Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

Using Builtin VMware Agent Backup Policy to Backup VMs causing non snapshot removal and orphaned CTK files

Created: 30 Apr 2013 | 3 comments

Hi

 

We have an issue with Backup Exec 2010 R3 using vmware agent backup on a built in full/differential policy whereby the following happens:

 

  • Backup takes a snapshot 
  • Backup completes
  • Backup does not remove the snapshot taken 
  • CTK files remain after manual snapshot removal

The problem is that it is causing come VMDKs to be inaccessible by the virtual machine espcially SQL VMs. They list in device manager but come up as inaccessible "Device not Ready" This shows even before snapshot removal and is down to the CTK files 

Changing the vmx file to ctkenabled:false for all drives and booting the VM fixes the issue but we have to do this after every backup from BExec.

Is there are reason why this is happening? Its seems Changed Block Tracking is not working and need to understand if it is vmware at fault here or Backup Exec

Operating Systems:

Comments 3 CommentsJump to latest comment

pkh's picture

What version of ESX are you using?  I hope it is not ESX 5.1 because it is not supported.

Have you seen this blog?

https://www-secure.symantec.com/connect/blogs/remo...

Colin Weaver's picture

OK I am going to be editing this technote shortly to add a second registry key so if you end up seeing a copy that only mentions one registry key to create/amend then also consider adding the extra one I mention below this link

http://www.symantec.com/docs/TECH200709

HKEY_LOCAL_MACHINE\SOFTWARE\Symantec\Backup Exec For Windows\Backup Exec\Engine\VMware Agent\RemoveSnapshotRetryCount

This is also a DWORD and set this to non-zero to retry snapshot removal (would suggest maybe setting it to 2 initially)

 

Please note there are some circumstances when the reason for not being able to remove a snapshot is outside of our control and you may have to work with VMware to isolate the cause. This is because we send a snapshot removal request via the VMware programming interface. Once the request has been received what happens after that is within the VMware code.