepAlive=0,ConnsAsyncWriting=0,ConnsTotal=0,IdleWorkers=49,Load1=0.01,Load15=0,Load5=0,ParentServerConfigGeneration=3,ParentServerMPMGeneration=2,ReqPerSec=0.00497512,ServerUptimeSeconds=201,TotalAccesses=1,TotalkBytes=0,Uptime=201,scboard_closing=0,scboard_dnslookup=0,scboard_finishing=0,scboard_idle_cleanup=0,scboard_keepalive=0,scboard_logging=0,scboard_open=100,scboard_reading=0,scboard_sending=1,scboard_starting=0,scboard_waiting=49 1502489900000000000
+```

plugins/inputs/apache/apache.go

@@ -8,6 +8,7 @@ import (
 	"net/url"
 	"strconv"
 	"strings"
+	"sync"
 	"time"
 
 	"github.com/influxdata/telegraf"
@@ -28,18 +29,22 @@ type Apache struct {
 	SSLKey string `toml:"ssl_key"`
 	// Use SSL but skip chain & host verification
 	InsecureSkipVerify bool
+
+	client *http.Client
 }
 
 var sampleConfig = `
-  ## An array of Apache status URI to gather stats.
+  ## An array of URLs to gather from, must be directed at the machine
+  ## readable version of the mod_status page including the auto query string.
   ## Default is "http://localhost/server-status?auto".
   urls = ["http://localhost/server-status?auto"]
-  ## user credentials for basic HTTP authentication
-  username = "myuser"
-  password = "mypassword"
 
-  ## Timeout to the complete conection and reponse time in seconds
-  response_timeout = "25s" ## default to 5 seconds
+  ## Credentials for basic HTTP authentication.
+  # username = "myuser"
+  # password = "mypassword"
+
+  ## Maximum time to receive response.
+  # response_timeout = "5s"
 
   ## Optional SSL Config
   # ssl_ca = "/etc/telegraf/ca.pem"
@@ -65,55 +70,51 @@ func (n *Apache) Gather(acc telegraf.Accumulator) error {
 		n.ResponseTimeout.Duration = time.Second * 5
 	}
 
-	var outerr error
-	var errch = make(chan error)
-
-	for _, u := range n.Urls {
-		addr, err := url.Parse(u)
-		if err != nil {
-			return fmt.Errorf("Unable to parse address '%s': %s", u, err)
-		}
-
-		go func(addr *url.URL) {
-			errch <- n.gatherUrl(addr, acc)
-		}(addr)
-	}
-
-	// Drain channel, waiting for all requests to finish and save last error.
-	for range n.Urls {
-		if err := <-errch; err != nil {
-			outerr = err
-		}
-	}
-
-	return outerr
-}
-
-func (n *Apache) gatherUrl(addr *url.URL, acc telegraf.Accumulator) error {
-
-	var tr *http.Transport
-
-	if addr.Scheme == "https" {
-		tlsCfg, err := internal.GetTLSConfig(
-			n.SSLCert, n.SSLKey, n.SSLCA, n.InsecureSkipVerify)
+	if n.client == nil {
+		client, err := n.createHttpClient()
 		if err != nil {
 			return err
 		}
-		tr = &http.Transport{
-			ResponseHeaderTimeout: time.Duration(3 * time.Second),
-			TLSClientConfig:       tlsCfg,
-		}
-	} else {
-		tr = &http.Transport{
-			ResponseHeaderTimeout: time.Duration(3 * time.Second),
+		n.client = client
+	}
+
+	var wg sync.WaitGroup
+	wg.Add(len(n.Urls))
+	for _, u := range n.Urls {
+		addr, err := url.Parse(u)
+		if err != nil {
+			acc.AddError(fmt.Errorf("Unable to parse address '%s': %s", u, err))
+			continue
 		}
+
+		go func(addr *url.URL) {
+			defer wg.Done()
+			acc.AddError(n.gatherUrl(addr, acc))
+		}(addr)
+	}
+
+	wg.Wait()
+	return nil
+}
+
+func (n *Apache) createHttpClient() (*http.Client, error) {
+	tlsCfg, err := internal.GetTLSConfig(
+		n.SSLCert, n.SSLKey, n.SSLCA, n.InsecureSkipVerify)
+	if err != nil {
+		return nil, err
 	}
 
 	client := &http.Client{
-		Transport: tr,
-		Timeout:   n.ResponseTimeout.Duration,
+		Transport: &http.Transport{
+			TLSClientConfig: tlsCfg,
+		},
+		Timeout: n.ResponseTimeout.Duration,
 	}
 
+	return client, nil
+}
+
+func (n *Apache) gatherUrl(addr *url.URL, acc telegraf.Accumulator) error {
 	req, err := http.NewRequest("GET", addr.String(), nil)
 	if err != nil {
 		return fmt.Errorf("error on new request to %s : %s\n", addr.String(), err)
@@ -123,7 +124,7 @@ func (n *Apache) gatherUrl(addr *url.URL, acc telegraf.Accumulator) error {
 		req.SetBasicAuth(n.Username, n.Password)
 	}
 
-	resp, err := client.Do(req)
+	resp, err := n.client.Do(req)
 	if err != nil {
 		return fmt.Errorf("error on request to %s : %s\n", addr.String(), err)
 	}

plugins/inputs/apache/apache_test.go

@@ -41,7 +41,7 @@ func TestHTTPApache(t *testing.T) {
 	}
 
 	var acc testutil.Accumulator
-	err := a.Gather(&acc)
+	err := acc.GatherError(a.Gather)
 	require.NoError(t, err)
 
 	fields := map[string]interface{}{

plugins/inputs/bcache/README.md

@@ -70,7 +70,7 @@ Using this configuration:
 When run with:
 
 ```
-./telegraf -config telegraf.conf -input-filter bcache -test
+./telegraf --config telegraf.conf --input-filter bcache --test
 ```
 
