[Pacemaker] Master/Slave failover during reboot

[Eliot Gable]

Yes, I realized that after upgrading to 1.0.4. When using 1.0.3, when I simply issued a 'reboot' command at the prompt, the node would go offline in UNCLEAN state. With 1.0.4, it just shows as OFFLINE without the UNCLEAN modifier. I am just going to assume that I somehow botched the 1.0.3 install and not worry about, since it seems to be doing the right thing now.


-----Original Message-----
From: Lars Marowsky-Bree [mailto:lmb at suse.de]
Sent: Tuesday, July 21, 2009 9:00 AM
To: pacemaker at oss.clusterlabs.org
Subject: Re: [Pacemaker] Master/Slave failover during reboot

On 2009-07-20T15:27:23, Eliot Gable <egable at broadvox.com> wrote:

> I have a resource that is configured as a Master/Slave resource. If I
> kill a resource it is dependent on, it properly fails over to the
> other node. However, if I reboot the master node, it does not fail
> over. What I see is that the master node switches to UNCLEAN -
> Offline, the master resource stops running (crm_mon shows only the
> slave node running) and then it just sits there until the master node
> finishes booting. Once the rebooted node re-joins the cluster, it
> figures out which is master/slave and one of them gets promoted. The
> entire time, the partition says it has quorum. This is Pacemaker
> 1.0.4.

Nodes in "Unclean - offline" state should be fenced. Since you make no
allusion to fencing/STONITH occuring, I wonder if you have fencing
configured?


Regards,
    Lars

--
Architect Storage/HA, OPS Engineering, Novell, Inc.
SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg)
"Experience is the name everyone gives to their mistakes." -- Oscar Wilde


_______________________________________________
Pacemaker mailing list
Pacemaker at oss.clusterlabs.org
http://oss.clusterlabs.org/mailman/listinfo/pacemaker

CONFIDENTIAL.  This e-mail and any attached files are confidential and should be destroyed and/or returned if you are not the intended and proper recipient.