Monitoring "chkconfig" settings

[forrie]

We've encountered an issue with RHEL where the chkconfig settings of some daemons (that were previously configured via --add) seem to be removed from startup.

I'm looking for a way to monitor these -- perhaps with a Nagios script (we use Nagios internally, so that makes sense).

This may be more of a configuration management issue (aka Puppet), but... the daemons I'm monitoring are critical to operations here, so it qualifies for alerts.

Suggestions?

 

[lgarner]

You could store a "good" chkconfig --list somewhere and have Nagios run that periodically and report differences.

I'd probably monitor the daemons themselves rather than the startup status. An alert action could even be to start the non-running service.