Highlighted
Contributor
Contributor

Problems with ESXi 4.1 and snapshopts from Backup

Hi all,

i have a problem with our backup Software. We tested the software VEEAM and now the license expired and we ran into an error.

The customer didn´t order the license and the trial period exeeded now.

The Software uses the snapshot technology to successfully back up the servers.

After the backup, the snapshot will be deleted and the new files are committed into the vmdk file.

But this is only theory 😞 the snapshots won´t be deleted anymore.

Whe i browse the datastore i can see 8 snapshots but in the snapshot manager from the virtual machine there is no one listed.

I tried to manually create a snapshot. Then i can see the following szenario in snapshot manager:

Computername

|

- Consolidate Helper - 0

   |

    -manually created Snapshot

     |

     - you ar here

When i try to delete all snapshots i only recieve an error:

Unable to access file <unspecified filename> since it is locked

Please help me.

Thanks in advance,

Frank

0 Kudos
5 Replies
Highlighted
User Moderator
User Moderator

Hi Frank,

Welcome to the forums.

You can try the solutions indicated in this kb:

Good luck.

Regards

Franck

Highlighted
Contributor
Contributor

Hi Franck,

thanks for your quick answer. I checked most of the hints from this KB article.

The last points i am going to check tonight.

- power off the virtual machine and retry committing the snaphots

- Use the vCenter converter to create a clone from the machine.

The Problem is that the vm is an Exchange Server with a huge DB with about 100 GB.

This job will last long using the converter.

If anybody knows an easier way of committing the snapshots, please let me know.

Thanks,

Frank

0 Kudos
Highlighted
Virtuoso
Virtuoso

I seem to recall that the consolidated helper is a hidden snapshot that you should not see. Since the Veeam product simply calls the VMware API to create/commit snapshots, the root cause may not be a Veeam issue. I would check to see if perhaps the VMDK is currently tied to the backup appliance, as that would cause a snapshot commit to fail. Part of the Veeam backup process is to take a VM snapshot, then "steal" the VMDK over to their backup server and back up the data, then it "returns" the VMDK back to the VM and commits the snapshot.

You might also check with Veeam. Their forums are open to non-purchasing customers, and many of the employees are quick to respond to help people.

VCDX #104 (DCV, NV) ஃ WahlNetwork.com ஃ @ChrisWahl ஃ Author, Networking for VMware Administrators
0 Kudos
Highlighted
Contributor
Contributor

Hi Franck,

thanks for your advice.

I tested all points in that Knowledge Base article.

I powered off the machine, created manually a snapshot and tried to delete all... but that didn´t solve my problem.

The error that a file is locked didn´t appear but the snapshots were still in the datastore but not in the snapshot manager...

So i left the machine powered off and cloned the complete machine into another datastore

While cloning the machine, the snapshot were committed back into the .vmdk file.

It took some time but now i solved that issue and i am happy.

Now i am able again to expand the disks and so on.

Thanks for your quick advice.

Best regards,

Frank

0 Kudos
Highlighted
Contributor
Contributor

Hi,

i have solved that issue.

When the snapshot will stay in the Manager and you cannot commit it to the vmdk you have to restart the management services.

Sometimes a prozess is still locking the file.

You can restart the service on the console or via ssl.

Via SSL you can run "/sbin/services.sh restart" to restart the services.

After you have done this you are able to delete all snapshots and commit them into your vmdk File

On ESX Server i usesd this command via ssh

"service mgmt-vmware restart" to restart the managemet services.

Sometimes you also have to restart the vCenter Agent with the following command via ssh

"service vmware-vpxa restart"

Best regard,

Frank


0 Kudos