Unusual .vbk File Size?

[name4521]

Hello, everyone!

We're backing up a single virtual machine that is 336GB in size in a single Veeam backup job. Currently there are six restore points. The .vbk and .vrb files in the file system look like this:

Code: Select all

04/09/2009  12:57 AM     1,339,457,151 file2009-04-08T165439.vrb
04/10/2009  11:32 AM     3,697,467,795 file2009-04-08T213623.vrb
04/11/2009  10:33 AM     4,441,551,646 file2009-04-09T221135.vrb
04/12/2009  03:58 PM     1,500,336,760 file2009-04-10T221129.vrb
04/13/2009  01:39 PM     3,884,064,977 file2009-04-11T221121.vrb
04/13/2009  01:39 PM 1,177,070,543,547 file.vbk

Is it unusual for the .vbk file to be several times larger than the virtual machine it is backing up?

[toreagan]

I have the same issue with a few VM's. One VM in particular is only 116 GB, and the .VBK file grows to around 410GB. I am also wondering how come the backups are so much larger than the vm size?? I'm using version 3.0.1.

~Trent

[Gostev]

Hello, this may happen due to a few various scenarios:
- you had another VM in this job before that got removed after
- the job was stopped a few times during actual VM backup (manually or due to some issue that made backup terminate abnormally)

We will need to investigate the logs to find out the exact reason, but the 2 reasons above are most common. Under normal circumstances, this should not be happening.

[toreagan]

Thanks Gostev, the jobs did terminate abnormally a few times due to the backup storage location running out of disk space. So if the job does terminate abnormally its recommended to delete all backups under this job and start it from scratch again? to keep the vbk file size manageable?

Thanks,
~Trent

[Gostev]

Yes, with the current version it would be best to do this.

[name4521]

Yeah, there were a few times that the jobs terminated abnormally here too. We have a mix of the following two results:

Code: Select all

17 of 17 files processed 

Total VM size: 336.00 GB
Processed size: 336.00 GB
Avg. performance rate: 20 MB/s
Backup mode: VCB SAN

Start time: 4/9/2009 10:11:35 PM
End time: 4/10/2009 3:04:12 AM

Completing current object backup process
VCB error:  Not attempting to renew disk lease for snapshot-5991[Bulk Datastore 2] w2k3r2.fs05/w2k3r2.fs05_2.vmdk anymore (have tried 6 times already). Too many lost disk leases.  The managed object of type "vim.host.DiskManager.Lease" identified by the MoRef "session[B892A17C-B809-404E-A009-1142F38AA34F]C48037BA-9059-4F54-A2DA-307ED0871FC2" could not be found. Error: Failed to export the disk: I/O Operation failed An error occurred, cleaning up...

Code: Select all

17 of 17 files processed 

Total VM size: 336.00 GB
Processed size: 336.00 GB
Avg. performance rate: 14 MB/s
Backup mode: VCB SAN

Start time: 4/10/2009 10:11:29 PM
End time: 4/11/2009 5:15:07 AM

Completing current object backup process
VCB error:  No active paths found. Error: Failed to export the disk: The requested resource is in use An error occurred, cleaning up...

[ssimakov]

name4521, all these errors indicate issues with your storage subsystem and yes, frequent failures like these may lead to growing VBKs.

[name4521]

We currently have VirtualCenter installed on a VM. I suspect that the host it's on is overworked, so we're moving it to a more powerful host in an effort to minimize the disk lease and the other VCB error.

We have a VM out there with twelve attached .vmdk files. Checking the logs, it takes about 50 minutes to release the snapshot when the backup finishes 10 hours after the snapshot was created. I'm not sure how big the deltas are getting, but could it cause the .vbk size problem when Veeam times out waiting for the snapshot cleanup process to complete?

[Gostev]

Since it happens in the end of backup job when VBK population is over, this specific issue should not results in backup file grow. To my understanding, only stopping/terminating currently running backup process (while it interacts with VBK) may result in this.