|
|
|
|
|
|
|
|
|
|
xen-users
Re: [Xen-users] Debian Etch, Xen 3.2-1 Error: Device 0 (vif) could not b
I think problem is not 'vif4'. Look closely to xl list (xm list). If you
see some '-d-' or '-s-' domains - congratulation, you won the main xen
pain-in-the-ass problem: 'zombile' domains, which can not be destroyed
until dom0 application will not return shared memory pages to Xen.
On 31.10.2011 05:48, Jan Bakuwel wrote:
Hi,
> From time to time this problem pops up. The server runs happily. VMs are
started and restarted a number of times. Then after a while a VM (as far
as I know always a HVM) no longer boots. A reboot of the server "fixes"
the problem but I'm looking for a better way.
I've noticed that subsequent attempts to start this particular (HVM)
domain will start building up entries in xenstore. I've already tried to
remove them with xenstore-rm, which doesn't make a difference.
It seems that in this case "vif4.0" is getting stuck. When the VM is
attempting to boot, xentop shows a (very brief) error message at the top
saying that vif4.0 exist but the domain doesn't. vif4.0 doesn't show in
ifconfig but it is listed in /sys/class/net.
My gut feeling tells me that if I would be able to get rid of vif4.0,
I'd be able to start the VM again.
Any ideas how I can get this unstuck so I can start this VM without
rebooting dom0 and all other domUs?
regards,
Jan
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
|
|
|
|