[Pacemaker] Xen, Pacemaker and live migration

[Andreas Kurz]

Hello Frank,

On 01/21/2012 05:25 PM, emmanuel segura wrote:
> Hello Frank
> 
> The first thing I found wrong in your config is
> 
> =====================================================
> colocation CLVMCloneWithLVMforVM1 inf: LVMforVM1 clvm-clone
> =====================================================

That part is ok, LVMforVM1 depends on clvm-clone ... see bellow for some
comments ...

> 
> clvm-clone must be a resource running on every cluster node, not just
> where lvm for the vm1 it's running
> 
> 2012/1/20 Frank Meier <Frank.Meier at hr-group.de
> <mailto:Frank.Meier at hr-group.de>>
> 
>     Hi there,
> 
>     I'm new in clustering and
>     I have problems to get xen live migration running.
> 
>     I configure cLVM and manual migration works, but xen goes down and
>     starts on the second node new.
> 
>     node xencluster1 \
>            attributes standby="off"
>     node xencluster2 \
>            attributes standby="off"
>     primitive LVMforVM1 ocf:heartbeat:LVM \
>            meta target-role="started" is-managed="true" \
>            operations $id="LVMforVM1-operations" \
>            op monitor interval="10" timeout="30" \
>            params volgrpname="vm1VG" exclusive="yes"
>     primitive STONITH--Node1 stonith:meatware \
>            meta target-role="Started" \
>            operations $id="STONITH--Node1-operations" \
>            op monitor interval="15" timeout="15" start-delay="15" \
>            params hostlist="xencluster1.reno.de"

The hostlist needs to inculde the node names as used by
Pacemaker/Corosync ... so xencluster1, xenscluster2

>     primitive STONITH-Node2 stonith:meatware \
>            meta target-role="started" \
>            operations $id="STONITH-Node2-operations" \
>            op monitor interval="15" timeout="15" start-delay="15" \
>            params hostlist="xencluster2.reno.de
>     <http://xencluster2.reno.de>"
>     primitive VM1 ocf:heartbeat:Xen \
>            meta allow-migrate="true" target-role="started" \
>            operations $id="VM1-operations" \
>            op monitor interval="10" timeout="30" \
>            params xmfile="/etc/xen/vm/VM1"
>     primitive clvm ocf:lvm2:clvmd \
>            operations $id="clvm-operations" \
>            op monitor interval="10" timeout="20" \
>            params daemon_timeout="30"
>     primitive dlm ocf:pacemaker:controld \
>            operations $id="dlm-operations" \
>            op monitor interval="10" timeout="20" start-delay="0" \
>            meta target-role="started"
>     clone clvm-clone clvm \
>            meta target-role="started" ordered="true" interleave="true"
>     clone dlm-clone dlm \
>            meta target-role="started" ordered="true" interleave="true"
>     location STONITHnode1Constraint STONITH--Node1 -inf: xencluster1
>     location cli-prefer-VM1 VM1 \
>            rule $id="cli-prefer-rule-VM1" inf: #uname eq xencluster2
>     colocation CLVMCloneWithLVMforVM1 inf: LVMforVM1 clvm-clone
>     colocation DLMCloneWithcLVMClone inf: dlm-clone clvm-clone

The clvm depends on the dlm ... inf: clvm-clone dlm-clone

>     colocation VM1WithLVM1forVM1 inf: VM1 LVMforVM1
>     order DLMCloneBeforecLVMClone : dlm-clone clvm-clone
>     order LVMforVM1BeforeVM1 : LVMforVM1 VM1
>     order cLVMCloneBeforeLVMforVM1 : clvm-clone LVMforVM1
>     property $id="cib-bootstrap-options" \
>            dc-version="1.1.2-2e096a41a5f9e184a1c1537c82c6da1093698eb5" \

It's a good idea to update.

>            no-quorum-policy="ignore" \
>            stonith-enabled="false" \

Don't forget to enable stonith.

>            default-resource-stickiness="100000" \
>            last-lrm-refresh="1327076030" \
>            cluster-infrastructure="openais" \
>            expected-quorum-votes="2"
> 
> 
>     Any clue?

What is the error you see when running the migration? For a live system
I would expect the default timeout of 20s for migration actions to be
too low ... any errors because of timeouts in "crm_mon -1fr"?

Regards,
Andreas

-- 
Need help with Pacemaker?
http://www.hastexo.com/now

> 
> 
> 
>     Mit freundlichen Grüßen
> 
>     Frank Meier
>     UNIX-Basis
> 
>     Hamm Reno Group GmbH
>     Industriegebiet West | D-66987 Thaleischweiler-Fröschen
>     T.+49(0)6334 444-322 <tel:%2B49%280%296334%20444-322> | F.+49(0)6334
>     444-190 <tel:%2B49%280%296334%20444-190>
>     frank.meier at hr-group.de <mailto:frank.meier at hr-group.de> |
>     www.my-reno.com <http://www.my-reno.com>
>     ___________________________________________________________________
> 
>     Sitz: Am Tie 7 | D-49086 Osnabrück
>     Handelsregister Osnabrück HRB 19587
>     Geschäftsführer: Hans-Jürgen de Fries,
>     Jens Gransee, Manfred Klumpp,
>     Robert Reisch, Albert Scherer
> 
>     _______________________________________________
>     Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
>     <mailto: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
> 
> 
> 
> 
> -- 
> esta es mi vida e me la vivo hasta que dios quiera
> 
> 
> 
> _______________________________________________
> 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


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 222 bytes
Desc: OpenPGP digital signature
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20120123/1e7c9a96/attachment-0004.sig>