[Pacemaker] Problem during Cluster Upgrade
Dan Frincu
df.cluster at gmail.com
Thu Apr 12 08:54:44 UTC 2012
Hi,
On Wed, Apr 11, 2012 at 5:26 PM, Karl Rößmann <K.Roessmann at fkf.mpg.de> wrote:
> Hi all,
>
> I'm upgrading a three node cluster from SLES 11 SP1 to SLES SP2 node by
> node.
> the upgrade includes:
> corosync-1.3.3-0.3.1 to corosync-1.4.1-0.13.1
> pacemaker-1.1.5-5.9.11.1 to pacemaker-1.1.6-1.27.26
> kernel 2.6.32.54-0.3-xen to 3.0.13-0.27-xen
>
> After Upgrading the first node and restarting the cluster I get these
> never ending messages on the DC (which is not the updated node)
>
> Apr 11 14:19:26 orion14 corosync[6865]: [TOTEM ] Type of received message
> is wrong... ignoring 6.
> Apr 11 14:19:27 orion14 corosync[6865]: [TOTEM ] Type of received message
> is wrong... ignoring 6.
> Apr 11 14:19:28 orion14 corosync[6865]: [TOTEM ] Type of received message
> is wrong... ignoring 6.
> Apr 11 14:19:29 orion14 corosync[6865]: [TOTEM ] Type of received message
> is wrong... ignoring 6.
I think the question relates more to corosync, added the proper group in CC.
>
> the updated node is still in STANDBY mode.
> Should I ignore the message and put the mode to ONLINE ?
> I don't want the cluster to crash, there are running services
> on the other two nodes.
> So now I stopped the openais on the updated node: no more messages.
> the other two nodes are still up and working.
>
> Any ideas ?
I don't know exactly if a rolling upgrade is possible (I may be wrong
on this one) but putting the cluster in maintenance-mode, upgrading
corosync and pacemaker on all 3 nodes and then re-probing for the
resources is a more common upgrade path. If there are no issues on the
reprobe, then you could take the cluster out of maintenance-mode.
Also, do you have a support contract with Suse? I think their support
can help out more on this.
HTH,
Dan
>
> Karl
>
>
>
> --
> Karl Rößmann Tel. +49-711-689-1657
> Max-Planck-Institut FKF Fax. +49-711-689-1632
> Postfach 800 665
> 70506 Stuttgart email K.Roessmann at fkf.mpg.de
>
> _______________________________________________
> 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://bugs.clusterlabs.org
--
Dan Frincu
CCNA, RHCE
More information about the Pacemaker
mailing list