[Pacemaker] Corosync hanging during stop

[Andrew Beekhof]

On 17 Oct 2013, at 8:05 pm, D.Gossrau <dgossrau at andtek.com> wrote:

> Hi Lars,
> 
> On 10/12/2013 02:14 AM, Lars Ellenberg wrote:
>> On Thu, Oct 10, 2013 at 04:06:46PM +0200, Detlef Gossrau wrote:
>>> Hi,
>>> 
>>> I created a cluster installation with two nodes. Everything is
>>> running smoothly most of the time. But in case a resource was
>>> switched into unmanaged state and I reboot the node on which the
>>> resource is running, corosync is hanging during stop. Only after
>>> 'shutdown-escalation' time corosync is stopping.
>> 
>> *corosync* is hanging?
>> or rather *pacemaker*, maybe?
> 
> I assume corosync is hanging. On the console I see
> 'Waiting for corosync services to unload: ...'
> 
> The problem seems to be a resource that takes to long for stopping. In that case the resource is set to unmanage and is blocking corosync.
> 
> Is there a general possibility to get informed if a resource is changing the state to unmanage ? Other state changes I observe via ClusterMon. But the state change to unmanage is not reported.

Not currently, but might be a useful feature.
If you file a bug for it, it will get added to the todo list :)

> 
> Kind regards,
> Detlef
> 
> 
> _______________________________________________
> 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