I understand how to monitor long running processes with monit or god,
but I don’t understand how to monitor Cron jobs (for memory bloating
notably) since such processes
- always have different PIDs when they start
- are precisely supposed to stop when they are done, then restart when
asked (vs running all the time).
I have been looking on the web for a while, but didn’t find any simple
way. Is there a recommended “basic strategy” to monitor Cron jobs?
When the process starts, or when it is started by a wrapper, the
process writes it’s pid to a file such as /var/run/my_app.pid
As the pid is always written to the same file monit can read the file
and find the pid of the process and check that there is a process with
that pid running.
One thing though: my script/runner doesn’t seem to write a PID file
(or at least I can’t locate it in the “typical” places - it might be a
right issue). Is there a possibility to manually specify a PID file
(either a runner option, or a shell command)? I can’t find such