 It produces:

plugins/inputs/bond/README.md

@@ -0,0 +1,85 @@
+# Bond Input Plugin
+
+The Bond input plugin collects network bond interface status for both the
+network bond interface as well as slave interfaces.
+The plugin collects these metrics from `/proc/net/bonding/*` files.
+
+### Configuration:
+
+```toml
+[[inputs.bond]]
+  ## Sets 'proc' directory path
+  ## If not specified, then default is /proc
+  # host_proc = "/proc"
+
+  ## By default, telegraf gather stats for all bond interfaces
+  ## Setting interfaces will restrict the stats to the specified
+  ## bond interfaces.
+  # bond_interfaces = ["bond0"]
+```
+
+### Measurements & Fields:
+
+- bond
+  - active_slave (for active-backup mode)
+  - status
+
+- bond_slave
+  - failures
+  - status
+
+### Description:
+
+```
+active_slave
+  Currently active slave interface for active-backup mode.
+
+status
+  Status of bond interface or bonds's slave interface (down = 0, up = 1).
+
+failures
+  Amount of failures for bond's slave interface.
+```
+
+### Tags:
+
+- bond
+  - bond
+
+- bond_slave
+  - bond
+  - interface
+
+### Example output:
+
+Configuration:
+
+```
+[[inputs.bond]]
+  ## Sets 'proc' directory path
+  ## If not specified, then default is /proc
+  host_proc = "/proc"
+
+  ## By default, telegraf gather stats for all bond interfaces
+  ## Setting interfaces will restrict the stats to the specified
+  ## bond interfaces.
+  bond_interfaces = ["bond0", "bond1"]
+```
+
+Run:
+
+```
+telegraf --config telegraf.conf --input-filter bond --test
+```
+
+Output:
+
+```
+* Plugin: inputs.bond, Collection 1
+> bond,bond=bond1,host=local active_slave="eth0",status=1i 1509704525000000000
+> bond_slave,bond=bond1,interface=eth0,host=local status=1i,failures=0i 1509704525000000000
+> bond_slave,host=local,bond=bond1,interface=eth1 status=1i,failures=0i 1509704525000000000
+> bond,bond=bond0,host=isvetlov-mac.local status=1i 1509704525000000000
+> bond_slave,bond=bond0,interface=eth1,host=local status=1i,failures=0i 1509704525000000000
+> bond_slave,bond=bond0,interface=eth2,host=local status=1i,failures=0i 1509704525000000000
+```

plugins/inputs/bond/bond.go

@@ -0,0 +1,204 @@
+package bond
+
+import (
+	"bufio"
+	"fmt"
+	"io/ioutil"
+	"os"
+	"path/filepath"
+	"strconv"
+	"strings"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/inputs"
+)
+
+// default host proc path
+const defaultHostProc = "/proc"
+
+// env host proc variable name
+const envProc = "HOST_PROC"
+
+type Bond struct {
+	HostProc       string   `toml:"host_proc"`
+	BondInterfaces []string `toml:"bond_interfaces"`
+}
+
+var sampleConfig = `
+  ## Sets 'proc' directory path
+  ## If not specified, then default is /proc
+  # host_proc = "/proc"
+
+  ## By default, telegraf gather stats for all bond interfaces
+  ## Setting interfaces will restrict the stats to the specified
+  ## bond interfaces.
+  # bond_interfaces = ["bond0"]
+`
+
+func (bond *Bond) Description() string {
+	return "Collect bond interface status, slaves statuses and failures count"
+}
+
+func (bond *Bond) SampleConfig() string {
+	return sampleConfig
+}
+
+func (bond *Bond) Gather(acc telegraf.Accumulator) error {
+	// load proc path, get default value if config value and env variable are empty
+	bond.loadPath()
+	// list bond interfaces from bonding directory or gather all interfaces.
+	bondNames, err := bond.listInterfaces()
+	if err != nil {
+		return err
+	}
+	for _, bondName := range bondNames {
+		bondAbsPath := bond.HostProc + "/net/bonding/" + bondName
+		file, err := ioutil.ReadFile(bondAbsPath)
+		if err != nil {
+			acc.AddError(fmt.Errorf("error inspecting '%s' interface: %v", bondAbsPath, err))
+			continue
+		}
+		rawFile := strings.TrimSpace(string(file))
+		err = bond.gatherBondInterface(bondName, rawFile, acc)
+		if err != nil {
+			acc.AddError(fmt.Errorf("error inspecting '%s' interface: %v", bondName, err))
+		}
+	}
+	return nil
+}
+
+func (bond *Bond) gatherBondInterface(bondName string, rawFile string, acc telegraf.Accumulator) error {
+	splitIndex := strings.Index(rawFile, "Slave Interface:")
+	if splitIndex == -1 {
+		splitIndex = len(rawFile)
+	}
+	bondPart := rawFile[:splitIndex]
+	slavePart := rawFile[splitIndex:]
+
+	err := bond.gatherBondPart(bondName, bondPart, acc)
+	if err != nil {
+		return err
+	}
+	err = bond.gatherSlavePart(bondName, slavePart, acc)
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+func (bond *Bond) gatherBondPart(bondName string, rawFile string, acc telegraf.Accumulator) error {
+	fields := make(map[string]interface{})
+	tags := map[string]string{
+		"bond": bondName,
+	}
+
+	scanner := bufio.NewScanner(strings.NewReader(rawFile))
+	for scanner.Scan() {
+		line := scanner.Text()
+		stats := strings.Split(line, ":")
+		if len(stats) < 2 {
+			continue
+		}
+		name := strings.TrimSpace(stats[0])
+		value := strings.TrimSpace(stats[1])
+		if strings.Contains(name, "Currently Active Slave") {
+			fields["active_slave"] = value
+		}
+		if strings.Contains(name, "MII Status") {
+			fields["status"] = 0
+			if value == "up" {
+				fields["status"] = 1
+			}
+			acc.AddFields("bond", fields, tags)
+			return nil
+		}
+	}
+	if err := scanner.Err(); err != nil {
+		return err
+	}
+	return fmt.Errorf("Couldn't find status info for '%s' ", bondName)
+}
+
+func (bond *Bond) gatherSlavePart(bondName string, rawFile string, acc telegraf.Accumulator) error {
+	var slave string
+	var status int
+
+	scanner := bufio.NewScanner(strings.NewReader(rawFile))
+	for scanner.Scan() {
+		line := scanner.Text()
+		stats := strings.Split(line, ":")
+		if len(stats) < 2 {
+			continue
+		}
+		name := strings.TrimSpace(stats[0])
+		value := strings.TrimSpace(stats[1])
+		if strings.Contains(name, "Slave Interface") {
+			slave = value
+		}
+		if strings.Contains(name, "MII Status") {
+			status = 0
+			if value == "up" {
+				status = 1
+			}
+		}
+		if strings.Contains(name, "Link Failure Count") {
+			count, err := strconv.Atoi(value)
+			if err != nil {
+				return err
+			}
+			fields := map[string]interface{}{
+				"status":   status,
+				"failures": count,
+			}
+			tags := map[string]string{
+				"bond":      bondName,
+				"interface": slave,
+			}
+			acc.AddFields("bond_slave", fields, tags)
+		}
+	}
+	if err := scanner.Err(); err != nil {
+		return err
+	}
+	return nil
+}
+
+// loadPath can be used to read path firstly from config
+// if it is empty then try read from env variable
+func (bond *Bond) loadPath() {
+	if bond.HostProc == "" {
+		bond.HostProc = proc(envProc, defaultHostProc)
+	}
+}
+
+// proc can be used to read file paths from env
+func proc(env, path string) string {
+	// try to read full file path
+	if p := os.Getenv(env); p != "" {
+		return p
+	}
+	// return default path
+	return path
+}
+
+func (bond *Bond) listInterfaces() ([]string, error) {
+	var interfaces []string
+	if len(bond.BondInterfaces) > 0 {
+		interfaces = bond.BondInterfaces
+	} else {
+		paths, err := filepath.Glob(bond.HostProc + "/net/bonding/*")
+		if err != nil {
+			return nil, err
+		}
+		for _, p := range paths {
+			interfaces = append(interfaces, filepath.Base(p))
+		}
+	}
+	return interfaces, nil
+}
+
+func init() {
+	inputs.Add("bond", func() telegraf.Input {
+		return &Bond{}
+	})
+}

