[Nagiosplug-help] question about failing check

[C. Bensend]

> Is there a way to change the interval of the check if it fails the first
> time though? If it fails, instead of waiting another 5 minutes to check
> again, can the check run in 10 seconds? For example, if apache doesn’t
> respond, check again in 10 seconds to make sure it isn't just a network
> lag
> or hiccup, and if that fails then execute the event handler? From my
> understanding, the failure count can be set but the check interval is a
> constant, or am I wrong on that? Thank you for the reply.

Nagios does that itself (see retry_interval).

I would recommend against 10 seconds; chances are whatever problem
you're experiencing would still be there, and by default Nagios
doesn't really have that fine-grained timing.

I run with a great majority of my services on a five-minute
check_interval, and a one-minute retry_interval.  That means that
services are checked once per five minutes normally, and then
retried once per minute when they are in a non-OK soft state.  That
seems to be a happy medium for me for catching problems promptly,
while trying to minimize nuisance alerts.

Benny


-- 
"Hairy ape nads."        -- Colleen, playing Neverwinter Nights