[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] Stuck blkback


I still haven't seen a single error but this issue is occurring about once every 3 weeks or so on one of our hypervisors. 

This started happening just this year.  We are running Debian Squeeze with all the latest packages as of about 2 months ago.  I added a second core to our dom0's and set it to float.  They have also been completely patched this year for the time jump fix.  A week or two ago I pinned the second core, so if it has something to do with multiple dom0 cores the only thing I have left is to remote the second core and reboot everything.  We have also started using pvgrub more heavily although I would think the guest shouldn't be able to prevent the dom0 from destroying the blkback device/process.

Is anybody else seeing/seen this issue?


On Sat, Apr 13, 2013 at 4:56 AM, Tom Fritz <tfritz@xxxxxxxxx> wrote:

We have had a few occasions in the past few months where a VPS will need a hard reboot.  After the
command "xm shutdown" the VM would no longer boot.  I have traced this to the device mapper device
having an open count of 1, and then to the fact that there is a blkback process running for a domain
that doesn't exist.  I'm going to go out on a limb and suggest the blkback device was for the VM in
question.  I can't tell if this is a symptom or cause(the trouble with the blkback process) however I
don't know of any way to close out that backend process without a reboot so it renders the LV
unusable:/  Has anybody witnessed this behavior?

        Yes I have seen this a number of times om our production machines. It will render the machine unable to start/restart a VM. I have not been able to find a solution to the issue. If you figure something out I would be very interested in hearing about it.


Xen-users mailing list

Xen-users mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.