Linux HA Heartbeat/Monitoring with Nagios

Requirements

You need to have a Nagios NRPE server running on the host.

The host need to have the command /usr/bin/cl_status accessible to user nrpe.

Linux HA Server

Script

The script will return a Nagios code of:

• OK: when the number of registered nodes and active nodes is equal

• WARNING: If at least one node is active

• CRITICAL: If the cluster status could not be accessed or if no nodes are active

You can get it here: media:check_heartbeat.sh

Copy it to your nagios plugins directory, in my case: /usr/local/nagios/libexec/

# cp check_hearbeat.sh /usr/local/nagios/libexec/check_heartbeat.sh
# chmod 755 /usr/local/nagios/libexec/check_heartbeat.sh

Configuration

Add the following to /etc/nagios/nrpe_local.cfg

command[check_heartbeat]=/usr/local/nagios/libexec/check_heartbeat.sh

and reload nrpe server:

# /etc/init.d/nagios-nrpe-server reload

Nagios Server

template.cfg

Add the following template:

# check LVS Heartbeat
define service{
    use generic-service
    name check-heartbeat
    service_description Linux HA LVS
    check_command   check_nrpe!check_heartbeat
}

host file

Add the following service definition:

#check heartbeat
define service{
    use check-heartbeat
    host_name   lvs1, lvs2
}