22.12.2017 16:43, Kevin Wolf wrote:
Am 12.12.2017 um 17:04 hat Vladimir Sementsov-Ogievskiy geschrieben:
The test creates two vms (vm_a, vm_b), create dirty bitmap in
the first one, do several writes to corresponding device and
then migrate vm_a to vm_b with dirty bitmaps.
For now, only migration through shared storage for persistent
bitmaps is available, so it is tested here. Only offline variant
is tested for now (a kind of suspend-resume), as it is needed
to test that this case is successfully fixed by recent patch.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsement...@virtuozzo.com>
Reviewed-by: John Snow <js...@redhat.com>
[...]
+ self.vm_b.launch()
+ self.vm_b.event_wait("RESUME", timeout=10.0)
Wasn't this racy because the RESUME event may be sent before we connect
to the socket?
Didn't know (and didn't faced) it. If it is so, it should be changed
like you've changed similar place in my previous iotest.
Kevin
--
Best regards,
Vladimir
