Simple plugin for nagios that wrap the docker info command.
It supports a fews critical and warning alerts like service status and ram issues.
It follows the guidelines from this site https://www.monitoring-plugins.org/doc/guidelines.html
Clone the repository
git clone https://github.com/sirbotta/docker-nagios-plugin.git
Install the dependecies
cd docker-nagios-plugin
bundle install
Link or copy the script to your nagios plugins folder
ln -s docker_info.rb /usr/lib/nagios/plugins/check_docker.rb
Wire up your script as usual in the nrpe server and nagios server. The plugin must be used with the check_nrpe command.
NRPE server
in /etc/nagios/nrpe.cfg add this line
command[check_docker]=/usr/lib/nagios/plugins/check_docker.rb
restart the nrpe server
sudo service nagios-nrpe-server restart
NAGIOS server
In your /etc/nagios3/command.cfg add something like
define command {
command_name check_docker
command_line $user1$/check_nrpe -H $HOSTADDRESS$ -c check_docker
}
In /etc/nagios3/conf.d/your-host.cfg add
define service{
use generic-service
host_name your-host
check_command check_docker
}
restart nagios
sudo service nagios3 restart
./docker_info.rb
or
ruby docker_info.rb
if everything is correct the result should be like this
DOCKERINFO OK: StorageDriver is aufs ,RootDir is /var/lib/docker/aufs ,ExecutionDriver is native-0.1 ,KernelVersion is 3.11.0-15-generic|Containers=1 ,Images=4 ,Dirs=6
also you can set up custom thresholds for containers with -w (warning) and -c (critical)
rise warning if more of 10 containers are running in the docker server
./docker_info.rb -w10
rise warning if more of 10 containers and rise a critical if 20 or more are running in the docker server
./docker_info.rb -w10 -c20
OK if everything is under the thresholds (default unlimited)
WARNING if [No swap limit] warning from docker or surpassed a warning threshold
CRITICAL if permission denied on the nagios user, the docker daemon is not running or surpassed a critical threshold