[PATCH 09/33] qemu_migration.c: forbid powernv domains migration
Daniel P. Berrangé
berrange at redhat.com
Tue Jan 25 15:00:19 UTC 2022
On Mon, Jan 24, 2022 at 10:09:30AM -0300, Daniel Henrique Barboza wrote:
>
>
> On 1/21/22 11:33, Daniel P. Berrangé wrote:
> > On Thu, Jan 20, 2022 at 10:52:12AM -0300, Daniel Henrique Barboza wrote:
> > > The PowerNV machine does not implement any form of migration.
> >
> > What do you mean by that ?
> >
> > Migration is a general feature in QEMU, not typically something
> > that a machine types opts in/out of.
>
> What I meant with this patch is that migration, albeit possible, will not work.
> These emulations doesn't implement vmstate() for their internal states or its
> specific devices.
>
> The reason is a more complicated version of "because no one bother to do it". It's
> technically doable of course (as you said, every QEMU machine is migratable per
> default) but not something that we have in our long-term scope.
>
>
> This patch had the intention to tell users to 'don't even bother migrating these
> domains, it doesn't work'. We can live without it and let QEMU error out normally.
>
>
> >
> > It is possible for devices to register migration blockers to
> > prevent it, but libvirt shouldn't try to second guess that.
>
>
> I agree that this kind of handling is something that belongs to QEMU and it's
> best that Libvirt doesn't speculate about it.
>
>
> >
> > Overall I'd like to see a clear justification for why libvirt
> > should enforce a policy here, as opposed to letting QEMU
> > accept or reject the migration.
>
>
> I'm fine with dropping this patch.
Ok, lets drop this one and rely on QEMU reporting the lack of support.
Regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
More information about the libvir-list
mailing list