Nagios Support Forum

Yes. thank you.

Once we hear back from 'amprantino' we'll close the thread out, just let us know!

Unfortunately no: many times I want a service to be checked multiple times before changing to a hard state so I can avoid an incorrect notification.

You basically have no choice but to deal with what we call the "Nagios Notification Timeline." Just like box293 wrote. Grab a spreadsheet and start plotting your possible service check timings, and you will find your notification events. From there, you can build backwards to adjust your timings and max_checks and whatever other parameters to try to eliminate the possibility of cross-notifications.

You may want to try having one service that doesn't notify, but is the control service for another service that does notify, and make the second one dependent on the first one being in a CRITICAL state. They can both check the same thing, but only the second one notifies and has zero SOFT states. That way, you can have multiple SOFT states on the first one - with no notifications - but the second one doesn't check unless the first one is HARD CRITICAL. Then notify.

You can accomplish the same thing with an event handler that knows how many times the service has failed and in what specific state it is in (HARD/SOFT, etc). Notify from there instead.

There are options, but if you need to be that picky, your options become programming choices.

@amprantino: Did eloyd's post provide any insight?