OpenVZ Forum


Home » General » Support » Another NFS-related issue (vzdump --suspend can't stop lockd)
Another NFS-related issue (vzdump --suspend can't stop lockd) [message #33274] Fri, 03 October 2008 09:19 Go to next message
fatbrother
Messages: 12
Registered: September 2008
Location: Novosibirsk, Russia
Junior Member
I have same setup as in my previous posting ( http://forum.openvz.org/index.php?t=tree&goto=33262& #msg_33262):
HN debian/FZA 2.6.18-12-fza-amd64 kernel, Debian Etch, vzctl 3.0.22-1dso1,
VE Centos 5.2, nfstools 1.0.9, has NFS mount.

I'm trying to do vzdump --suspend --exclude-path /home VE.
(BTW, do I ever need --exclude-path, won't it exclude NFS mount automatically?)

First sync goes OK, but on second sync I get:
INFO: suspend vps
Setting up checkpoint...
suspend...
Can not suspend VE: Resource temporarily unavailable
Error: foreign process 22159/21135(lockd) inside VE (e.g. vzctl enter or vzctl exec).
Error: suspend is impossible now.
Error: foreign process 22159/21135(lockd) inside VE (e.g. vzctl enter or vzctl exec).
Error: suspend is impossible now.
Error: foreign process 22159/21135(lockd) inside VE (e.g. vzctl enter or vzctl exec).
Error: suspend is impossible now.
Checkpointing failed

As far as I understand, that's NFS lockd, right?
What should I do? Not using lockd is unacceptable for several reasons. Umounting NFS renders VE unusable, so it'd be as good as vzdump --stop. If you suggest using vzdump --checkpoint, I have another issue for ya.

Report message to a moderator

Re: Another NFS-related issue (vzdump --suspend can't stop lockd) [message #33279 is a reply to message #33274] Fri, 03 October 2008 15:49 Go to previous messageGo to next message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Hi,

How lockd was started? Did you start it from inside the HN like "vzctl exec ..."?

Report message to a moderator

Re: Another NFS-related issue (vzdump --suspend can't stop lockd) [message #33288 is a reply to message #33279] Fri, 03 October 2008 17:40 Go to previous messageGo to next message
fatbrother
Messages: 12
Registered: September 2008
Location: Novosibirsk, Russia
Junior Member
maratrus wrote on Fri, 03 October 2008 22:49

Hi,

How lockd was started? Did you start it from inside the HN like "vzctl exec ..."?


No. I hit this problem, stopped the container, did the backup and started the container again. And run vzdump again to post correct text of error message to the forum. So all this was after a clean container restart.

Report message to a moderator

Re: Another NFS-related issue (vzdump --suspend can't stop lockd) [message #33300 is a reply to message #33288] Mon, 06 October 2008 14:26 Go to previous message
maratrus is currently offline  maratrus
Messages: 1495
Registered: August 2007
Location: Moscow
Senior Member
Hello,

could you please check that:

1. Using our 2.6.18 kernel from git repository you are able to reproduce that.
http://git.openvz.org/ (linux-2.6.18-openvz project)
or may be you can get/compile the kernel from here (http://download.openvz.org/kernel/branches/2.6.18/current/)
(may be this issue was already fixed)

2. Please make sure that "vzctl chkpnt $VEID" bumps into that error.

Then please file the bug in OpenVZ bugzilla http://bugzilla.openvz.org/.
Thank You very much for the assistance!

Report message to a moderator

Previous Topic: can not login webmail (plesk - horde)
Next Topic: Some questions
Goto Forum:
  

[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Mon Aug 12 13:52:54 GMT 2024

Total time taken to generate the page: 0.02821 seconds
Powered by FUDforum Powered by Virtuozzo Containers