[Pacemaker] pacemaker-1.0.10 - failure of resource causes migration of other resources

[Dejan Muhamedagic]

Hi,

On Wed, Dec 29, 2010 at 11:15:12PM +0100, Nikola Ciprich wrote:
> Hello,
> while doing some tests of 1.0.10, I've noticed strange behaviour.

Does that mean that you didn't observe this with earlier
versions?

> I have lots of mutualy unrelated resources (virtual machines) + some other needed stuff (DRBD, CLVM, etc)
> when one of virtual machines fails on a node, all other machines are migrated.
> I'm not sure this is proper behaviour, and if it is, can I somehow disable it?

No, that shouldn't happen.

> In case hb_report could help, I've uploaded it here:
> http://nelide.cz/downloads/hbreport-101229.tar.bz2
> If somebody could take a look at this, I'd be very grateful.

What I've noticed are these strange log messages:

Dec 29 22:55:07 vbox1 pengine: [28757]: notice: clone_rsc_colocation_rh: Cannot pair clvmd:1 with instance of cl-dlm

Even though all clones run on both nodes, there seems to be some
confusion. For instance:

Dec 29 22:55:00 vbox1 pengine: [28757]: info: find_clone: Internally renamed clvmd:0 on vbox1 to clvmd:2 (ORPHAN)
Dec 29 22:55:00 vbox1 pengine: [28757]: info: clone_rsc_order_lh: Inhibiting clvmd:2 from being active
Dec 29 22:55:00 vbox1 pengine: [28757]: notice: LogActions: Leave resource clvmd:0      (Started vbox2)

Does this reproduce from start or only after the cluster has been
running for a while?

Otherwise, this will have to wait for Andrew.

Thanks,

Dejan

> Of course, if I can provide any further needed information or be somehow
> helpful, I'll be more then happy to provid what I can.
> with best regards
> 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