Nagios Support Forum

One of the problems with the CPU check is that sometimes 100% CPU utilization is OK. Generally LOADAVE is a much better indicator of system that's saturated and thus a more worthy plugin for notification sending.

CPU checks can cause too many notifications because of the ceiling. You can't set a threshold greater than 100% utilization (or 0% idle, is it also may be measured).

So I'm thinking it would be good to have a critical_max_check_attempts for CPU. This way, we could have a state change based on number of consecutive non-OK checks, and not necessarily the value of the return result.

I guess this could be achieved with a smart plugin, that tracked its own attempt count, returning 2 when it achieved critical_max_check_attempts. But also may be useful as a server config. What do you think ?

I think this is the job of a plugin, not the nagios process itself. If it were to be implemented in the nagios process, we would have a near-endless amount of config options for every plugin, and some people may not even use a given plugin. This leads to a lot of clutter in the configs.