high cpu irq load and slow boot after update from 10.4 to 11.2

Gerrit Kühn gerrit.kuehn at aei.mpg.de
Mon Nov 26 13:14:07 UTC 2018


On Mon, 26 Nov 2018 19:34:43 +0700 Eugene Grosbein <eugen at grosbein.net>
wrote about Re: high cpu irq load and slow boot after update from 10.4 to
11.2:


> > Any ideas?

> Maybe this box has some clocking problems incompatible with tickless
> kernel. 

Is there anything I could look out for in dmesg or similar to spot the
root cause for this behaviour? The CPUs are Xeon E5606 on a Supermicro
X8DTU mainboard.

> Try get back to old periodic ticking with sysctl
> kern.eventtimer.periodic=1 instead of now default 0.

I'll try that as soon as I spot the issue again.

> Of, if you are curious, run ntpd if it is not already running, wait
> about an hour then look to its /var/db/ntpd.drift file to see if system
> clock is good or not.

ntpd is always running. Right now it looks ok to me (but the issue is not
there, either).

root at storage:~ # cat /var/db/ntpd.drift
-1.366

> Perhaps, you can get better behaviour changing default value
> of kern.timecounter.hardware to another one from kern.timecounter.choice;
> same with kern.eventtimer.timer and kern.eventtimer.choice

Would that work while I see the issue (i.e., should it make the issue go
away then), or should this be set on (re)boot?

Which settings would be recommended to try? This is what I have now:

---
root at storage:~ # sysctl kern.timecounter.hardware
kern.timecounter.hardware: TSC

root at storage:~ # sysctl kern.timecounter.choice
kern.timecounter.choice: ACPI-safe(850) HPET(950) i8254(0) TSC(1000)
dummy(-1000000)

root at storage:~ # sysctl kern.eventtimer.timer
kern.eventtimer.timer: LAPIC

root at storage:~ # sysctl kern.eventtimer.choice
kern.eventtimer.choice: LAPIC(600) HPET(350) HPET1(340) HPET2(340)
HPET3(340) i8254(100) RTC(0)
---


Thanks for your input.


cu
  Gerrit


More information about the freebsd-stable mailing list