Yes you can configure it there but the default settings working normally
without problem also on my bigger systems.
The strange behavior on your system is that all checks are triggered at the
same time at 0:00.
One Test i would do is search for all services with the same name "Check_MK
Discovery" and then issue the manual reschedule command for all over a
given timeperiod of 60 minutes or so.

Command Menu -> Various Commands -> Reschedule active checks -> 60 min and
go

Now you should see all the discovery checks spread over 60 minutes and from
now on this should stay the same. If you see again at midnight all checks
at the same time then there is another problem inside your installation.
Something like the core makes an restart at 0:00 and forgets all the
planned states or forgets the old state.

br
Andreas

Gilles Pion <gilles.pion+check...@gmail.com> schrieb am Mo., 12. Feb. 2018,
08:10:

> Strangely, according to the setting of
> "service_inter_check_delay_method"  (set to "s" in the CheckMK
> generated nagios.cfg file) this should not happen
>
> However, I've found that in the same file "max_concurrent_checks is
> set to 0 (which I guess mean "unlimited").
> I thing I'd better tune that setting.
> Can someone confirm if Is ~etc/nagios/nagios.d/timing.cfg the good
> place to customiza this setting?
>
> References:
> https://assets.nagios.com/downloads/nagioscore/docs/nagioscore/4/en/configmain.html
>
> 2018-02-10 8:15 GMT+01:00 Gilles Pion <gilles.pion+check...@gmail.com>:
> > What I need is to stagger the checks (you found the exact word I's was
> > looking for!). The default time frame is OK for me, the real problem
> > is that currently, once per day, a huge number of checks is activated
> > *in the same second* and thus most of them are timing out.
> > Unfortunately Implementing this with different time periods would
> > really not be practical  (there is many hundredth of hosts in my
> > current setup).
> > I'm surprised not having faced this issue with the previous old
> > CheckMK implementation (with even more hosts). Could there be an
> > "maximum-parallel-checks" setting somewhere which have not be
> > reimplemented of that I've missed?
> > Other approach: a little bit of "randomness" in the schedule could be
> > an option too.
> > But I guess those workarounds would be Nagios engine settings, not on
> > CheckMK side.
>
_______________________________________________
checkmk-en mailing list
checkmk-en@lists.mathias-kettner.de
http://lists.mathias-kettner.de/mailman/listinfo/checkmk-en

Reply via email to