Re: [libvirt] [PATCH 5/6] allow non-zero success values from a domain's MigratePerform
On Wed, Oct 07, 2009 at 04:54:27PM +0200, Paolo Bonzini wrote:
> Excellent, sounds fine then, but we need the rebased version once
>Dan patches are applied.
Of course. But it sounds like Dan Berrange doesn't fancy 5/6 very much
because he wants the "old virsh, new daemon" case to work too. He also
didn't like the hack I posted (which I understand even more...).
For v2, I could add a driver feature saying whether the driver supports
VIR_MIGRATE_PAUSED, and if it does I'll do the resume from
virDomainMigrate. This will fix the QEMU driver, and will cost (almost)
nothing if the driver can actually transfer the suspended/running state.
That sounds like a reasonable approach. If the client is too old, then
everything will work as currently. If the server is too old, the client
can report a error to caller that this capability isn't available. If
the client & server are new,then everything will work
Daniel
--
|: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :|
|: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :|
|: http://autobuild.org -o- http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|