Nagios 4 Load issues

[abrist]

and possibly the auto scheduler.

Definitely. As of now, it may be best to turn off the auto scheduler and deal with the elevated load levels instead of the possible check latency issues of the auto-scheduler . . . .

[liquidcool]

abrist,

Thank you for getting back.

I think what we will do is keep the it enabled because even though the load does fluctuate it does not get as high as it would when it is turned off. Most of the time the load actually is fine, as long as when I see it starting to get out of control, I just switch it off for a few hours and then enable it again, it starts working fine.

So are you confirming there is an issue with the auto scheduler ?

If you need me to do any further testing please do not hesitate to ask me and I will see what I can do.

At the moment I have set these settings to see if it has any affect:
auto_rescheduling_window=360 (from 180 - the default)
auto_rescheduling_interval=30 (I had set it to 60)

[abrist]

So are you confirming there is an issue with the auto scheduler ?

Essentially, yes.

At the moment I have set these settings to see if it has any affect:
auto_rescheduling_window=360 (from 180 - the default)
auto_rescheduling_interval=30 (I had set it to 60)

I find the most check execution issues when the window is set to more seconds than you shortest check (which for most installs is a 1 minute retry).
So, if you are going to run the auto-rescheduler, I suggest setting the window to like 45 seconds instead of 360.

Are all you checks running on time as expected?

[liquidcool]

Ok, I will set the window as you suggest and see what happens.

From what I can see all the checks run as expected, with very good latency timings. When the scheduler is behaving itself, which I have been able to get to about 6-7 hours now before the load spikes. There are none that hit the timeout threshold of 60 seconds ... that only happened when the load ended up hitting about 700 that one time ...

[abrist]

Ah, well keep us abreast of the situation. More news about ndo to come!

[liquidcool]

I noticed that you guys have released 4.1.0 rc1. Have any of these issue about load and autoscheduling been addressed in this release ? I don't see anything specific in the change log.

[abrist]

No, my suggestion is still to avoid the auto_reschedule(r) if you have issues with it.