Check_Jenkins_*

Origin

I've used Jenkins as a Continuous Integration for a while.

Surprisingly the company I joined recently used it also as an orchestration tool :

We don't only manage our code with jenkins but also let it deal with jobs that process data on other jenkins instances. In that context it was important to monitor the success or failure of a job, but most importantly make sure they don't take too long (timeout-plugin was not an option)

Nagios also supports escalation which is a nice feature in term of process and SLA in our environment.

Check Jenkins

Usage

Command line

You can test it quickly with :

./check_jenkins.py  -w 200 -c 300  -H builds.apache.org -S -j Hadoop-Common-trunk 

That check will verify that Hadoop-Common-trunk on https://builds.apache.org is not taking more than 200 minutes (crit at 300min)and th elast build was successful

Please don't hammer builds.apache.org

Nagios

Define a command

# check Jenkins job 
# Note : we have ssl ON by default
define command{
    command_name    check_jenkins_job
    command_line    $USER2$/check_jenkins.py -S -H $HOSTNAME$ -j $ARG1$ -w $ARG2$ -c $ARG3$ -u $ARG4$ -p $ARG5$ 
}

Then a service

define service{
     use                     generic-service
     service_description     check_jenkins Process data
     check_command           check_jenkins!Large_data_process!360!540!nagios!readonly!
     host_name               data-process-01.acme.tld
}

I'd recommend to put the various scripts in a folder defined with $USER2$ in resource.cfg, to avoid having it with system package based checks in $USER1$

Check Jenkins Last Successful Build

Usage

Command line

You can test it quickly with :

./check_jenkins_lsb.py  -w 1d -c 2d  -H builds.apache.org -S -j Hadoop-Common-trunk 

That check will verify that Hadoop-Common-trunk on https://builds.apache.org successfully run since a day or at worst two days

Please don't hammer builds.apache.org

Nagios

Define a command

# check Jenkins job 
# Note : we have ssl ON by default
define command{
    command_name    check_jenkins_job_lsb
    command_line    $USER2$/check_jenkins_lsb.py -S -H $HOSTNAME$ -j $ARG1$ -w $ARG2$ -c $ARG3$ -u $ARG4$ -p $ARG5$ 
}

Then a service

define service{
     use                     generic-service
     service_description     check_jenkins Process data
     check_command           check_jenkins_lsb!Large_data_process!1d!2d!nagios!readonly!
     host_name               data-process-01.acme.tld
}

I'd recommend to put the various scripts in a folder defined with $USER2$ in resource.cfg, to avoid having it with system package based checks in $USER1$