automatic vzdump on one machine always fails

[kastnermario]

Aug 04 03:01:10 INFO: Starting Backup of VM 101 (openvz)
Aug 04 03:01:10 INFO: CTID 101 exist mounted running
Aug 04 03:01:10 INFO: status = running
Aug 04 03:01:10 INFO: mode failure - unable to dump into snapshot (use option --dumpdir)
Aug 04 03:01:10 INFO: trying 'suspend' mode instead
Aug 04 03:01:10 INFO: backup mode: suspend
Aug 04 03:01:10 INFO: ionice priority: 7
Aug 04 03:01:10 INFO: starting first sync /var/lib/vz/private/101/ to /mnt/backup//dump/vzdump-openvz-101-2012_08_04-03_01_10.tmp
Aug 04 03:03:53 INFO: Number of files: 55978
Aug 04 03:03:53 INFO: Number of files transferred: 44132
Aug 04 03:03:53 INFO: Total file size: 4710312517 bytes
Aug 04 03:03:53 INFO: Total transferred file size: 4710083817 bytes
Aug 04 03:03:53 INFO: Literal data: 4710084260 bytes
Aug 04 03:03:53 INFO: Matched data: 0 bytes
Aug 04 03:03:53 INFO: File list size: 1396237
Aug 04 03:03:53 INFO: File list generation time: 0.001 seconds
Aug 04 03:03:53 INFO: File list transfer time: 0.000 seconds
Aug 04 03:03:53 INFO: Total bytes sent: 4713967527
Aug 04 03:03:53 INFO: Total bytes received: 878381
Aug 04 03:03:53 INFO: sent 4713967527 bytes received 878381 bytes 28836978.03 bytes/sec
Aug 04 03:03:53 INFO: total size is 4710312517 speedup is 1.00
Aug 04 03:03:53 INFO: first sync finished (163 seconds)
Aug 04 03:03:53 INFO: suspend vm
Aug 04 03:03:53 INFO: Setting up checkpoint...
Aug 04 03:03:53 INFO: suspend...
Aug 04 03:03:53 INFO: Can not suspend container: Device or resource busy
Aug 04 03:03:53 INFO: Error: task 4630/1516(ntpd) uses posix timers
Aug 04 03:03:53 INFO: Checkpointing failed
Aug 04 03:03:56 ERROR: Backup of VM 101 failed - command 'vzctl --skiplock chkpnt 101 --suspend' failed: exit code 16

I have ~20 container but only this one always fails. Manual backup works fine. Automatic backup fails.

What can/should I do?

Thank you.

 

[tom]

your run ntp inside your container? disable it and try again.

 

[kastnermario]

your run ntp inside your container? disable it and try again.

Why should I disable NTP? NTP runs in every (!) container. Is this a well known Problem? But why does this only effects this one machine and not the other machines too? This seems not to be a logic conclusion. Any other hint?

Thx

btw: Disabling NTP in this container does not solve the issue.

 

[tom]

you should run ntp only once. most run it on the host and never in a container. some run it in one container only.

 

[tom]

"mode failure - unable to dump into snapshot"

find out why snapshot backup is not working. does it work for other containers?

analyse your LVM configuration.

 

[kastnermario]

"mode failure - unable to dump into snapshot"

find out why snapshot backup is not working. does it work for other containers?

analyse your LVM configuration.

Hi tom,

yes it works well for any other container and vm (20 Container and 37 KVM). Just this one is not backing up. Every of this machines is hosted on the same environment so this could not be a side effect by a misconfigured LVM. This will affect other containers and vm's too.

Any other hint?

Thx