On 20/07/10 13:42, Dag Wieers wrote:
On Mon, 19 Jul 2010, nodata wrote:
On 19/07/10 18:52, Dag Wieers wrote:
During a discussion it was apparent some people expect the KVM guests to
be cleanly shut down when the host is being rebooted/shut down. Clearly
this is not done as part of the /etc/init.d/libvirtd sysv script stop.
So what likely is happening is that /etc/init/halt is sending TERM
signals to each remaining process, waits 5 seconds and then KILLs them.
Which would be fine if the kvm/qemu processes perform a shutdown on a
TERM signal _and_ they would in fact shut down cleanly within 5 seconds.
That's never going to happen. Five seconds is never enough.
My point exactly.
Isn't this worth opening a ticket about? Killing a guest is a bug
whatever way you look at it. The default behaviour should make sense and
it doesn't.
It's also a regression: the Xen replacement, KVM, should behave as Xen did.
_______________________________________________
rhelv5-list mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/rhelv5-list