plugins/inputs/bond/bond_test.go

@@ -0,0 +1,77 @@
+package bond
+
+import (
+	"testing"
+
+	"github.com/influxdata/telegraf/testutil"
+)
+
+var sampleTest802 = `
+Ethernet Channel Bonding Driver: v3.5.0 (November 4, 2008)
+
+Bonding Mode: IEEE 802.3ad Dynamic link aggregation
+Transmit Hash Policy: layer2 (0)
+MII Status: up
+MII Polling Interval (ms): 100
+Up Delay (ms): 0
+Down Delay (ms): 0
+
+802.3ad info
+LACP rate: fast
+Aggregator selection policy (ad_select): stable
+bond bond0 has no active aggregator
+
+Slave Interface: eth1
+MII Status: up
+Link Failure Count: 0
+Permanent HW addr: 00:0c:29:f5:b7:11
+Aggregator ID: N/A
+
+Slave Interface: eth2
+MII Status: up
+Link Failure Count: 3
+Permanent HW addr: 00:0c:29:f5:b7:1b
+Aggregator ID: N/A
+`
+
+var sampleTestAB = `
+Ethernet Channel Bonding Driver: v3.6.0 (September 26, 2009)
+
+Bonding Mode: fault-tolerance (active-backup)
+Primary Slave: eth2 (primary_reselect always)
+Currently Active Slave: eth2
+MII Status: up
+MII Polling Interval (ms): 100
+Up Delay (ms): 0
+Down Delay (ms): 0
+
+Slave Interface: eth3
+MII Status: down
+Speed: 1000 Mbps
+Duplex: full
+Link Failure Count: 2
+Permanent HW addr:
+Slave queue ID: 0
+
+Slave Interface: eth2
+MII Status: up
+Speed: 100 Mbps
+Duplex: full
+Link Failure Count: 0
+Permanent HW addr:
+`
+
+func TestGatherBondInterface(t *testing.T) {
+	var acc testutil.Accumulator
+	bond := &Bond{}
+
+	bond.gatherBondInterface("bond802", sampleTest802, &acc)
+	acc.AssertContainsTaggedFields(t, "bond", map[string]interface{}{"status": 1}, map[string]string{"bond": "bond802"})
+	acc.AssertContainsTaggedFields(t, "bond_slave", map[string]interface{}{"failures": 0, "status": 1}, map[string]string{"bond": "bond802", "interface": "eth1"})
+	acc.AssertContainsTaggedFields(t, "bond_slave", map[string]interface{}{"failures": 3, "status": 1}, map[string]string{"bond": "bond802", "interface": "eth2"})
+
+	bond.gatherBondInterface("bondAB", sampleTestAB, &acc)
+	acc.AssertContainsTaggedFields(t, "bond", map[string]interface{}{"active_slave": "eth2", "status": 1}, map[string]string{"bond": "bondAB"})
+	acc.AssertContainsTaggedFields(t, "bond_slave", map[string]interface{}{"failures": 2, "status": 0}, map[string]string{"bond": "bondAB", "interface": "eth3"})
+	acc.AssertContainsTaggedFields(t, "bond_slave", map[string]interface{}{"failures": 0, "status": 1}, map[string]string{"bond": "bondAB", "interface": "eth2"})
+}

plugins/inputs/cassandra/README.md

@@ -1,5 +1,8 @@
+
 # Telegraf plugin: Cassandra
 
+### **Deprecated in version 1.7**: Please use the [jolokia2](https://github.com/influxdata/telegraf/tree/master/plugins/inputs/jolokia2) plugin with the [cassandra.conf](/plugins/inputs/jolokia2/examples/cassandra.conf) example configuration.
+
 #### Plugin arguments:
 - **context** string: Context root used for jolokia url
 - **servers** []string: List of servers with the format "<user:passwd@><host>:port"

plugins/inputs/cassandra/cassandra.go

@@ -4,13 +4,14 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
-	"github.com/influxdata/telegraf"
-	"github.com/influxdata/telegraf/plugins/inputs"
 	"io/ioutil"
 	"log"
 	"net/http"
 	"net/url"
 	"strings"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/inputs"
 )
 
 type JolokiaClient interface {
@@ -61,7 +62,8 @@ func newCassandraMetric(host string, metric string,
 func addValuesAsFields(values map[string]interface{}, fields map[string]interface{},
 	mname string) {
 	for k, v := range values {
-		if v != nil {
+		switch v.(type) {
+		case int64, float64, string, bool:
 			fields[mname+"_"+k] = v
 		}
 	}
@@ -118,13 +120,13 @@ func (j javaMetric) addTagsFields(out map[string]interface{}) {
 		switch t := values.(type) {
 		case map[string]interface{}:
 			addValuesAsFields(values.(map[string]interface{}), fields, attribute)
-		case interface{}:
+		case int64, float64, string, bool:
 			fields[attribute] = t
 		}
 		j.acc.AddFields(tokens["class"]+tokens["type"], fields, tags)
 	} else {
-		fmt.Printf("Missing key 'value' in '%s' output response\n%v\n",
-			j.metric, out)
+		j.acc.AddError(fmt.Errorf("Missing key 'value' in '%s' output response\n%v\n",
+			j.metric, out))
 	}
 }
 
