On 12/12/2011 11:35 PM, Amit Shah wrote:
On (Mon) 12 Dec 2011 [17:00:57], Eric Blake wrote:
> Testing results - upgrading libvirtd with a running domain of type
> fedora-13 passed, and I noticed all persistent configurations were
> properly upgraded as part of the new libvirtd. All attempts to modify a
> running fedora-13 machine were converted to pc-0.13, as expected by the
> code.
>
> HOWEVER, pc-0.13 is wrong. I had one rather significant test failure:
>
>
> succeeded. My conclusion? Qemu must be saving the machine type
> somewhere in the migration file format (virsh save is a migration to
> file, which means you'd see the same behavior when migrating between two
> machines); and pc-0.13 is missing something related to kvmclock, and is
> thus unable to reparse the migration data. But pc-0.14 works just fine,
> which means your patch is _almost_ right - just change fedora-13 to
> pc-0.14, and I think we're set.
The fedora-13 machine type was just a slightly modified pc-0.12
machine type, so using pc-0.12 should be the safest.
That's wrong. The whole reason fedora-13 was introduced was because it
exposed MORE features than stock pc-0.12, so it is closer either to
pc-0.13 or to pc-0.14. I already proved that we can't do migration from
fedora-13 to pc-0.13, but can do migration to pc-0.14, which means that,
at least in F16, fedora-13 is a synonym to pc-0.14, not pc-0.12.
--
Eric Blake eblake(a)redhat.com +1-919-301-3266
Libvirt virtualization library
http://libvirt.org