Public | Automated Build

Last pushed: 2 years ago
Short Description
Short description is empty for this repo.
Full Description

CollectD -> InfluxDB Docker Container

tl;dr

  • sends data to a InfluxDB that has the collectD plugin enabled
  • uses Confd for hooking up InfluxDB location with CollectD

Summary

CAdvisor and Heapster work well for gathering metrics on the containers, but we still need to monitor the docker host, right? Right?? :)

This requires 2 things:

  • InfluxDB with the collectD plugin enabled. See skippy/influxdb config file
  • the InfluxDB hosts are expected to have the following key in etcd /services/influxdb/hosts/<IP-ADDR> If you are curious how to self-register a InfluxDB cluster (using Fleet), see the skippy/influxdb fleet files

Usage

There are two ways to start this container:

docker run --name collectd --net=host skippy/collectd-influxdb

or without --net=host

docker run --name collectd -e ETCD_IP=<ETCD_IP> skippy/collectd-influxdb

Dependencies

For anything dealing with clustering, they currently are:

<a name="todos"></a>

TODOs:

  • use a standard build of ConfD instead of the custom build from the master branch of kelseyhightower/confd
  • right now we direct all collectD traffic to the first influxDB host registered in etcd. This should be randomized
  • confd does not pass along output from STDOUT. So I have to send collectD to log and then tail the log. This is a problem in case the log file fills up (highly unlikely I know, but still). Confd seemed to used to be able to pickup and pass back STDOUT, but no more.
Docker Pull Command
Owner
skippy
Source Repository

Comments (0)