Need Official Process for Stopping Checkpoint Removal without risk of corruption

[mark95677]

Hello,

We have an issue where we rebooted our Veeam Backup Server (Windows) after patching and it corrupted something so that Checkpoints removal process failed and left orphaned checkpoints within our Wasabi cloud storage.

I recently learned that this Checkpoint Removal process runs 24/7, starts at 3am, and does not provide an option to stop it for system maintenance.

We have the following Veeam Support case:
Veeam Support - Case # 07802077 - Checkpoint Removal Failure on Wasabi Repository

We were informed by the engineer that I needed to submit a Feature request on this Forum.

My request is the following:
There needs to be an approved and documented process/procedure so that we can service the backup server without causing a corruption of checkpoints on Wasabi storage to occur. Please work to come up with an ‘Official’ process or utility to help stop/pause the Checkpoint Removal process so that the server can be rebooted when needed or product updated.

I personally consider this a defect and not a feature request, but I figured I would do what I can to get answers.

Thank you for your time.

[KirillChuxlancev]

Hello,

Kirill from Object Storage QA team here. I'll try to clarify the mechanism of the Background Checkpoint Removal job.

The session itself is indeed running on the background constantly, but if no Checkpoint Removal tasks are assigned to it, the session is simply idling. You may observe the session log in the Shell Console to find out if any active Checkpoint Removal task in being processed at the moment.

In fact, the mechanism is very similar to the previous approach, when Checkpoint Removal was executed on Backup Job post processing. With the new approach, these tasks are executed after the Backup Job run within the scope of a separate session (Background Checkpoint Removal), allowing backup jobs to finish, effectively distributing the load.

Considering all the above, I can confirm it is safe to service the backup server if no Checkpoint Removal tasks are being executed by the Background Checkpoint Removal job.

Hope that helps!

[veremin]

Additionally, I’d like to note that even though we don’t recommend rebooting the backup server while it’s performing background checkpoint deletion, we are confident that such operations should not lead to backup chain corruption—the checkpoint deletion process should resume from where it left off after the reboot.

If you’re still certain that the backup chain was corrupted specifically due to the backup server reboot and the background checkpoint deletion, we’ll double-check the debug logs to verify whether these claims are justified. In any case, we plan to work with your support engineer to investigate the issue in detail.

Thanks!