[Openstack-operators] Upgrading to Havana: nova-manage db sync fails.

Robert van Leeuwen Robert.vanLeeuwen at spilgames.com
Tue Oct 22 11:21:57 UTC 2013


Tried creating the indexes but it start complaining about different ones. 
Seems like the database migration partly worked and became in a inconsistent state.
I fixed it by upgrading python-migrate / sqlalchemy and ran it agains a fresh grizzly database.

I think this could be fixed if the RDO packages: require the correct versions of python-migrate / sqlalchemy.

Cheers,
Robert



________________________________________
From: mikalstill at gmail.com [mikalstill at gmail.com] on behalf of Michael Still [mikal at stillhq.com]
Sent: Tuesday, October 22, 2013 12:07 PM
To: Robert van Leeuwen
Cc: openstack-operators at lists.openstack.org
Subject: Re: [Openstack-operators] Upgrading to Havana: nova-manage db sync fails.

Is that message truncated?

I think that you can just add the index and then retry the upgrade and
you should be ok.

Cheers,
Michael

On Tue, Oct 22, 2013 at 8:07 PM, Robert van Leeuwen
<Robert.vanLeeuwen at spilgames.com> wrote:
> Hi,
>
> I'm trying to upgrade our dev environment to Havana
> When running the nova-manage  db sync command it gives a traceback in migrate.versioning.api [-] 184 -> 185:
>
> OperationalError: (OperationalError) (1091, "Can't DROP 'uniq_task_name_x_host_x_period_beginning_x_period_ending'; check that column/key exists") 'ALTER TABLE task_log DROP INDEX uniq_task_name_x
>
> Database started with Folsom, upgraded to  Grizzly in the past.
> Running on SL6.4 and RDO packages.
>
> Anyone hit this? Suggestion on how to fix this?
>
> Thx,
> Robert van Leeuwen
>
> _______________________________________________
> OpenStack-operators mailing list
> OpenStack-operators at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators



--
Rackspace Australia



More information about the OpenStack-operators mailing list