Recently I had to configure nagios to monitor a non-standard port (9999/TCP).  This is a service of ours that apparently decides to die at random times.  In order to find out more about the issue (this service is not live yet), I configured an instance of nagios to monitor this port.

In order to do this, I used the pre-existing check_tcp with a specific run-time parameter (namely, -p 9999).

First, define a command for this service check:

define command{
     command_name check_tcp_9999
     command_line $USER1$/check_tcp -h $HOSTADDRESS$ -p 9999 -4
}

Note that we manually defined the -p (port) and -h (host) parameters to use the host using this command, as well as port 9999, respectively.

Once the command has been defined, create a new service for the host in question.  If you haven’t set one up yet, below is an example to follow:

define host{
     use generic-host   ; template name, available by default
     host_name svr-001  ; unique name of the host being defined
     alias server 001   ; description of the host
     address 10.0.0.254 ; IP of the host
}

To get the service check we defined above running against this host, reference the following stanza.

define service{
     use generic-service                          ; template name, available by default
     host_name svr-001                            ; the host against which to run this check
     service_description 'tcp check of 9999/tcp'  ; self-explanatory
     check_command check_tcp_9999                 ; name of the command we defined earlier
     check_interval 5                             ; check this service every 5 minutes
     check_period 24x7                            ; time period in which to monitor
     retry_interval 3                             ; if a check fails, re-try in 3 minutes
     max_check_attempts 3                         ; the max number of times a failed service will be checked
     notification_interval 60                     ; reminder of alerts sent every 60 minutes
     notification_period 24x7                     ; time period in which alerts are sent
     notification_options w,c,u                   ; the conditions upon which to send alerts
     contact_groups contacts                      ; contact group to alert in the event of failure
}

Verify that the additions you made to the config file(s) were valid by executing:

# nagios3 -v /etc/nagios3/nagios.cfg

If the pre-flight check goes well, update your nagios installation by executing:

# /etc/init.d/nagios3 restart

That’s it!