[Users] Live storage migration fails on CentOS 6.4 + ovirt3.3 cluster

[Sander Grendelman]

On Wed, Nov 6, 2013 at 5:41 PM, Martijn Grendelman
<martijn.grendelman at isaac.nl> wrote:
>
> So is qemu-kvm-rhev from mentioned source RPM a drop-in replacement for
> qemu-kvm from CentOS ? Would it make sense to install it instead of
> qemu-kvm?

That's the way it worked for me.

- Put one host in maintenance
- Install *-rhev packas on that host
      yum localinstall qemu-*rhev*.rpm)
  this replaces the regular packages:
      Nov 06 16:15:38 Installed: 2:qemu-img-rhev-0.12.1.2-2.355.el6.9.x86_64
      Nov 06 16:15:40 Installed: 2:qemu-kvm-rhev-0.12.1.2-2.355.el6.9.x86_64
      Nov 06 16:15:40 Installed:
2:qemu-kvm-rhev-tools-0.12.1.2-2.355.el6.9.x86_64
      Nov 06 16:15:40 Erased: qemu-kvm-tools
      Nov 06 16:15:41 Erased: qemu-kvm
      Nov 06 16:15:41 Erased: qemu-img
- Restart vdsm, enable host migrate test VM to host, test live storage
migration: OK
- Put host back in maintenance and reboot.
- Activate host and repeat maintenance/installation steps with other hosts.

The "solution" for the live snaphost problem as proposed in
https://bugzilla.redhat.com/show_bug.cgi?id=1009100 is:

     |BZ| Vdsm should check (hopefully via libvirt) if the underlying
qemu supports live snapshot, and report this feature to Engine.
     |BZ| If it does not, block this feature in UI.

That is something that should be implemented, however,  there should
also be some way to use a full-featured kvm with oVirt.

Can oVirt carry the qemu-*rhev* packages?
Or is this something that should be in epel? Or CentOS?