On 06/12/2014 05:51 AM, Peter Krempa wrote:
On 06/11/14 18:27, Eric Blake wrote:
> A future patch is going to wire up qemu active block commit jobs;
> but as they have similar events and are canceled/pivoted in the
> same way as block copy jobs, it is easiest to track all bookkeeping
> for the commit job by reusing the <mirror> element. This patch
> adds domain XML to track which job was responsible for creating a
> mirroring situation, and adds a job='copy' attribute to all
> existing uses of <mirror>. Along the way, it also massages the
> qemu monitor backend to read the new field in order to generate
> the correct type of libvirt job (even though it requires a
> future patch to actually cause a qemu event that can be reported
> as an active commit).
>
>
ACK, although this patch pointed me to a problem. If you do a block-copy
of a network disk to a local file (which seems from the code that it's
allowed) and the mirror is being pivoted to the new disk we don't update
the disk type in the code and thus would end up with a very damaged VM.
Fortunately for block-copy this isn't a major problem as the VM has to
be transient, but will be fatal for the active-commit code.
Oh good point. I'm hoping your series for copying a virStorageSource
will help out here (including however we end up resolving the issue at
tracking the security manager stuff).
--
Eric Blake eblake redhat com +1-919-301-3266
Libvirt virtualization library
http://libvirt.org