VM is locked ????????

[schick]

I can't seem to do anything with the Virtual Machine Control Panel.
It appears my attempts to reboot, or force a shutdown are failing because the VM is locked. What the heck does that mean?

I wanted to reboot becasue I couldn't login in via SSH. Using numerous clients, I found that after being prompted for a password, and entering my password, nothing happens. It just sits there. WTF?

It seems like things around here are a little bit screwy lately.

[dorogoy]

Exactly the same to me on VM1.

[nsuan]

I'm on VM1 and have the same problem, almost everything froze (It kept passing packets), tried to reboot and now the VM is locked.

[delta]

Same here, Xen domain 'delta' on vm1. I guess we might need to submit a ticket for this issue.

[schick]

What the hell is going on around here lately?

[matta]

Yes... it is a definite problem, has happenned before. It looks like the Xend daemon is not locking, but not allocating network devices. If xend is stopped it does not start back up and requires a host reboot. So far it is only happenning on vm1, it might be due to it using software versus hardware RAID or just is waiting to happen on other hosts. We've actually been trying to purchase a support contract for Xen from XenSource in event of problems such as this... for now we will see what we can get from the Xen developers.

[dorogoy]

If case it helps:

I don't know if the problem was already there but everything started badly for me after I tried to do a snapshot. The SSH connection was OK. After trying the snapshot nothing happened at all (no locked VPS) and the system told me that the snapshot was made (a snapshot of more than a GB in less than a second??? Wow!!!). From there I tried to reboot because I suspected that something was wrong and then, the locked state. The prompt in SSH failed a few seconds later.

I wonder if this problems is related to corrupted data on the disc. The corrupted data doesn't allow to perform a snapshot and everything goes upside down. That is probably what happened before as well.

[matta]

I've thought that also, which is why I made the snapshot do a fsck before any actions... unfortunately this is not easily reproduced so it is difficult to fix without another crash. I will make some changes and add some debugging logs and see what we can find.

[schick]

I am able to login via secure shell now. It seems that everything is up again.
What ever you have done to remedy this, has apparenty worked.

Thanks.

[dorogoy]

I did a snapshot and everything went smoothly now ;)

[dorogoy]

Matt, I was just wondering that but ... Don't you think that, if we are talking about a problem related to snapshots, the main problem here is TAR? Maybe if we use for snapshots a low level tool like AFIO:


Afio makes cpio-format archives. It deals somewhat gracefully with input data corruption, supports multi-volume archives during interactive operation, and can make compressed archives that are much safer than compressed tar or cpio archives. Afio is best used as an `archive engine' in a backup script.


or even DAR manage pretty well data corruption.