@@ -155,8 +157,8 @@ func (c cassandraMetric) addTagsFields(out map[string]interface{}) {
 				addCassandraMetric(k, c, v.(map[string]interface{}))
 			}
 		} else {
-			fmt.Printf("Missing key 'value' in '%s' output response\n%v\n",
-				c.metric, out)
+			c.acc.AddError(fmt.Errorf("Missing key 'value' in '%s' output response\n%v\n",
+				c.metric, out))
 			return
 		}
 	} else {
@@ -164,8 +166,8 @@ func (c cassandraMetric) addTagsFields(out map[string]interface{}) {
 			addCassandraMetric(r.(map[string]interface{})["mbean"].(string),
 				c, values.(map[string]interface{}))
 		} else {
-			fmt.Printf("Missing key 'value' in '%s' output response\n%v\n",
-				c.metric, out)
+			c.acc.AddError(fmt.Errorf("Missing key 'value' in '%s' output response\n%v\n",
+				c.metric, out))
 			return
 		}
 	}
@@ -173,7 +175,11 @@ func (c cassandraMetric) addTagsFields(out map[string]interface{}) {
 
 func (j *Cassandra) SampleConfig() string {
 	return `
-  # This is the context root used to compose the jolokia url
+  ## DEPRECATED: The cassandra plugin has been deprecated.  Please use the
+  ## jolokia2 plugin instead.
+  ##
+  ## see https://github.com/influxdata/telegraf/tree/master/plugins/inputs/jolokia2
+
   context = "/jolokia/read"
   ## List of cassandra servers exposing jolokia read service
   servers = ["myuser:mypassword@10.10.10.1:8778","10.10.10.2:8778",":8778"]
@@ -257,6 +263,16 @@ func parseServerTokens(server string) map[string]string {
 	return serverTokens
 }
 
+func (c *Cassandra) Start(acc telegraf.Accumulator) error {
+	log.Println("W! DEPRECATED: The cassandra plugin has been deprecated. " +
+		"Please use the jolokia2 plugin instead. " +
+		"https://github.com/influxdata/telegraf/tree/master/plugins/inputs/jolokia2")
+	return nil
+}
+
+func (c *Cassandra) Stop() {
+}
+
 func (c *Cassandra) Gather(acc telegraf.Accumulator) error {
 	context := c.Context
 	servers := c.Servers
@@ -274,8 +290,8 @@ func (c *Cassandra) Gather(acc telegraf.Accumulator) error {
 				m = newCassandraMetric(serverTokens["host"], metric, acc)
 			} else {
 				// unsupported metric type
-				log.Printf("I! Unsupported Cassandra metric [%s], skipping",
-					metric)
+				acc.AddError(fmt.Errorf("E! Unsupported Cassandra metric [%s], skipping",
+					metric))
 				continue
 			}
 
@@ -283,7 +299,8 @@ func (c *Cassandra) Gather(acc telegraf.Accumulator) error {
 			requestUrl, err := url.Parse("http://" + serverTokens["host"] + ":" +
 				serverTokens["port"] + context + metric)
 			if err != nil {
-				return err
+				acc.AddError(err)
+				continue
 			}
 			if serverTokens["user"] != "" && serverTokens["passwd"] != "" {
 				requestUrl.User = url.UserPassword(serverTokens["user"],
@@ -291,8 +308,12 @@ func (c *Cassandra) Gather(acc telegraf.Accumulator) error {
 			}
 
 			out, err := c.getAttr(requestUrl)
+			if err != nil {
+				acc.AddError(err)
+				continue
+			}
 			if out["status"] != 200.0 {
-				fmt.Printf("URL returned with status %v\n", out["status"])
+				acc.AddError(fmt.Errorf("URL returned with status %v - %s\n", out["status"], requestUrl))
 				continue
 			}
 			m.addTagsFields(out)

plugins/inputs/cassandra/cassandra_test.go

@@ -151,7 +151,7 @@ func TestHttpJsonJavaMultiValue(t *testing.T) {
 
 	var acc testutil.Accumulator
 	acc.SetDebug(true)
-	err := cassandra.Gather(&acc)
+	err := acc.GatherError(cassandra.Gather)
 
 	assert.Nil(t, err)
 	assert.Equal(t, 2, len(acc.Metrics))
@@ -180,7 +180,7 @@ func TestHttpJsonJavaMultiType(t *testing.T) {
 
 	var acc testutil.Accumulator
 	acc.SetDebug(true)
-	err := cassandra.Gather(&acc)
+	err := acc.GatherError(cassandra.Gather)
 
 	assert.Nil(t, err)
 	assert.Equal(t, 2, len(acc.Metrics))
@@ -197,16 +197,17 @@ func TestHttpJsonJavaMultiType(t *testing.T) {
 }
 
 // Test that the proper values are ignored or collected
-func TestHttpJsonOn404(t *testing.T) {
+func TestHttp404(t *testing.T) {
 
-	jolokia := genJolokiaClientStub(validJavaMultiValueJSON, 404, Servers,
+	jolokia := genJolokiaClientStub(invalidJSON, 404, Servers,
 		[]string{HeapMetric})
 
 	var acc testutil.Accumulator
-	err := jolokia.Gather(&acc)
+	err := acc.GatherError(jolokia.Gather)
 
-	assert.Nil(t, err)
+	assert.Error(t, err)
 	assert.Equal(t, 0, len(acc.Metrics))
+	assert.Contains(t, err.Error(), "has status code 404")
 }
 
 // Test that the proper values are ignored or collected for class=Cassandra
@@ -214,7 +215,7 @@ func TestHttpJsonCassandraMultiValue(t *testing.T) {
 	cassandra := genJolokiaClientStub(validCassandraMultiValueJSON, 200, Servers, []string{ReadLatencyMetric})
 
 	var acc testutil.Accumulator
-	err := cassandra.Gather(&acc)
+	err := acc.GatherError(cassandra.Gather)
 
 	assert.Nil(t, err)
 	assert.Equal(t, 1, len(acc.Metrics))
@@ -246,7 +247,7 @@ func TestHttpJsonCassandraNestedMultiValue(t *testing.T) {
 
 	var acc testutil.Accumulator
 	acc.SetDebug(true)
-	err := cassandra.Gather(&acc)
+	err := acc.GatherError(cassandra.Gather)
 
 	assert.Nil(t, err)
 	assert.Equal(t, 2, len(acc.Metrics))

plugins/inputs/ceph/README.md

@@ -117,7 +117,7 @@ All fields are collected under the **ceph** measurement and stored as float64s.
   * recovering\_objects\_per\_sec (float)
 
