I've come across this problem with a Windows 2008 R2 server with VMware Data Recovery a few months ago that fit the situation as described here:
http://www.blueshiftblog.com/?p=473
However, I came across a problem today where the environment is at ESX Update 1 and the Windows 2008 R2 server was deployed as a fresh install. When reviewing the disk.EnableUUID variable, it's already set to true but when a backup is executed, the process would reach 98% then fail with message:
"Cannot create a quiesced snapshot because the create snapshot operation exceeded the time limit for holding off I/O in the frozen virtual machine."
What's strange is that I went ahead to change the disk.EnableUUID variable to false and I was then able to backup the VM. Does anyone know why this is the case? Thanks.
VMware Data Recovery version: 1.2.0.37
ESXi version 4.1.0 build 260247
On 2k8 R2, setting the disk.EnableUUID to "FALSE" will essentially force "file system" quiescing as opposed to "application" quiescing. What applications are running inside the VM? SQL? Exchange? Is there a heavy amount of I/O going on at the time of the backup? What happens if you try and create a manual snapshot through the VI Client? (checking quiescing and unchecking memory)
Hi,
I should have included more information for you:
Server Operating System: WIndows Server 2008 R2 64-bit
Server role: VMware vCenter
Installed application: SQL Server 2008 STD (for vCenter & VUM)
vCenter version: 4 with Update 1
When disk.UUID is set to true, executing a snapshot from VI Client as you suggested fails exactly the same as executing VMware Backup Recovery job with the same message:
"Cannot create a quiesced snapshot because the create snapshot operation exceeded the time limit for holding off I/O in the frozen virtual machine."
If disk.UUID needs to be true to enable application quiescing then I suspect there is most likely a problem with one of the VSS services so maybe it's worth to investigate each and every VSS service to determine what might be the problem?
The VMware Snapshot Provider is the problem. Reviewing and implementing
the fixes in the current version of the following VMware KB does not
resolve the issue. They may fix it soon when they see this.
Cannot create a quiesced snapshot because the snapshot operation
exceeded the time limit for holding off I/O in the frozen virtual
machine
Use this method if you are getting the following error in the system
event log when trying to start the VMware Snapshot Provider Service in
the guest VM:
Error 1053: The service did not respond to the start or control request in a timely fashion VMWare Snapshot Provider
Also, you will probably be getting the following error when trying to
take a snapshot of the VM with Memory unchecked and Quiesce checked or
when trying to do a VDR backup:
Cannot create a quiesced snapshot because the snapshot operation
exceeded the time limit for holding off I/O in the frozen virtual
machine.
To fix the problem Uninstall and Reinstall VMWare Snapshot Provider as follows:
1. Uninstall Volume Shadow Copy Service (using Modify from VMware Tools setup)
2. Delete VSS folder from C:\Program Files\VMWare\VMWare Tools\Drivers
3. Delete VMWare Snapshot Provider Service by deleting the key
HKLM\System\Current Control Set\Services\vmvss (Just delete the vmvss
key)
4. Delete VMware Snapshot Provider
{564d7761-7265-2056-5353-2050726f7669} from HKLM\system\current control
set\services\VSS\Providers
(If you do not have this key, just omit this step.)
5. Reboot
6. Start the two COM+ services mentioned in kb.vmware.com/kb/1018194
7. Install Volume Shadow Copy Service (using Modify from VMWare Tools setup.exe)
8. Reboot
VMWare Snapshot Provider should now be Manual, and you can start it.You
should now be able to take a snapshot and back up with vdr while the VM
is powered on.
( Also see this Symantec article with a full step by step walkthrough of steps 1-4: )
Marcus
IT Wheel
Consider awarding points for "helpful" and/or "correct" answers.
I've got exactly the same problem, setting the UUID statement to false allows me to backup with BEWS 2010.
Has anyone tried the above process? Its hard for me to get a reboot organised of a production VM.
This fix works, Trust me.