Nagios Support Forum

Hi All,

I'll not bore you with my excuses but let it be said that I'm definitely a nagios novice.

We've have a recently installed nagios Core instance (3.5.0) on Freebsd 9.2 which was built to replace a much older version, a number of config files were migrated across to make the migration easier...

Nagios appears to be monitoring the correct hosts/services and notifications all work correctly (email and SMS). We've had no problems until it was recently noticed that we weren't always receiving service notifications when a host failed. Service checks/notifications work fine when the host check is still ok. It appears that when the Host check goes critical (HARD or SOFT) Service checks for that host stop processing. This actually seems quite sensible but as it wasn't the the case in the original nagios instance the actions based on failed service checks are no longer firing off. I've also failed miserably to find documentation that refers to this function.

So I suppose the question is:
Is there a global option that would allow service checks to continue regardless of the host check status?

Thanks

Quin

This has been the way Nagios works since I can remember (going back to Nagios 1.3). I know of no way to notify that services on a downed host are not working, but then I've never looked, either.

Honestly I would just set the host check to check_dummy and make it always OK, then make a service for the real state of the host. That way if you later want the services to stop checking based on the host's state, you can use dependencies to do so.

Thanks for the replies, I suspected that it was operating correctly and that the previous nagios admin had managed to 'bodge' it to do what he wanted.

Good suggestion on the use of a dummy check, that will server as a good temporary measure at least until I get round to rearranging the way the service checks work.

Many thanks for you input.

Quin