The infamous 'VSS_WS_FAILED_AT_PREPARE_SNAPSHOT' problem..

[ap_]

Another way around this, although wouldn't recommend it on backups jobs for VMs with high I/O that are failing to create at preparing a snapshot/checkpoint; create one manually from Hyper-V Manager or SCVMM.
Once completed, delete the snapshot. All AVHDX files associated with the VMs will be merged but do check the storage location in case any are left. You should be able to delete them manually and run a backup job.

I haven't had a backup failure since.

Thanks for this information Bennon, very much appreciated! I couldn't find any orphaned AVHDX files (perhaps I was looking in the wrong place) but your solution of creating and deleting a checkpoint (followed by a restart) on the guest vm has fixed my issue.

[migmjt]

Case ID 03391638.
I had this problem on one VM and turned off integration services for a while. I then tried turning them back on after some successful backups but it failed again. I turned integration services off again and it started backing up again. Then another VM stopped backing up recently. I raised a case with Veeam and got a lot of advice about VSS being in stable state, restarting services, updating integration services, updating Veeam itself etc etc. Nothing helped.

I then tried making checkpoints in Hyper-V for both VMs and they started backing up again. I removed the checkpoints and they failed again.

In the end I had a look in the VM folders on the cluster storage and noticed there were some 5 month old Autorecovery.avhdx files sitting there. I removed all checkpoints and then renamed the avhdx files as .old. I did this with the VMs running to be sure they weren't required - I figured they'd be locked if they were in use so was fairly confident they were obsolete when I could rename them. (I deleted them once I was sure I didn't need them). The backups started working fine again.

I watched the backup process and saw that Veeam creates these files briefly for each VM then they disappear. I guess something went wrong one time and they got left behind (both VMs had the same timestamp on these files). Having the old files must get in the way of Veeam making new ones I suppose.

So if you get the 'VSS_WS_FAILED_AT_PREPARE_SNAPSHOT' problem when backing up with Veeam in Hyper-V, have a look in the VMs folders on you Hyper-V host / cluster storage and check you don't have any rogue checkpoint files. If you do try moving them out of the folder or renaming them and try a backup.

[dsapergia]

The cause of my encountering this issue was the fact that one of my Virtual Machine's VHDX files was in the root of its Cluster shared volume and not in a folder. ie c:\ClusterSharedVolumes\server1\serverdisk2.vhdx moving it to c:\clusterSharedVolumes\server1\vhdx\serverdisk2.vhdx resolved the issue. For some reason MS Servers can't properly work with these files in the root of CSV volumes, we first encountered the issue with Virtual Machine Manager.

[Bennon Maina]

If you're running Windows Server 2012 R2 DC/STD/ESS stand alone or WSFC hosts, make sure you install this hotfix if you haven't done so (although I suspect it has been included in rollup/security updates by now) https://support.microsoft.com/en-gb/help/3060678. It seems to be a known issue where snapshots are not deleted after you perform a backup operation by using VSS.

[ahodzic]

I ran into the same issue.
Ended up:
1. Deleting manually created snapshot
2. Restarting "Volume Shadow Copy" service on the VM

Now I have a successful backup!

[phoenixconstruct]

This might help someone:

This issue has been giving me headaches, VEEAM support was not helpful, just pointed it out as an MS problem.
Turns out it was the VSS volume size limit, reduced my volumes and bingo!

https://support.microsoft.com/en-us/hel ... in-windows

[ntottle]

Great news! Backups are working perfectly now!

I am now experiencing exactly the same issue.
I have Hyper-v 2012 R2 host with the VSS writer fails. I also have the .autorecovery.avhdx files.

Can I clarify, did you just delete the auto files to get this working again?? Thanks

I appreciate this post is 4 years old and you may not be able to remember. Thank you.

[wishr]

Hi Neil,

The easiest way to solve it would be to open a case with Microsoft support.

Thanks

[PetrM]

Hello,

I recommend to involve our support team as well to simplify communication with Microsoft support if they have some questions related to our product. Please upload logs and share a case ID with us.

Thanks!

[ntottle]

Thank you.

I have already logged a call with Veeam - Case # 05461009
We have established this is a Microsoft issue.

I will contact their support.

Thanks

[zlatan]

Hi,
Any updates from MS about this issue?
Thanks!

[PetrM]

Hi Zlatan and Welcome to Veeam R&D Forums!

Please open a case with our support team and share case ID over here. The error "VSS_WS_FAILED_AT_PREPARE_SNAPSHOT" is a generic one which can be thrown due to several different causes, it just indicates presence of an issue at the level of guest OS. Therefore, the solution which helped in one case might be entirely useless in your scenario. That's why I recommend to let our engineers to work on it. Also, you may collect debug logs and upload bundle to the support case. You can add as well Windows Event Logs (System, Application) from the affected VM to the bundle.

Thanks!