VCB...Failed to export the disk errors

[Jwoods]

Every now and then I'll get the error Failed to export the disk: The requested resource is in use[/b] after performing a VCB backup and it's usually after the second VMDK has been backed up. It goes completely through the backup and at the point of removing the snapshot errors out. Has anyone run across this error? Have any solutions? Or is this a bug???

[kkuphal]

I'm seeing the same thing on one of our servers. It happens to be Exchange where we have disabled the LGTO_Sync driver as recommended in the KB. A normal file level VCB works fine but the FullVM job fails with this error. Have you made any progress on this yourself?

[Jwoods]

I have found errors in the logs and shipped them off to VMware support. I've also found a post from GBromage here[/url], but other than that, not much more. It's more consistent when I run single backups--one after another. However, the error still displays. But it's more frequent and consistent with 2 or more backups running simultaneously.

[kkuphal]

Interesting note.

I've been in the process of upgrading the ESX servers to 3.0.1 and finished yesterday and installed the VCB 3.0.1 software and the updated BackupExec integration scripts and tested the failing job I mentioned.

Today it's working. Could be something that is fixed in 3.0.1 of VCB

[Jwoods]

Unfortunately, that's the version I've been working with. Esx 301 and VCB 301 w/ VC 201p1

[kkuphal]

Any more progress here? My thought of 3.01 was also mistaken as I'm now seeing these again.

\[2007-03-26 08:01:21.060 'BaseLibs' 3804 warning] \[Vmdb_Unset] Unsetting unknown path: /vmomi/

Converting "H:\mnt\us09exm01.maclean-fogg.com-fullVM\scsi0-0-0-US09EXM01.vmdk" (compact file):

0%=====================50%=====================100%

**************************************************

\[2007-03-26 08:09:55.784 'BlockList' 5440 warning] Could not log out (us09rpl01.maclean-fogg.com:905:Administrator)

\[2007-03-26 08:09:55.784 'vcbMounter' 5440 error] Error: Failed to export the disk: The requested resource is in use

\[2007-03-26 08:09:55.784 'vcbMounter' 5440 error] An error occured, cleaning up...

failed to prepare us09exm01.maclean-fogg.com for backup, PrepareForBackup() returned error 1

External command failed. See error above.

Exit Code: 1

It will fail repeated but then also start working again with a later retry of the backup job without any changes on our part. Very frustrating.

[admin]

http://kb.vmware.com/kb/5096672 ?

[kkuphal]

Nope. The machines will back up successfully one attempt but not another. The job log I just posted is running right now successfully after a retry without any changes on my part to the infrastructure, job, machine, etc. It just seems to randomly fail (and then many VMs will fail to back up) and then later work fine.

[admin]

Ah well was worth an ask.

Are you backing up multiple VMs simultaneously when the failure occurs, or just the one?

[kkuphal]

I would say generally multiple because this normally occurs during the backup cycle. I think we have 4 maximum to run at a time (really our minimum because we backup to disk using 4 partitions). I have long suspected that this might be related to multiple jobs but I have turned it down about as far as I can go given our environment

[Jwoods]

I have eliminated multiple simultaneously running jobs and have gone to one job at a time and things have been running consistently well for the past 3 weeks. Unfortunately, this takes forever.

During testing though, whenever a job either creates a snapshot or is in the process of removing a snapshot and another job is performing one of those tasks, that "resource is in use" error appears.

Interestingly, I know multiple snapshots can be committed/deleted simultaneously during non-VCB related tasks, but for some reason, VCB doesn't respond well to this.