Open Stack

Hi Jean-Philippe,

we had the same issue with ceph as backend. This fixed the problem in our
setup:
https://ask.openstack.org/en/question/87545/cinder-high-availability/

Although the above link talks about an active-active setup, the official
docs mention the hostname in the configuration also for an active-passive
setup:
https://docs.openstack.org/ha-guide/storage-ha-block.html#configure-block-storage-api-service

regards
Christian Zunker

Jean-Philippe Méthot <jp.methot at planethoster.info> schrieb am Fr., 6. Juli
2018 um 07:11 Uhr:

> Hi,
>
> We’ve been running on Openstack for several years now and our setup has
> always counted a single controller. We are currently testing switching to a
> dual controller HA solution, but an unexpected issue has appeared,
> regarding storage. See, we use Dell compellent SAN for our block devices. I
> notice that when I create a volume on one controller, I am unable to make
> any operation on the same volume on the second controller (this is with an
> active/passive cinder-volume). Worse, this affects VMs directly as they
> can’t be migrated if the active controller isn’t the one that created their
> block device.
>
> I know this issue doesn’t happen on Ceph, so I’ve been wondering, is this
> a limitation of Openstack or the SAN driver? Also, is there actually a way
> to reach even active-passive high availability with this current storage
> solution?
>
>
> Jean-Philippe Méthot
> Openstack system administrator
> Administrateur système Openstack
> PlanetHoster inc.
>
>
>
>
> _______________________________________________
> OpenStack-operators mailing list
> OpenStack-operators at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20180706/45a66b5e/attachment.html>