37 lines
1.3 KiB
Markdown
37 lines
1.3 KiB
Markdown
# Telegraf Input Plugin: Fleet
|
|
|
|
The plugin will gather names of running units from [fleet](https://github.com/coreos/fleet) and the sum total of each running unit. It uses the fleet v1 API to gather data.
|
|
|
|
### Configuration:
|
|
|
|
```toml
|
|
# Description
|
|
[[inputs.fleet]]
|
|
## Works with Fleet HTTP API
|
|
## Multiple Hosts from which to read Fleet stats:
|
|
hosts = ["http://localhost:49153/fleet/v1/state"]
|
|
```
|
|
|
|
### Measurements & Fields:
|
|
|
|
The fields are dynamically generated from the output of the fleet API. Using the ```name``` value.. The values of those fields are the number of containers with the ```systemdSubState``` value of "running".
|
|
<insert example json output here of both running and not to show what is included and not included>
|
|
|
|
The unit names will have their instanced id and the @ symbol stripped off.
|
|
For example if you had a unit named ```nginx-1.10.1@35``` the field name would be ```nginx-1.10.1```.
|
|
|
|
- fleet
|
|
- ```<dynamic unit name>``` (int)
|
|
|
|
### Tags:
|
|
|
|
- All measurements have the following tags:
|
|
- server (name of the host/container telegraf is running on)
|
|
|
|
### Example Output:
|
|
|
|
```
|
|
$ ./telegraf -config telegraf.conf -input-filter example -test
|
|
* Plugin: fleet, Collection 1
|
|
> fleet,host=localhost.local,server=http://fleet.testserver.com:49153/fleet/v1/state some-api=2i,test-application=1i,webapp=1i,nginx=2i,redis=1i 1470615664000000000
|
|
``` |