elk-collectd

Collectd plugins to get some stats and metrics from Elasticsearch and Logstash. Using collectd-python.

Requirements

• collectd 4.9+
• python 2.6+
• Elasticsearch 6
• Logstash 6

Installation

1. Place *.py files into a directory on the host. For example into /usr/share/collectd/plugins.
2. Configure collectd (see below).
3. Restart collectd.

Configuration

1. collectd.conf:

LoadPlugin python

<Plugin python>
  ModulePath "/usr/share/collectd/plugins"
  Import "elasticsearch-collectd"
  <Module "elasticsearch-collectd">
    Host "localhost" # elasticsearch host, default is localhost
    Port 9200 # elasticsearch API port, default is 9200
    Name "elasticsearch" # instance name, default is elasticsearch
    Stats_enabled "nodes cluster_health" # Which stats you should check, default is to check both nodes and cluster_health
    Timeout 20 # stat fetching timeout in secs, default is 20
  </Module>
<Plugin>

<Plugin python>
  ModulePath "/usr/share/collectd/plugins"
  Import "logstash-collectd"
  <Module "logstash-collectd">
    Host "localhost" # logstash host, default is localhost
    Port 9600 # logstash API port, default is 9600
    Name "logstash" # instance name, default is logstash
  </Module>
<Plugin>

2. You can use multiple Module blocks with different options if you need to collect data from multiple instances.
3. If you skip one or more option or whole Module block, default values will be used.