Escalating an event to critical right away will actually defeat the
basic design behind nocol (avoid false alarms).
However, you can get notified of particular hosts right away by adding
a line in the noclogd to get paged at the WARNING level for particular
hosts (just run an external program which ignores all except what are
critical hosts).
-vikas
Stefan `Sec` Zehl wrote:
> Our webcache frequently crashes (we are debugging it, but i have no
> idea why) in which case we need to restart it.
> If I assign it 'critical' level in portmon-conf, it slowly rises from
> 'warning','error' to critical (one with each check). As portmon rechecks
> every 10 minutes means that our webcache is 30 minutes down before
> someone will get paged. Is there a way to circumvent this 'slow rise'
> or, even better, a way to trigger an earlier recheck in case something
> fails ?(so fatal errors would rise more quickly)
|