 * ceph\_pgmap\_state
-  * state name e.g. active+clean (float)
+  * count (float)
 
 * ceph\_usage
   * bytes\_used (float)
@@ -186,7 +186,7 @@ All measurements will have the following tags:
 
 *Cluster Stats*
 
-* ceph\_pg\_state has the following tags:
+* ceph\_pgmap\_state has the following tags:
   * state (state for which the value applies e.g. active+clean, active+remapped+backfill)
 * ceph\_pool\_usage has the following tags:
   * id
@@ -200,7 +200,7 @@ All measurements will have the following tags:
 *Admin Socket Stats*
 
 <pre>
-telegraf -test -config /etc/telegraf/telegraf.conf -config-directory /etc/telegraf/telegraf.d  -input-filter ceph
+telegraf --config /etc/telegraf/telegraf.conf --config-directory /etc/telegraf/telegraf.d --input-filter ceph --test
 * Plugin: ceph, Collection 1
 > ceph,collection=paxos, id=node-2,role=openstack,type=mon accept_timeout=0,begin=14931264,begin_bytes.avgcount=14931264,begin_bytes.sum=180309683362,begin_keys.avgcount=0,begin_keys.sum=0,begin_latency.avgcount=14931264,begin_latency.sum=9293.29589,collect=1,collect_bytes.avgcount=1,collect_bytes.sum=24,collect_keys.avgcount=1,collect_keys.sum=1,collect_latency.avgcount=1,collect_latency.sum=0.00028,collect_timeout=0,collect_uncommitted=0,commit=14931264,commit_bytes.avgcount=0,commit_bytes.sum=0,commit_keys.avgcount=0,commit_keys.sum=0,commit_latency.avgcount=0,commit_latency.sum=0,lease_ack_timeout=0,lease_timeout=0,new_pn=0,new_pn_latency.avgcount=0,new_pn_latency.sum=0,refresh=14931264,refresh_latency.avgcount=14931264,refresh_latency.sum=8706.98498,restart=4,share_state=0,share_state_bytes.avgcount=0,share_state_bytes.sum=0,share_state_keys.avgcount=0,share_state_keys.sum=0,start_leader=0,start_peon=1,store_state=14931264,store_state_bytes.avgcount=14931264,store_state_bytes.sum=353119959211,store_state_keys.avgcount=14931264,store_state_keys.sum=289807523,store_state_latency.avgcount=14931264,store_state_latency.sum=10952.835724 1462821234814535148
 > ceph,collection=throttle-mon_client_bytes,id=node-2,type=mon get=1413017,get_or_fail_fail=0,get_or_fail_success=0,get_sum=71211705,max=104857600,put=1413013,put_sum=71211459,take=0,take_sum=0,val=246,wait.avgcount=0,wait.sum=0 1462821234814737219
@@ -213,7 +213,8 @@ telegraf -test -config /etc/telegraf/telegraf.conf -config-directory /etc/telegr
 <pre>
 > ceph_osdmap,host=ceph-mon-0 epoch=170772,full=false,nearfull=false,num_in_osds=340,num_osds=340,num_remapped_pgs=0,num_up_osds=340 1468841037000000000
 > ceph_pgmap,host=ceph-mon-0 bytes_avail=634895531270144,bytes_total=812117151809536,bytes_used=177221620539392,data_bytes=56979991615058,num_pgs=22952,op_per_sec=15869,read_bytes_sec=43956026,version=39387592,write_bytes_sec=165344818 1468841037000000000
-> ceph_pgmap_state,host=ceph-mon-0 active+clean=22952 1468928660000000000
+> ceph_pgmap_state,host=ceph-mon-0,state=active+clean count=22952 1468928660000000000
+> ceph_pgmap_state,host=ceph-mon-0,state=active+degraded count=16 1468928660000000000
 > ceph_usage,host=ceph-mon-0 total_avail_bytes=634895514791936,total_bytes=812117151809536,total_used_bytes=177221637017600 1468841037000000000
 > ceph_pool_usage,host=ceph-mon-0,id=150,name=cinder.volumes bytes_used=12648553794802,kb_used=12352103316,max_avail=154342562489244,objects=3026295 1468841037000000000
 > ceph_pool_usage,host=ceph-mon-0,id=182,name=cinder.volumes.flash bytes_used=8541308223964,kb_used=8341121313,max_avail=39388593563936,objects=2075066 1468841037000000000

plugins/inputs/ceph/ceph.go

@@ -4,13 +4,14 @@ import (
 	"bytes"
 	"encoding/json"
 	"fmt"
-	"github.com/influxdata/telegraf"
-	"github.com/influxdata/telegraf/plugins/inputs"
 	"io/ioutil"
 	"log"
 	"os/exec"
 	"path/filepath"
 	"strings"
+
+	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/plugins/inputs"
 )
 
 const (
@@ -100,15 +101,15 @@ func (c *Ceph) gatherAdminSocketStats(acc telegraf.Accumulator) error {
 	for _, s := range sockets {
 		dump, err := perfDump(c.CephBinary, s)
 		if err != nil {
-			log.Printf("E! error reading from socket '%s': %v", s.socket, err)
+			acc.AddError(fmt.Errorf("E! error reading from socket '%s': %v", s.socket, err))
 			continue
 		}
 		data, err := parseDump(dump)
 		if err != nil {
-			log.Printf("E! error parsing dump from socket '%s': %v", s.socket, err)
+			acc.AddError(fmt.Errorf("E! error parsing dump from socket '%s': %v", s.socket, err))
 			continue
 		}
-		for tag, metrics := range *data {
+		for tag, metrics := range data {
 			acc.AddFields(measurement,
 				map[string]interface{}(metrics),
 				map[string]string{"type": s.sockType, "id": s.sockId, "collection": tag})
@@ -244,25 +245,19 @@ type taggedMetricMap map[string]metricMap
 
 // Parses a raw JSON string into a taggedMetricMap
 // Delegates the actual parsing to newTaggedMetricMap(..)
-func parseDump(dump string) (*taggedMetricMap, error) {
+func parseDump(dump string) (taggedMetricMap, error) {
 	data := make(map[string]interface{})
 	err := json.Unmarshal([]byte(dump), &data)
 	if err != nil {
 		return nil, fmt.Errorf("failed to parse json: '%s': %v", dump, err)
 	}
 
