[osuosl-openpower] Instance migration issues

[Lance Albertson]

I have completed the migration of the instances. I believe this was caused
by the fact that these instances did not have an additional flavor
specification which excludes it from the NVMe nodes. I'm going to work on
fixing those in the coming days so that this doesn't happen again.

If you still have any issues with your VMs still, please let me know.

Thanks!

On Tue, Apr 5, 2022 at 11:30 AM Lance Albertson <lance at osuosl.org> wrote:

> All,
>
> I need to reboot some VM instances today due to an issue we discovered
> while doing live migrations last week. Some instances ended up on a
> hypervisor node that is only configured for local (NVMe) storage.
> Unfortunately, we can't live migrate these instances off the hypervisors
> and will need to do some manual database changes to get them to the correct
> nodes. In addition, it appears that if you reboot the instances, the VM
> might not come back online due to this configuration issue.
>
> I will need to stop these instances, make the change in the database and
> then start the instances on the new hypervisor.
>
> Thankfully, it appears there is no storage loss as long as we get these
> instances back on a ceph-enabled hypervisor.
>
> I'm working through the instances right now and will send an update once
> I've completed the migration.
>
> --
> Lance Albertson
> Director
> Oregon State University | Open Source Lab
>
>

-- 
Lance Albertson
Director
Oregon State University | Open Source Lab
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osuosl.org/pipermail/openpower/attachments/20220405/7f88a1bb/attachment-0001.html>