On Fri, Sep 13, 2013 at 06:40:43PM +1000, Andrew Lau wrote:
Hi,
I recently upgraded to the oVirt Nightly repo because I needed some patches
not yet in 3.3, but it appears to have broken the live migration.
Both of the nodes have the storage (glusterfs replicate), oVirt gives the
error:
Migration failed due to Error: Fatal error during migration (VM: gl01,
Source: HV02, Destination: HV01).
It seems to successfully transfer but then the receiving vdsm node gives
the following error:
Thread-58935::ERROR::2013-09-13
18:36:52,004::vm::2062::vm.Vm::(_startUnderlyingVm)
vmId=`3867341c-b465-4285-9c35-0c5972527808`::The vm start process failed
if ret is None:raise libvirtError('virDomainLookupByUUIDString()
failed', conn=self)
I would need to see much more logs from source and destination nodes in
order to understand why the VM failed to reach to destination. Would you
provide them?
Dan.