-	tmm := newTaggedMetricMap(data)
-
-	if err != nil {
-		return nil, fmt.Errorf("failed to tag dataset: '%v': %v", tmm, err)
-	}
-
-	return tmm, nil
+	return newTaggedMetricMap(data), nil
 }
 
 // Builds a TaggedMetricMap out of a generic string map.
 // The top-level key is used as a tag and all sub-keys are flattened into metrics
-func newTaggedMetricMap(data map[string]interface{}) *taggedMetricMap {
+func newTaggedMetricMap(data map[string]interface{}) taggedMetricMap {
 	tmm := make(taggedMetricMap)
 	for tag, datapoints := range data {
 		mm := make(metricMap)
@@ -271,7 +266,7 @@ func newTaggedMetricMap(data map[string]interface{}) *taggedMetricMap {
 		}
 		tmm[tag] = mm
 	}
-	return &tmm
+	return tmm
 }
 
 // Recursively flattens any k-v hierarchy present in data.
@@ -376,36 +371,53 @@ func decodeStatusPgmap(acc telegraf.Accumulator, data map[string]interface{}) er
 	return nil
 }
 
-func decodeStatusPgmapState(acc telegraf.Accumulator, data map[string]interface{}) error {
+func extractPgmapStates(data map[string]interface{}) ([]interface{}, error) {
+	const key = "pgs_by_state"
+
 	pgmap, ok := data["pgmap"].(map[string]interface{})
 	if !ok {
-		return fmt.Errorf("WARNING %s - unable to decode pgmap", measurement)
+		return nil, fmt.Errorf("WARNING %s - unable to decode pgmap", measurement)
 	}
-	fields := make(map[string]interface{})
-	for key, value := range pgmap {
-		switch value.(type) {
-		case []interface{}:
-			if key != "pgs_by_state" {
-				continue
-			}
-			for _, state := range value.([]interface{}) {
-				state_map, ok := state.(map[string]interface{})
-				if !ok {
-					return fmt.Errorf("WARNING %s - unable to decode pg state", measurement)
-				}
-				state_name, ok := state_map["state_name"].(string)
-				if !ok {
-					return fmt.Errorf("WARNING %s - unable to decode pg state name", measurement)
-				}
-				state_count, ok := state_map["count"].(float64)
-				if !ok {
-					return fmt.Errorf("WARNING %s - unable to decode pg state count", measurement)
-				}
-				fields[state_name] = state_count
-			}
+
+	s, ok := pgmap[key]
+	if !ok {
+		return nil, fmt.Errorf("WARNING %s - pgmap is missing the %s field", measurement, key)
+	}
+
+	states, ok := s.([]interface{})
+	if !ok {
+		return nil, fmt.Errorf("WARNING %s - pgmap[%s] is not a list", measurement, key)
+	}
+	return states, nil
+}
+
+func decodeStatusPgmapState(acc telegraf.Accumulator, data map[string]interface{}) error {
+	states, err := extractPgmapStates(data)
+	if err != nil {
+		return err
+	}
+	for _, state := range states {
+		stateMap, ok := state.(map[string]interface{})
+		if !ok {
+			return fmt.Errorf("WARNING %s - unable to decode pg state", measurement)
 		}
+		stateName, ok := stateMap["state_name"].(string)
+		if !ok {
+			return fmt.Errorf("WARNING %s - unable to decode pg state name", measurement)
+		}
+		stateCount, ok := stateMap["count"].(float64)
+		if !ok {
+			return fmt.Errorf("WARNING %s - unable to decode pg state count", measurement)
+		}
+
+		tags := map[string]string{
+			"state": stateName,
+		}
+		fields := map[string]interface{}{
+			"count": stateCount,
+		}
+		acc.AddFields("ceph_pgmap_state", fields, tags)
 	}
-	acc.AddFields("ceph_pgmap_state", fields, map[string]string{})
 	return nil
 }
 

plugins/inputs/ceph/ceph_test.go

@@ -1,15 +1,17 @@
 package ceph
 
 import (
+	"encoding/json"
 	"fmt"
-	"github.com/influxdata/telegraf/testutil"
-	"github.com/stretchr/testify/assert"
 	"io/ioutil"
 	"os"
 	"path"
 	"strconv"
 	"strings"
 	"testing"
+
+	"github.com/influxdata/telegraf/testutil"
+	"github.com/stretchr/testify/assert"
 )
 
 const (
@@ -24,15 +26,38 @@ func TestParseSockId(t *testing.T) {
 func TestParseMonDump(t *testing.T) {
 	dump, err := parseDump(monPerfDump)
 	assert.NoError(t, err)
-	assert.InEpsilon(t, 5678670180, (*dump)["cluster"]["osd_kb_used"], epsilon)
-	assert.InEpsilon(t, 6866.540527000, (*dump)["paxos"]["store_state_latency.sum"], epsilon)
+	assert.InEpsilon(t, int64(5678670180), dump["cluster"]["osd_kb_used"], epsilon)
+	assert.InEpsilon(t, 6866.540527000, dump["paxos"]["store_state_latency.sum"], epsilon)
 }
 
 func TestParseOsdDump(t *testing.T) {
 	dump, err := parseDump(osdPerfDump)
 	assert.NoError(t, err)
-	assert.InEpsilon(t, 552132.109360000, (*dump)["filestore"]["commitcycle_interval.sum"], epsilon)
-	assert.Equal(t, float64(0), (*dump)["mutex-FileJournal::finisher_lock"]["wait.avgcount"])
+	assert.InEpsilon(t, 552132.109360000, dump["filestore"]["commitcycle_interval.sum"], epsilon)
+	assert.Equal(t, float64(0), dump["mutex-FileJournal::finisher_lock"]["wait.avgcount"])
+}
+
+func TestDecodeStatusPgmapState(t *testing.T) {
+	data := make(map[string]interface{})
+	err := json.Unmarshal([]byte(clusterStatusDump), &data)
+	assert.NoError(t, err)
+
+	acc := &testutil.Accumulator{}
+	err = decodeStatusPgmapState(acc, data)
+	assert.NoError(t, err)
+
+	var results = []struct {
+		fields map[string]interface{}
+		tags   map[string]string
+	}{
+		{map[string]interface{}{"count": float64(2560)}, map[string]string{"state": "active+clean"}},
+		{map[string]interface{}{"count": float64(10)}, map[string]string{"state": "active+scrubbing"}},
+		{map[string]interface{}{"count": float64(5)}, map[string]string{"state": "active+backfilling"}},
+	}
+
+	for _, r := range results {
+		acc.AssertContainsTaggedFields(t, "ceph_pgmap_state", r.fields, r.tags)
+	}
 }
 
