(ccing Kevin)
On Mon, Sep 16, 2019 at 10:14:44 +0100, Daniel Berrange wrote:
> On Fri, Sep 13, 2019 at 02:43:53PM +0200, Peter Krempa wrote:
> > To my knowledge, everything in libvirt is now prepared to fully use
> > -blockdev way to configure disks in qemu.
> >
> > There is one known qemu bug though: Internal snapshots don't work with
> > -blockdev:
> >
> >
https://bugzilla.redhat.com/show_bug.cgi?id=1658981
>
> What are the chances of that bug actually getting fixed any time in the
> forseeable future ?
>
> savevm has been a pain for a long time since it blocks QEMU and we've
> been waiting for it to be fixedd by QEMU devs for at least 7-8 years
> now, and I don't see anyone from QEMU block maintainers commenting on
> this new BZ to indicate they're going to fix this latest problem.
>
> > Since I can't in good faith ask for merging this patchset yet I'd like
> > to give it some more testing I'm suggesting that we push it and revert
> > it during freeze or add a capability check once qemu is fixed.
>
> I'd only be in favour of enabling it if there is some clear time frame
> on whicht hat bug above is likely to be fixed. I'm guesing it is pretty
> unlikely to be fixed before our freeze, so I thinking speculatively
> enabling it is premature.
I'm not sure. Kevin, any idea when we can expect the fix?
Otherwise the only option to enable blockdev would be to just enable it
and blame qemu for internal snapshots not working.
The main problem with implementing a fix was that it's unclear what the
desired result even is. After some discussion with Max, I think we'll
try it with restricting snapshotting to everything that is directly
attached to a BlockBackend and monitor-owned nodes that don't have any
parents.
Kevin