Locked VM's

[Stampede]

Hi All,

I'm having a reoccuring issue with 2 or 3 of my VM's. The issue is they get locked in vSphere and prevent Veeam from backing them up. Now I can do a backup if I shut down the VM and run Veeam as that lifts the lock. The issue is, doing a shutdown every time I need to do a backup isn't practical in our business environment.

The VM's are primarily SQL databases, and I have enabled quiescence but it still doesn't let Veeam in. I've been searching every resource I can find but I still haven't found a solution. Has anyone found a good work around for this issue?

Thanks,
Peter

[foggy]

Peter, do you see any specific messages in the job session log? Have you contacted support already (as it is advised to do with any technical issue)?

[Vitaliy S.]

Are you able to create VM snapshot for these VMs while they are running?

[Stampede]

Hi foggy, Vitaliy,

The reason in the log is: "Error: File does not exist or locked. VMFS path: [[ESX5_IS01] QCDB/QCDB.vmx]. Please, try to download specified file using connection to the ESX server where the VM registered. Failed to create NFC download stream. NFC path: "

I havn't opened a support ticket for this issue yet.

I am going to take the snapshot but I will wait until this evening as this is a critical system that is being actively used. So once it's free of any users I will create the snapshot, and then remove it. Actually would the best practice in this case be to merge the back to the original snapshot, or to just delete the old new snapshot? I would be able to just delete it, so as not to interrupt VM activity by merging.

Thanks,
Peter

[Vitaliy S.]

Hi Peter,

There are existing topics with a similar issue, opening a support case would be the right decision if you want to make it work > Error: File does not exist or locked.

Actually would the best practice in this case be to merge the back to the original snapshot, or to just delete the old new snapshot? I would be able to just delete it, so as not to interrupt VM activity by merging.

Just click delete all/consolidate to merge data from the snapshot to a virtual disk of the VM. Based on the error message above, you don't need to do that, as vmx lock issues is not related to the snapshot operation.

Thanks!

[Stampede]

Ok, interesting journey to my solution. Hopefully this helps anyone who experiences something similar.
So I searched through all the help docs and tried everything. I submitted a ticket with Veeam and VMware and ran through the suggestions both offered.
Snapshot's worked, after returning the VM's to their original host which was holding the lock (as determined by the MAC address) I was able to download the VMX file locally, but VEEAM still kept erroring out.
I even restarted the host, and then created a generic VM removed the data stores from the old and attached it to the new. Then I made sure to remove the old VM from inventory. The VM ran great, but Veeam still failed due to locked VMDK/VMX.

There was a great tech support specialist, Antony, on Veeam Phone Support that saved the day. He realized the version of Veeam was outdated and a patch to v8 fixed the issue. What is frustrating is I downloaded the latest version of Veeam from the website maybe 2 months ago, so I'm not sure why I would have an ancient version of the software. Literally first release of the software, old. Luckily I got an update patch which brought my version up to speed and then allowed all backups to process normally.
So if your having this same type of lock problem, check your Veeam version first! This is a known error on the early v8 releases. My current version of Backup & Replication v8 is 8.0.0.2084.

Here is the update patch link, http://www.veeam.com/kb2068


Thanks everyone for all your help!
Peter

[Vitaliy S.]

Hi Peter,

Thanks for taking time to post your solution. One tip to avoid these situations in future - Anton is sending Veeam Community Digest email every weekend with all information about patches, updates and other important news, so it's worth looking it through.

P.S. I will pass your appreciation to the support engineer.

[Stampede]

Thanks Vitaliy,

Be sure to thank Bryan for me as well!
Thank you for the tip, I didn't know that.


Peter

[lando_uk]

It's annoying that the base install image of VBR is always the first release, you then have to run the updates. It would be nice when an major update comes that that the base image is also updated, like VMware/Windows etc offers.

[Vitaliy S.]

Hi Mark,

Thanks for your feedback! While this is definitely very useful to have the ISO image always updated, but releasing a new ISO (aka maintenance release) is a significant overhead for our R&D that we prefer to avoid, unless absolutely necessary. However, we do understand potential downsides of this approach and are thinking on how to address this in future releases.

Thank you for your understanding.

[lando_uk]

Something in the initial setup that connects to your servers to see if there are any pending updates then downloads and installs, so new users are always on the latest build.

[Vitaliy S.]

Product auto-update feature! Yes, this should be a good way our for this situation.