Nagios Support Forum

We have a rather large nagios installation (Hosts: 128 / Services: 1557) and we often get notifications which we think could be avoided.

This happens when one of the routers go down with many hosts and services behind it.
but also when a host reboots and comes back up again with some of the services taking a little longer to get ready.

I have searched for something like a grace period. Mean:
a host which is depended upon (parent) is considered up only if its up state is at least e.g. 5 minutes old.

this should apply to all dependency relations in nagios.

we are grateful for any hints or suggestions.

(Nagios Core 3.2.1, from current debian stable)

You could define a service escalation with a first_notification delay of 5 min or so. This would only send a notification is the state is HARD after 5 minutes. Tune to your liking. Just a thought.