[Pacemaker] 1.0.7 upgraded, restarting resources problem

Martin Gombač martin at isg.si
Fri Feb 5 15:29:51 UTC 2010


Bug was confirmed. :-)
http://developerbugs.linux-foundation.org/show_bug.cgi?id=2317

Hopefully fix will follow soon.

Regards,
M.

Martin Gombač wrote:
> Shoud i file a bug for the problem (restarting resources) described 
> below?
> Regards,
> M.
>
> Dejan Muhamedagic wrote:
>> Hi,
>>
>> On Mon, Jan 18, 2010 at 10:00:39PM +0100, Martin Gombač wrote:
>>  
>>> Hi,
>>>
>>> i have one m/s drbd resource and one Xen instance on top. Both m/s
>>> are primary.
>>> When i restart node that's _not_ hosting the Xen instance (ibm1),
>>> pacemaker restarts running Xen instance on the other node (ibm2).
>>> There is no need to do that. I thought it got fixed
>>> (http://developerbugs.linux-foundation.org/show_bug.cgi?id=2153).
>>> Didn't it?
>>>
>>> Here is my config once more. Please note the WARNING showed up only
>>> after upgrade.
>>> (BTW setting drbd0predHosting score to 0 doesn't restart it. But it
>>> doesn't help resource ordering either.)
>>>
>>> [root at ibm1 etc]# crm configure show
>>> WARNING: notify: operation name not recognized
>>>     
>>
>> That's from the shell, please ignore it. Strange, the operation
>> list should've been updated a long time ago.
>>
>> Thanks,
>>
>> Dejan
>>
>>  
>>> node $id="3d430f49-b915-4d52-a32b-b0799fa17ae7" ibm2
>>> node $id="4b2047c8-f3a0-4935-84a2-967b548598c9" ibm1
>>> primitive Hosting ocf:heartbeat:Xen \
>>>    params xmfile="/etc/xen/Hosting.cfg" shutdown_timeout="303" \
>>>    meta target-role="Started" allow-migrate="true" is-managed="true" \
>>>    op monitor interval="120s" timeout="506s" start-delay="5s" \
>>>    op migrate_to interval="0s" timeout="304s" \
>>>    op migrate_from interval="0s" timeout="304s" \
>>>    op stop interval="0s" timeout="304s" \
>>>    op start interval="0s" timeout="202s"
>>> primitive drbd_r0 ocf:linbit:drbd \
>>>    params drbd_resource="r0" \
>>>    op monitor interval="15s" role="Master" timeout="30s" \
>>>    op monitor interval="30s" role="Slave" timeout="30s" \
>>>    op stop interval="0s" timeout="501s" \
>>>    op notify interval="0s" timeout="90s" \
>>>    op demote interval="0s" timeout="90s" \
>>>    op promote interval="0s" timeout="90s" \
>>>    op start interval="0s" timeout="255s"
>>> ms ms_drbd_r0 drbd_r0 \
>>>    meta notify="true" master-max="2" inteleave="true"
>>> is-managed="true" target-role="Started"
>>> order drbd0predHosting inf: ms_drbd_r0:promote Hosting:start
>>> property $id="cib-bootstrap-options" \
>>>    dc-version="1.0.7-b1191b11d4b56dcae8f34715d52532561b875cd5" \
>>>    cluster-infrastructure="Heartbeat" \
>>>    stonith-enabled="false" \
>>>    no-quorum-policy="ignore" \
>>>    default-resource-stickiness="10" \
>>>    last-lrm-refresh="1263845352"
>>>
>>> All i want is to have just one resource Hosting started, after drbd
>>> was promoted(/primary) on the node that's it's starting.
>>> Please advise me if you can.
>>>
>>> Thank you,
>>> regards,
>>> M.
>>>
>>> _______________________________________________
>>> Pacemaker mailing list
>>> Pacemaker at oss.clusterlabs.org
>>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>>     
>>
>> _______________________________________________
>> Pacemaker mailing list
>> Pacemaker at oss.clusterlabs.org
>> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>>   
>
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker




More information about the Pacemaker mailing list