[Pacemaker] preventing cluster doing any action when load reaches some threshold

Dejan Muhamedagic dejanmm at fastmail.fm
Mon Jun 20 10:12:30 UTC 2011


Hi,

On Sun, Jun 19, 2011 at 04:34:24PM +0200, Nikola Ciprich wrote:
> Hello Andrew et al,
> 
> few times, it happened to me, that cluster node got loaded too much
> (especially I/O load), so cluster actions (monitor) started to timeout.
> So cluster manager decided to restart services etc, thus causing even
> higher load (especially when cluster has tens of services). Is there 
> any way, to tell that when loadavg reaches some threshold (ie 10), 
> manager should postpone all actions?

No.

> Or is there some more elegant way to deal with this kind of situation?

What do your monitor timeouts look like? You should probably
adjust them. Take a look here for some notes about the problem:

http://www.linux-ha.org/wiki/File:Linuxtag-09-ha-paper.pdf
http://www.advogato.org/person/lmb/diary.html

Thanks,

Dejan

> Thanks a lot in advance!
> nik
> 
> 
> -- 
> -------------------------------------
> Ing. Nikola CIPRICH
> LinuxBox.cz, s.r.o.
> 28. rijna 168, 709 01 Ostrava
> 
> tel.:   +420 596 603 142
> fax:    +420 596 621 273
> mobil:  +420 777 093 799
> 
> www.linuxbox.cz
> 
> mobil servis: +420 737 238 656
> email servis: servis at linuxbox.cz
> -------------------------------------



> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> 
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker





More information about the Pacemaker mailing list