OpenVZ Forum


Home » General » Support » Server halting
Server halting [message #45247] Sun, 19 February 2012 08:20 Go to next message
Electron is currently offline  Electron
Messages: 7
Registered: November 2011
Junior Member
vzctl-lib-3.0.30.2-1.x86_64
vzkernel-2.6.32-042stab044.17.x86_64
vzkernel-2.6.32-042stab049.6.x86_64
vzquota-3.0.12-1.x86_64
vzkernel-2.6.32-042stab044.11.x86_64
vzctl-3.0.30.2-1.x86_64

Kernel - 2.6.32-042stab049.6

After random time of work server not work ( ping to server is good, but all containers and ssh on HN is not work ). After hard reboot I see at dmesg :

[    7.855423] BUG: Quota files for 6801 are broken: no quota engine running
[    7.855525] EXT4-fs (sda3): ext4_orphan_cleanup: deleting unreferenced inode 9830613
[    7.855530] BUG: Quota files for 6801 are broken: no quota engine running
[    7.855634] EXT4-fs (sda3): ext4_orphan_cleanup: deleting unreferenced inode 9830612
[    7.855638] BUG: Quota files for 6801 are broken: no quota engine running
[    7.855740] EXT4-fs (sda3): ext4_orphan_cleanup: deleting unreferenced inode 9830611
[    7.855745] BUG: Quota files for 6801 are broken: no quota engine running
[    7.855846] EXT4-fs (sda3): ext4_orphan_cleanup: deleting unreferenced inode 9830610
[    7.855855] BUG: Quota files for 6801 are broken: no quota engine running
[    7.855957] EXT4-fs (sda3): ext4_orphan_cleanup: deleting unreferenced inode 9830609

and many of :
[ 2363.775506] Fatal resource shortage: privvmpages, UB 6862.
[ 2363.775560] Fatal resource shortage: privvmpages, UB 6862.

before halting.

Overcommit at server normal :

CTID        LowMem  LowMem     RAM MemSwap MemSwap   Alloc   Alloc   Alloc
              util  commit    util    util  commit    util  commit   limit
Summary:      5.73  124.96   50.44   12.36  113.20   34.46  113.20  363.74


How can I fixed halt of server ?

Report message to a moderator

Re: Server halting [message #45277 is a reply to message #45247] Wed, 22 February 2012 02:47 Go to previous messageGo to next message
yasuyuki is currently offline  yasuyuki
Messages: 11
Registered: February 2012
Junior Member
I've seen your phenomenon.
and I also want the solution.

I can reproduce this problem when many 32 bit template containers are started/stopped at same time.
How about your environment?

Report message to a moderator

Re: Server halting [message #45294 is a reply to message #45277] Thu, 23 February 2012 07:18 Go to previous messageGo to next message
Electron is currently offline  Electron
Messages: 7
Registered: November 2011
Junior Member
At dmesg log after hard reboot I not see any messages about mass restarting containers at few seconds before halting.

"Fatal resource shortage" solved by add resources to container ( many failcnt on privvmpages )

"BUG: Quota files for XXXX are broken: no quota engine running" solved by restart quota ( vzctl stop XXXX, vzuota off XXXX, vzquota on XXXX, vzctl start XXXX )

Quota fixed by me after last halt.

At other HN with same environment ( kernel, OS version, version of vz packets, version of templates ) I not see any troubles.

Report message to a moderator

Re: Server halting [message #45315 is a reply to message #45247] Fri, 24 February 2012 09:33 Go to previous messageGo to next message
Electron is currently offline  Electron
Messages: 7
Registered: November 2011
Junior Member
I catch an error at KVM ( I can't insert [img] link and do attach )

If you know, how I can fix this, please help.
  • Attachment: ovz_fails.png
    (Size: 113.81KB, Downloaded 398 times)

Report message to a moderator

Re: Server halting [message #45318 is a reply to message #45247] Fri, 24 February 2012 11:39 Go to previous messageGo to next message
jmginer is currently offline  jmginer
Messages: 26
Registered: July 2008
Junior Member
Do you have containers that in the past was running with REHL 5 kernel and now are running in this REHL 6 ?

Maybe this bug -> bugzilla.openvz.org/show_bug.cgi?id=2188

Report message to a moderator

Re: Server halting [message #45319 is a reply to message #45318] Fri, 24 February 2012 12:29 Go to previous messageGo to next message
Electron is currently offline  Electron
Messages: 7
Registered: November 2011
Junior Member
Yes, we have containers at this HN , that migrated from RHEL5 node.
I understand, that this bug is not solved, but maybe anybody got temporary fix for this ?

Report message to a moderator

Re: Server halting [message #46009 is a reply to message #45247] Sun, 22 April 2012 14:56 Go to previous message
BeNe is currently offline  BeNe
Messages: 4
Registered: June 2009
Junior Member
We have the same problem here with Proxmox 2.0 and OpenVZ Containers.

Apr 22 05:00:20 miraculix kernel: EXT4-fs (dm-6): ext4_orphan_cleanup: deleting unreferenced inode 12059227
Apr 22 05:00:20 miraculix kernel: VZDQ: Tried to clean orphans on qmblk with 1 state
Apr 22 05:00:20 miraculix kernel: BUG: Quota files for 102 are broken: no quota engine running
Apr 22 05:00:20 miraculix kernel: EXT4-fs (dm-6): ext4_orphan_cleanup: deleting unreferenced inode 12058719
Apr 22 05:00:20 miraculix kernel: VZDQ: Tried to clean orphans on qmblk with 1 state
Apr 22 05:00:20 miraculix kernel: BUG: Quota files for 102 are broken: no quota engine running
Apr 22 05:00:20 miraculix kernel: EXT4-fs (dm-6): ext4_orphan_cleanup: deleting unreferenced inode 12058713
Apr 22 05:00:20 miraculix kernel: VZDQ: Tried to clean orphans on qmblk with 1 state
Apr 22 05:00:20 miraculix kernel: BUG: Quota files for 102 are broken: no quota engine running
Apr 22 05:00:20 miraculix kernel: EXT4-fs (dm-6): ext4_orphan_cleanup: deleting unreferenced inode 12192008
Apr 22 05:00:20 miraculix kernel: EXT4-fs (dm-6): 16 orphan inodes deleted
Apr 22 05:00:20 miraculix kernel: EXT4-fs (dm-6): recovery complete
Apr 22 05:00:20 miraculix kernel: EXT4-fs (dm-6): mounted filesystem with ordered data mode. Opts:

This happens during every backup of the openvz container with the snapshot mode.
Last week the whole node crashed. Shocked
There are no RHEL containers inside, never migrated from a RHEL node are something else.

Is this Bug related to the ext4 filesystem ?
If so, we would reinstall the whole node and use ext3 instead of ext4.

Thanks for any help!

Report message to a moderator

Previous Topic: centos + OpenVZ network issue
Next Topic: chromium in VE on gentoo host
Goto Forum:
  

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

Current Time: Sun Jul 14 13:52:24 GMT 2024

Total time taken to generate the page: 0.02328 seconds
Powered by FUDforum Powered by Virtuozzo Containers