VADP backups

Hi Karim

Please check the schedules configured for the VM clients. It could be the  problem with schedules not configured properly.

Thanks & Regards

Panwar

Panwar - Schedules are correct and other VMs in the group are getting current backup levels.  FULL on days when a FULL should run and INCR on days when INCR should run.

Bobby - I will check the daemon.raw file.  The VMs that are running a FULL everyday have had a successful FULL backup previously.

I was researching this issue and I came across a VMware article.  It talks about Change Block Tracking (CBT) and making sure it is enabled on the VMs.  I have not had a chance to check that as I have to shutdown the VM to look at its configuration parameters in vCenter.  Once I check if CBT is enabled or disabled, I will post results.

Thanks.

Hi Karim,

Check the daemon.raw on the server also to see if it tells you why it's doing a full - for example it may tell you "No full backups of this save set were found in the media database".

-Bobby

I found a VM that I could play with.  I have a Windows 2003 VM and it runs a FULL VADP backup every time, even if it already had a succesful FULL backup.  Following the VMware KB article - 1031873, I enabled CBT on the 2 virtual drives.  Rebooted the VM and kicked off a backup again.  It ran a FULL again.   The strange thing...When I look at the Save Seesion Details, it shows the Level as FULL.  When I look at the Group Detail, the level is shown as Inc.  See attached pics.

hi jumbo

what your mminfo says? are the schedules set on client and group?

Hello Jumbo.

Im experiencing the same issues you're having. This is when I try to backup a Linux virtual machine with CBT enabled.

(I dont have any Windows VMs to play with). Group Details says level incr and Save Session details says Full.

My NetWorker Server is RedHat Enterprise, 7.6.2SP2, my VADP proxy is a Windows2008R2 machine.

/Martin

Incr backups are only supported for Windows VMs. IF a windowVM is only doing FULL backups then you need to check if there are files in VM containing ASCII charctors.

From Networker VADP integration guide:

Note:

supported only for Windows virtual machines, and all attached disks must be NTFS file

systems.

Note also that CBT-based incremental backups are always file based. Image level recovery from

a CBT-based incremental backup is not supported.

I did read about the Unix / Linux VMs and how they do a FULL everytime based on the non-NTFS image which is OK.  This made me think about the Windows VMs that are getting a FULL everytime.  I looked at the savegroup completetion email that I get from the system and observed something interesting....  The test VM that I am working on, along with other production Windows VMs that are getting a FULL backup show the image as non-NTFS.  Very strange.  I checked the VM and the drives are NTFS partitioned.  Cannot explain why this is.  See below:

  nj06nt015:*FULL* savegrp: suppressed 63 lines of output.

* nj06nt015:*FULL*

* nj06nt015:*FULL*  Reading 7676Total Number of Virtual Disks read is 2.

* nj06nt015:*FULL* Completed save operation of the non-NTFS Image.

* nj06nt015:*FULL* Connecting directories saved successfully.

* nj06nt015:*FULL* Index entries commit session ended successfully.

* nj06nt015:*FULL* Total number of files backed up: 3.

* nj06nt015:*FULL* Total backup size: 393375 MB.

* nj06nt015:*FULL* Completion time: 1/12/2012 4:56:43 AM

* nj06nt015:*FULL* Total time taken: 4 hr(s) 24 min(s) 41 sec(s)

* nj06nt015:*FULL* Backup of the non-NTFS volume Success.

* nj06nt015:*FULL* Cleaningup VADP mount operation.

* nj06nt015:*FULL* Deleting snapshot for 'snapshot-1106' ...

* nj06nt015:*FULL*  Task is -1% complete

* nj06nt015:*FULL*  Task is 95% complete

* nj06nt015:*FULL*  Task is -1% complete

* nj06nt015:*FULL* Snapshot for VM nj06nt015 is deleted.

* nj06nt015:*FULL* Temporary directory 'F:\mnt\nj06nt015' for VADP deleted.

Disk Management screenshot from the server mentioned in the post...

Could you provide a screenshot of the Disk Management of the VM.

Update....

All the Windows VMs that are getting a FULL only and a combination of FULL and INCR, show as non-NTFS volumes when I check the savegroup completion emails.  All 3 below are Windows VMs with NTFS partitions:

* dplswkey1b:*FULL* Completed save operation of the non-NTFS Image.

* nj06nt015:*FULL* Completed save operation of the non-NTFS Image.

* nj06nt018:*FULL* Completed save operation of the non-NTFS Image.

If we go by the screenshot you have provided. VMs having this disk configuration should always do image level backups and file level recovery should not be possible.For more information you can run "savegrp -D9 "group name" >out.txt 2<&1"and check debug output file. Please also provide the file here.

Vermad - I am not sure what you mean by "VMs having this disk configuration should always do image level backups and file level recovery should not be possible".  I ran the savegrp command you suggested with the debug level of 9.  This time when I ran the backup job, instead of doing a FULL just like in the past 10 times, the job did not run succesfully.  The test group the client is in, is configured to do a incremental which I set in the client override properties but still it tried to do a FULL backup. Attached is the debug file for your review.

The debug output also says that the networker VMDK parser failed for non-NTFS volume80432:nsrvadp_save: Parser failed because of a non-NTFS volume inside the VM.
* nj06nt018:*FULL* Backup of the non-NTFS volume Success.
I have confirmed that if you have a non-NTFS volume in a VM, file level recoveries are not possible because networker VMDK parser is designed to only parse NTFS volumes.

I am not sure why the group has failed. adding -D9 in backup command should generate debug for the backup process. Please provide zipped nsr\tmp\sg\ folder for failure analysis.