telegraf/plugins/inputs/prometheus/README.md

106 lines
4.0 KiB
Markdown
Raw Normal View History

2016-03-01 16:12:23 +00:00
# Prometheus Input Plugin
2017-09-18 23:21:45 +00:00
The prometheus input plugin gathers metrics from HTTP servers exposing metrics
in Prometheus format.
2016-03-01 16:12:23 +00:00
### Configuration:
```toml
2017-09-18 23:21:45 +00:00
# Read metrics from one or many prometheus clients
2016-03-01 16:12:23 +00:00
[[inputs.prometheus]]
2017-09-18 23:21:45 +00:00
## An array of urls to scrape metrics from.
urls = ["http://localhost:9100/metrics"]
2016-03-01 16:12:23 +00:00
2017-09-18 23:21:45 +00:00
## An array of Kubernetes services to scrape metrics from.
# kubernetes_services = ["http://my-service-dns.my-namespace:9100/metrics"]
2017-09-18 23:21:45 +00:00
## Use bearer token for authorization
# bearer_token = /path/to/bearer/token
2017-09-18 23:21:45 +00:00
## Specify timeout duration for slower prometheus clients (default is 3s)
# response_timeout = "3s"
2016-03-01 16:12:23 +00:00
2018-05-04 23:33:23 +00:00
## Optional TLS Config
# tls_ca = /path/to/cafile
# tls_cert = /path/to/certfile
# tls_key = /path/to/keyfile
## Use TLS but skip chain & host verification
2017-09-18 23:21:45 +00:00
# insecure_skip_verify = false
2016-03-01 16:12:23 +00:00
```
2017-09-18 23:21:45 +00:00
#### Kubernetes Service Discovery
2017-09-18 23:21:45 +00:00
URLs listed in the `kubernetes_services` parameter will be expanded
by looking up all A records assigned to the hostname as described in
[Kubernetes DNS service discovery](https://kubernetes.io/docs/concepts/services-networking/service/#dns).
2017-09-18 23:21:45 +00:00
This method can be used to locate all
[Kubernetes headless services](https://kubernetes.io/docs/concepts/services-networking/service/#headless-services).
2017-09-18 23:21:45 +00:00
#### Bearer Token
2017-09-18 23:21:45 +00:00
If set, the file specified by the `bearer_token` parameter will be read on
each interval and its contents will be appended to the Bearer string in the
Authorization header.
### Usage for Caddy HTTP server
If you want to monitor Caddy, you need to use Caddy with its Prometheus plugin:
* Download Caddy+Prometheus plugin [here](https://caddyserver.com/download/linux/amd64?plugins=http.prometheus)
* Add the `prometheus` directive in your `CaddyFile`
* Restart Caddy
* Configure Telegraf to fetch metrics on it:
```
[[inputs.prometheus]]
# ## An array of urls to scrape metrics from.
urls = ["http://localhost:9180/metrics"]
```
> This is the default URL where Caddy Prometheus plugin will send data.
> For more details, please read the [Caddy Prometheus documentation](https://github.com/miekg/caddy-prometheus/blob/master/README.md).
2017-09-18 23:21:45 +00:00
### Metrics:
Measurement names are based on the Metric Family and tags are created for each
label. The value is added to a field named based on the metric type.
2016-03-01 16:12:23 +00:00
2017-09-18 23:21:45 +00:00
All metrics receive the `url` tag indicating the related URL specified in the
Telegraf configuration. If using Kubernetes service discovery the `address`
2017-09-18 23:21:45 +00:00
tag is also added indicating the discovered ip address.
2016-03-01 16:12:23 +00:00
2017-09-18 23:21:45 +00:00
### Example Output:
2016-03-01 16:12:23 +00:00
2017-09-18 23:21:45 +00:00
**Source**
2016-03-01 16:12:23 +00:00
```
# HELP go_gc_duration_seconds A summary of the GC invocation durations.
# TYPE go_gc_duration_seconds summary
2017-09-18 23:21:45 +00:00
go_gc_duration_seconds{quantile="0"} 7.4545e-05
go_gc_duration_seconds{quantile="0.25"} 7.6999e-05
go_gc_duration_seconds{quantile="0.5"} 0.000277935
go_gc_duration_seconds{quantile="0.75"} 0.000706591
go_gc_duration_seconds{quantile="1"} 0.000706591
go_gc_duration_seconds_sum 0.00113607
go_gc_duration_seconds_count 4
2016-03-01 16:12:23 +00:00
# HELP go_goroutines Number of goroutines that currently exist.
# TYPE go_goroutines gauge
go_goroutines 15
2017-09-18 23:21:45 +00:00
# HELP cpu_usage_user Telegraf collected metric
# TYPE cpu_usage_user gauge
cpu_usage_user{cpu="cpu0"} 1.4112903225816156
cpu_usage_user{cpu="cpu1"} 0.702106318955865
cpu_usage_user{cpu="cpu2"} 2.0161290322588776
cpu_usage_user{cpu="cpu3"} 1.5045135406226022
2016-03-01 16:12:23 +00:00
```
2017-09-18 23:21:45 +00:00
**Output**
2016-03-01 16:12:23 +00:00
```
2017-09-18 23:21:45 +00:00
go_gc_duration_seconds,url=http://example.org:9273/metrics 1=0.001336611,count=14,sum=0.004527551,0=0.000057965,0.25=0.000083812,0.5=0.000286537,0.75=0.000365303 1505776733000000000
go_goroutines,url=http://example.org:9273/metrics gauge=21 1505776695000000000
cpu_usage_user,cpu=cpu0,url=http://example.org:9273/metrics gauge=1.513622603430151 1505776751000000000
cpu_usage_user,cpu=cpu1,url=http://example.org:9273/metrics gauge=5.829145728641773 1505776751000000000
cpu_usage_user,cpu=cpu2,url=http://example.org:9273/metrics gauge=2.119071644805144 1505776751000000000
cpu_usage_user,cpu=cpu3,url=http://example.org:9273/metrics gauge=1.5228426395944945 1505776751000000000
2016-03-01 16:12:23 +00:00
```