[Users] ids sanlock error

[Itamar Heim]

On 06/26/2013 02:55 PM, Tony Feldmann wrote:
> Ii won't let me move the hosts as they have gluster volumes in that cluster.

so the cluster is both virt and gluster?

>
>
> On Wed, Jun 26, 2013 at 6:49 AM, Itamar Heim <iheim at redhat.com
> <mailto:iheim at redhat.com>> wrote:
>
>     On 06/26/2013 04:35 AM, Tony Feldmann wrote:
>
>         I was messing around with some things and force removed my DC.  My
>         cluster is still there with the 2 gluster volumes, however I
>         cannot move
>         that cluster into a new dc, I just get the following error in
>         engine.log:
>
>         2013-06-25 20:26:15,218 ERROR
>         [org.ovirt.engine.core.bll.__AddVdsSpmIdCommand]
>         (ajp--127.0.0.1-8702-2)
>         [7d1289a6] Command
>         org.ovirt.engine.core.bll.__AddVdsSpmIdCommand throw
>         exception: org.springframework.dao.__DuplicateKeyException:
>         CallableStatementCallback; SQL [{call insertvds_spm_id_map(?, ?,
>         ?)}];
>         ERROR: duplicate key value violates unique constraint
>         "pk_vds_spm_id_map"
>             Detail: Key (storage_pool_id,
>         vds_spm_id)=(084def30-1e19-__4777-9251-8eb1f7569b53, 1) already
>         exists.
>             Where: SQL statement "INSERT INTO
>         vds_spm_id_map(storage_pool___id,
>         vds_id, vds_spm_id)
>               VALUES(v_storage_pool_id, v_vds_id, v_vds_spm_id)"
>
>
>         I would really like to get this back into a dc without destroying my
>         gluster volumes and losing my data.  Can anyone please point me
>         in the
>         right direction?
>
>
>     but if you removed the DC, moving the cluster is meaningless - you
>     can just create a new cluster and move the hosts to it?
>     (the VMs reside in the DC storage domains, not in the cluster)
>
>     the above error message looks familiar - i think there was a bug
>     fixed for it a while back
>
>
>         On Tue, Jun 25, 2013 at 8:19 AM, Tony Feldmann
>         <trfeldmann at gmail.com <mailto:trfeldmann at gmail.com>
>         <mailto:trfeldmann at gmail.com <mailto:trfeldmann at gmail.com>>> wrote:
>
>              I have a 2 node cluster with engine running on one of the
>         nodes.  It
>              has 2 gluster volumes that replicate between the hosts as
>         its shared
>              storage.  Last night one of my systems crashed.  It looks
>         like all
>              of my data is present, however the ids file seems to be
>         corrupt on
>              my master domain.  I tried to do a hexdump -c on the ids
>         file, but
>              it just gave an input/output error.  Sanlock.log shows
>         error -5.  Is
>              there a way to rebuild the ids file, or can I tell ovirt to
>         use the
>              other domain as the master so I can get back up and running?
>
>
>
>
>         _________________________________________________
>         Users mailing list
>         Users at ovirt.org <mailto:Users at ovirt.org>
>         http://lists.ovirt.org/__mailman/listinfo/users
>         <http://lists.ovirt.org/mailman/listinfo/users>
>
>
>