Hi folks - I’ve recently started using the most recent Nagios 3.0 stable release. What I’m finding is that if a parent host is unreachable, I correctly don’t get alerts for services ‘behind’ this down host - but the service is being marked as in a critical state - not an unknown state. This makes availability figures suck - I’d rather the service was assumed to be unknown than down.
I was fairly familiar with 2.x which I used until around a year ago - and I can almost swear that this worked as I wanted it to, unless I’m missing a configuration parameter somewhere? (this is not unlikely!).
I ought to clarify a little. In this particular instance, a parent host was marked down for a period of 1/2 an hour. The parent host was a router in which intrusion detection kicked in and disabled ping (thus host was marked down).
The services behind this router were still responding, albeit sporadically.
Does host checking of child hosts suspend if the parent host is down? Do I need to add host dependencies to achieve this? I thought that in the case of hosts, if a parent host is down, the child hosts are assumed to be too (without the need for host dependencies).
Would service checking on those child hosts be suspended?
I think I may be making a few too many assumptions.