Configure and install the collectd[http://collectd.org/] monitoring daemon and plugins.
This cookbook has only been tested on Ubuntu.
Note: apache plugin has a dependency on libcurl. (see https://collectd.org/wiki/index.php/Plugin:Apache)
= ATTRIBUTES:
node['collectd']['name']- The hostname of the node as reported in statistics. Defaults tonode['fqdn'].node['collectd']['interval']- Time period in seconds to wait between data reads. Defaults to 10.node['collectd']['read_threads']- The number of threads performing data reads. Defaults to 5.
There is one main recipes provided:
collectd::default- Install a standalone daemon.collectd::attribute_driven- Install a standalone daemon.
There is one resource provided to manage collectd configuration.
The +collectd_plugin+ resource configures and enables standard collect plugins. Example:
collectd_plugin "interface" do
config :interface => "lo", :ignore_selected => true
end
The options hash is converted to collectd-style settings automatically. Any symbol key will be converted to camel-case. In the above example :ignore_selected will be output as the key "IgnoreSelected". If the key is already a string, this conversion is skipped. If the value is an array, it will be output as a separate line for each element.
The collectd_plugin resource can also use templates for more complex configuration. Example:
collectd_plugin "interface" do
template "generic_jmx.conf.erb"
cookbook "java-monitoring"
config :host => "127.0.0.1", :port => 2020
end
The collectd_plugin resource can also use content directly in recipe. Example:
collectd_plugin "interface" do
content 'LoadPlugin "foo"'
end
Note: if you get "Could not find plugin foobar" or "foobar wasn't loaded using LoadPlugin" then you might need to reinstall collectd.
Some plugins in collectd (e.g. mysql) require all the dependencies to be installed before you run configure && make and make install
on collectd.
Author: Noah Kantrowitz (nkantrowitz@crypticstudios.com) Copyright: 2010, Atari, Inc