[Pacemaker] 2 node cluster with clvm, configuration help needed...

[Andrew Beekhof]

On Thu, May 20, 2010 at 9:55 AM,  <Patrik.Rapposch at knapp.com> wrote:
>
>
>
>
>
> hy,
>
> i have a 2 node testcluster running with clvm. I defined one volume group
> and one logical volume which is placed in a group.
>
> with my existing configuration, when i stop openais on one node, the
> resource start of the volume group on the other node fails. and so is the
> group, which is depend on the vg00.

I don't have any experience with clvm
Perhaps add some logging to the vg00 resource to figure out what it's failing?

> when I then do a "cleanup resource" on the remaining node, the vg00 starts
> normaly and everything works.
>
> the next thing is: when i now start openais on the other node, the resource
> vg00 and the dependent group gets restarted and is put on the node which it
> was before restart of the former killed node.
>
> possibly this behaviour is based on my configuration. could you please give
> me an advice how to realize this?
>
> the behaviour should be following:
>
> if one node fails, the other should overtake the resources and keep them,
> even if the other node starts back up. the resources should not be
> restarted, they should just failover and stay where they are,
>
> thx for your advice in advance.
>
> attached you'll find my cib.xml.
>
> (See attached file: cib.xml)
> Mit freundlichen Grüßen / Best Regards
>
> Patrik Rapposch, Bsc.
> Systemadministration
>
> KNAPP Systemintegration GmbH
> Waltenbachstraße 9
> 8700 Leoben, Austria
> Phone: +43 3842 915
> Mobil:
> Fax: +43 3842 82930-990
> patrik.rapposch at knapp.com
> www.KNAPP.com
>
> Commercial register number: FN 138870x
> Commercial register court: Leoben
> _______________________________________________
> 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
>