On 10/7/16 15:03, [email protected] wrote:
When I add "for X cycles" to my statements in monitrc, e.g.,
if cpu usage > 95% for 5 cycles then alert
It has no effect. It alerts every cycle as if it was "for 1 cycles."
Is there something else I need to do for the limit to be reached multiple times
before alerting?
I am running monit-5.19.0 on Linux 3.19.8.
The CPU resource test doesn't use the word 'usage' in it - it's either just
if cpu > 95% [etc] or
if total cpu([user|system|wait]) > 95% [etc]
I found cpu too sensitive most of the time in my use-case, and stick
with just loadavg -
if loadavg (5min) is greater than 3.0 for 2 cycles then alert
that gives you a smooth(er) window against to test if the system is
really under continuous load or just dealing with a very brief spike or
spikes.
--
Paul Theodoropoulos
www.anastrophe.com
--
To unsubscribe:
https://lists.nongnu.org/mailman/listinfo/monit-general