 func TestGather(t *testing.T) {
@@ -685,3 +710,127 @@ var osdPerfDump = `
       "wait": { "avgcount": 0,
           "sum": 0.000000000}}}
 `
+var clusterStatusDump = `
+{
+  "health": {
+    "health": {
+      "health_services": [
+        {
+          "mons": [
+            {
+              "name": "a",
+              "kb_total": 114289256,
+              "kb_used": 26995516,
+              "kb_avail": 81465132,
+              "avail_percent": 71,
+              "last_updated": "2017-01-03 17:20:57.595004",
+              "store_stats": {
+                "bytes_total": 942117141,
+                "bytes_sst": 0,
+                "bytes_log": 4345406,
+                "bytes_misc": 937771735,
+                "last_updated": "0.000000"
+              },
+              "health": "HEALTH_OK"
+            },
+            {
+              "name": "b",
+              "kb_total": 114289256,
+              "kb_used": 27871624,
+              "kb_avail": 80589024,
+              "avail_percent": 70,
+              "last_updated": "2017-01-03 17:20:47.784331",
+              "store_stats": {
+                "bytes_total": 454853104,
+                "bytes_sst": 0,
+                "bytes_log": 5788320,
+                "bytes_misc": 449064784,
+                "last_updated": "0.000000"
+              },
+              "health": "HEALTH_OK"
+            },
+            {
+              "name": "c",
+              "kb_total": 130258508,
+              "kb_used": 38076996,
+              "kb_avail": 85541692,
+              "avail_percent": 65,
+              "last_updated": "2017-01-03 17:21:03.311123",
+              "store_stats": {
+                "bytes_total": 455555199,
+                "bytes_sst": 0,
+                "bytes_log": 6950876,
+                "bytes_misc": 448604323,
+                "last_updated": "0.000000"
+              },
+              "health": "HEALTH_OK"
+            }
+          ]
+        }
+      ]
+    },
+    "timechecks": {
+      "epoch": 504,
+      "round": 34642,
+      "round_status": "finished",
+      "mons": [
+        { "name": "a", "skew": 0, "latency": 0, "health": "HEALTH_OK" },
+        { "name": "b", "skew": -0, "latency": 0.000951, "health": "HEALTH_OK" },
+        { "name": "c", "skew": -0, "latency": 0.000946, "health": "HEALTH_OK" }
+      ]
+    },
+    "summary": [],
+    "overall_status": "HEALTH_OK",
+    "detail": []
+  },
+  "fsid": "01234567-abcd-9876-0123-ffeeddccbbaa",
+  "election_epoch": 504,
+  "quorum": [ 0, 1, 2 ],
+  "quorum_names": [ "a", "b", "c" ],
+  "monmap": {
+    "epoch": 17,
+    "fsid": "01234567-abcd-9876-0123-ffeeddccbbaa",
+    "modified": "2016-04-11 14:01:52.600198",
+    "created": "0.000000",
+    "mons": [
+      { "rank": 0, "name": "a", "addr": "192.168.0.1:6789/0" },
+      { "rank": 1, "name": "b", "addr": "192.168.0.2:6789/0" },
+      { "rank": 2, "name": "c", "addr": "192.168.0.3:6789/0" }
+    ]
+  },
+  "osdmap": {
+    "osdmap": {
+      "epoch": 21734,
+      "num_osds": 24,
+      "num_up_osds": 24,
+      "num_in_osds": 24,
+      "full": false,
+      "nearfull": false,
+      "num_remapped_pgs": 0
+    }
+  },
+  "pgmap": {
+    "pgs_by_state": [
+      { "state_name": "active+clean", "count": 2560 },
+      { "state_name": "active+scrubbing", "count": 10 },
+      { "state_name": "active+backfilling", "count": 5 }
+    ],
+    "version": 52314277,
+    "num_pgs": 2560,
+    "data_bytes": 2700031960713,
+    "bytes_used": 7478347665408,
+    "bytes_avail": 9857462382592,
+    "bytes_total": 17335810048000,
+    "read_bytes_sec": 0,
+    "write_bytes_sec": 367217,
+    "op_per_sec": 98
+  },
+  "mdsmap": {
+    "epoch": 1,
+    "up": 0,
+    "in": 0,
+    "max": 0,
+    "by_rank": []
+  }
+}
+`

plugins/inputs/cgroup/cgroup_linux.go

@@ -22,10 +22,11 @@ func (g *CGroup) Gather(acc telegraf.Accumulator) error {
 
 	for dir := range list {
 		if dir.err != nil {
-			return dir.err
+			acc.AddError(dir.err)
+			continue
 		}
 		if err := g.gatherDir(dir.path, acc); err != nil {
-			return err
+			acc.AddError(err)
 		}
 	}
 
@@ -224,7 +225,7 @@ var fileFormats = [...]fileFormat{
 }
 
 func numberOrString(s string) interface{} {
-	i, err := strconv.Atoi(s)
+	i, err := strconv.ParseInt(s, 10, 64)
 	if err == nil {
 		return i
 	}

plugins/inputs/cgroup/cgroup_test.go

@@ -24,24 +24,24 @@ var cg1 = &CGroup{
 func TestCgroupStatistics_1(t *testing.T) {
 	var acc testutil.Accumulator
 
