[Pacemaker] 2 Node Cluster: some principal questions

[Andreas Kurz]

On 03/22/2012 04:47 PM, Lentes, Bernd wrote:
> Hi,
>  
> thanks for the quick response. So the data is not destroyed if two
> identicals VM's access the LV simultaneouslay ? I thought it would be. I
> will have stonith too.
> What is about live migration ? Is it possible with my setup ?

Well, if you want live migration you must _not_ use an "exclusive" vg
but allow both nodes to access it simultaneously.

It is also possible to balance the VMs across the nodes.

The most important part in your setup regarding data protection is a
working STONITH setup ... and a correctly configured clvm setup will not
allow a node to activate a vg and its lvs without consensus about the
cluster state and it coordinates changes to the metadata but it does not
coordinate concurrent access to the data.

That is within the responsibility of Pacemaker to start VMs only on one
node ... and Pacemaker relies on STONITH to be certain about node states.

Regards,
Andreas

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


>  
>  
> Bernd
>  
> 
> --
> Bernd Lentes
> 
> Systemadministration
> Institut für Entwicklungsgenetik
> HelmholtzZentrum münchen
> bernd.lentes at helmholtz-muenchen.de
> phone: +49 89 3187 1241
> fax:   +49 89 3187 3826
> http://www.helmholtz-muenchen.de/idg
> 
> Wer Visionen hat, soll zum Hausarzt gehen
> Helmut Schmidt
> 
>  
> 
>     ------------------------------------------------------------------------
>     *From:* emmanuel segura [mailto:emi2fast at gmail.com]
>     *Sent:* Thursday, March 22, 2012 4:12 PM
>     *To:* The Pacemaker cluster resource manager
>     *Subject:* Re: [Pacemaker] 2 Node Cluster: some principal questions
> 
>     if you have in you cluster vg exclusive="yes" and remember it's good
>     idea have stonith configured
> 
>     Il giorno 22 marzo 2012 15:52, Lentes, Bernd
>     <bernd.lentes at helmholtz-muenchen.de
>     <mailto:bernd.lentes at helmholtz-muenchen.de>> ha scritto:
> 
>         Hi,
> 
>         i'd like to establish a 2 Node-Cluster. My OS will be SLES 11
>         and HAE. My services will run in VM's using KVM. I will install
>         the VM's in LV's, one LV for one VM. The LV's will be plain
>         partitions, having no filesystem. I have read this is quicker
>         with KVM than installing a VM in a LV with filesystem.
>         My LV's will reside on a FC SAN. Both nodes have access to the
>         SAN. For the LV's, i will use cLVM.
> 
>         My questions are:
> 
>         - let's assume that, due to an error, a VM will run on both
>         nodes. Does cLVM prevent that both VM's have simultaneously
>         access to the corresponding LV ? So that the data is not destroyed.
> 
>         - Is it possible that some VM's run on the first node and other
>         VM's run on the second node simultaneously ?
> 
>         - what do you think about my setup ?
> 
> 
> 
>         Bernd
> 
>         --
>         Bernd Lentes
> 
>         Systemadministration
>         Institut für Entwicklungsgenetik
>         HelmholtzZentrum münchen
>         bernd.lentes at helmholtz-muenchen.de
>         <mailto:bernd.lentes at helmholtz-muenchen.de>
>         phone: +49 89 3187 1241 <tel:%2B49%2089%203187%201241>
>         fax:   +49 89 3187 3826 <tel:%2B49%2089%203187%203826>
>         http://www.helmholtz-muenchen.de/idg
> 
>         Wer Visionen hat, soll zum Hausarzt gehen
>         Helmut Schmidt
> 
>         Helmholtz Zentrum München
>         Deutsches Forschungszentrum für Gesundheit und Umwelt (GmbH)
>         Ingolstädter Landstr. 1
>         85764 Neuherberg
>         www.helmholtz-muenchen.de <http://www.helmholtz-muenchen.de>
>         Aufsichtsratsvorsitzende: MinDir´in Bärbel Brumme-Bothe
>         Geschäftsführer: Prof. Dr. Günther Wess und Dr. Nikolaus Blum
>         Registergericht: Amtsgericht München HRB 6466
>         USt-IdNr: DE 129521671
> 
>         _______________________________________________
>         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
> 
> 
> ------------------------------------------------------------------------
> Helmholtz Zentrum München
> Deutsches Forschungszentrum für Gesundheit und Umwelt (GmbH)
> Ingolstädter Landstr. 1
> 85764 Neuherberg
> www.helmholtz-muenchen.de
> Aufsichtsratsvorsitzende: MinDir´in Bärbel Brumme-Bothe
> Geschäftsführer: Prof. Dr. Günther Wess und Dr. Nikolaus Blum
> Registergericht: Amtsgericht München HRB 6466
> USt-IdNr: DE 129521671
> 
> 
> _______________________________________________
> 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/20120322/54fdf0ad/attachment-0003.sig>