-	err := cg1.Gather(&acc)
+	err := acc.GatherError(cg1.Gather)
 	require.NoError(t, err)
 
 	tags := map[string]string{
 		"path": "testdata/memory",
 	}
 	fields := map[string]interface{}{
-		"memory.stat.cache":           1739362304123123123,
-		"memory.stat.rss":             1775325184,
-		"memory.stat.rss_huge":        778043392,
-		"memory.stat.mapped_file":     421036032,
-		"memory.stat.dirty":           -307200,
-		"memory.max_usage_in_bytes.0": 0,
-		"memory.max_usage_in_bytes.1": -1,
-		"memory.max_usage_in_bytes.2": 2,
-		"memory.limit_in_bytes":       223372036854771712,
+		"memory.stat.cache":           int64(1739362304123123123),
+		"memory.stat.rss":             int64(1775325184),
+		"memory.stat.rss_huge":        int64(778043392),
+		"memory.stat.mapped_file":     int64(421036032),
+		"memory.stat.dirty":           int64(-307200),
+		"memory.max_usage_in_bytes.0": int64(0),
+		"memory.max_usage_in_bytes.1": int64(-1),
+		"memory.max_usage_in_bytes.2": int64(2),
+		"memory.limit_in_bytes":       int64(223372036854771712),
 		"memory.use_hierarchy":        "12-781",
-		"notify_on_release":           0,
+		"notify_on_release":           int64(0),
 	}
 	acc.AssertContainsTaggedFields(t, "cgroup", fields, tags)
 }
@@ -56,17 +56,17 @@ var cg2 = &CGroup{
 func TestCgroupStatistics_2(t *testing.T) {
 	var acc testutil.Accumulator
 
-	err := cg2.Gather(&acc)
+	err := acc.GatherError(cg2.Gather)
 	require.NoError(t, err)
 
 	tags := map[string]string{
 		"path": "testdata/cpu",
 	}
 	fields := map[string]interface{}{
-		"cpuacct.usage_percpu.0": -1452543795404,
-		"cpuacct.usage_percpu.1": 1376681271659,
-		"cpuacct.usage_percpu.2": 1450950799997,
-		"cpuacct.usage_percpu.3": -1473113374257,
+		"cpuacct.usage_percpu.0": int64(-1452543795404),
+		"cpuacct.usage_percpu.1": int64(1376681271659),
+		"cpuacct.usage_percpu.2": int64(1450950799997),
+		"cpuacct.usage_percpu.3": int64(-1473113374257),
 	}
 	acc.AssertContainsTaggedFields(t, "cgroup", fields, tags)
 }
@@ -81,14 +81,14 @@ var cg3 = &CGroup{
 func TestCgroupStatistics_3(t *testing.T) {
 	var acc testutil.Accumulator
 
-	err := cg3.Gather(&acc)
+	err := acc.GatherError(cg3.Gather)
 	require.NoError(t, err)
 
 	tags := map[string]string{
 		"path": "testdata/memory/group_1",
 	}
 	fields := map[string]interface{}{
-		"memory.limit_in_bytes": 223372036854771712,
+		"memory.limit_in_bytes": int64(223372036854771712),
 	}
 	acc.AssertContainsTaggedFields(t, "cgroup", fields, tags)
 
@@ -108,14 +108,14 @@ var cg4 = &CGroup{
 func TestCgroupStatistics_4(t *testing.T) {
 	var acc testutil.Accumulator
 
-	err := cg4.Gather(&acc)
+	err := acc.GatherError(cg4.Gather)
 	require.NoError(t, err)
 
 	tags := map[string]string{
 		"path": "testdata/memory/group_1/group_1_1",
 	}
 	fields := map[string]interface{}{
-		"memory.limit_in_bytes": 223372036854771712,
+		"memory.limit_in_bytes": int64(223372036854771712),
 	}
 	acc.AssertContainsTaggedFields(t, "cgroup", fields, tags)
 
@@ -140,14 +140,14 @@ var cg5 = &CGroup{
 func TestCgroupStatistics_5(t *testing.T) {
 	var acc testutil.Accumulator
 
-	err := cg5.Gather(&acc)
+	err := acc.GatherError(cg5.Gather)
 	require.NoError(t, err)
 
 	tags := map[string]string{
 		"path": "testdata/memory/group_1/group_1_1",
 	}
 	fields := map[string]interface{}{
-		"memory.limit_in_bytes": 223372036854771712,
+		"memory.limit_in_bytes": int64(223372036854771712),
 	}
 	acc.AssertContainsTaggedFields(t, "cgroup", fields, tags)
 
@@ -167,16 +167,16 @@ var cg6 = &CGroup{
 func TestCgroupStatistics_6(t *testing.T) {
 	var acc testutil.Accumulator
 
-	err := cg6.Gather(&acc)
+	err := acc.GatherError(cg6.Gather)
 	require.NoError(t, err)
 
 	tags := map[string]string{
 		"path": "testdata/memory",
 	}
 	fields := map[string]interface{}{
-		"memory.usage_in_bytes":      3513667584,
+		"memory.usage_in_bytes":      int64(3513667584),
 		"memory.use_hierarchy":       "12-781",
-		"memory.kmem.limit_in_bytes": 9223372036854771712,
+		"memory.kmem.limit_in_bytes": int64(9223372036854771712),
 	}
 	acc.AssertContainsTaggedFields(t, "cgroup", fields, tags)
 }

plugins/inputs/chrony/README.md

@@ -63,6 +63,7 @@ Delete second or Not synchronised.
 ### Measurements & Fields:
 
 - chrony
+    - system_time (float, seconds)
     - last_offset (float, seconds)
     - rms_offset (float, seconds)
     - frequency (float, ppm)
@@ -82,9 +83,9 @@ Delete second or Not synchronised.
 ### Example Output:
 
 ```
-$ telegraf -config telegraf.conf -input-filter chrony -test
+$ telegraf --config telegraf.conf --input-filter chrony --test
 * Plugin: chrony, Collection 1
-> chrony,leap_status=normal,reference_id=192.168.1.1,stratum=3 frequency=-35.657,last_offset=-0.000013616,residual_freq=-0,rms_offset=0.000027073,root_delay=0.000644,root_dispersion=0.003444,skew=0.001,update_interval=1031.2 1463750789687639161
+> chrony,leap_status=normal,reference_id=192.168.1.1,stratum=3 frequency=-35.657,system_time=0.000027073,last_offset=-0.000013616,residual_freq=-0,rms_offset=0.000027073,root_delay=0.000644,root_dispersion=0.003444,skew=0.001,update_interval=1031.2 1463750789687639161
 ```
 
 

plugins/inputs/chrony/chrony.go

@@ -1,5 +1,3 @@
-// +build linux
-
 package chrony
 
 import (
@@ -92,7 +90,7 @@ func processChronycOutput(out string) (map[string]interface{}, map[string]string
 		}
 		name := strings.ToLower(strings.Replace(strings.TrimSpace(stats[0]), " ", "_", -1))
 		// ignore reference time
-		if strings.Contains(name, "time") {
+		if strings.Contains(name, "ref_time") {
 			continue
 		}
 		valueFields := strings.Fields(stats[1])