Azure BLOB storage - limitations?

[rkelley]

Greetings,

I am trying to understand the limitations of using Azure BLOB storage with Veeam. In my case, I have an on-prem Veeam-10 server with a local (NFS) repository and a scale-out repository (Azure BLOB). I have created a backup job with a couple of VMs pointing to the scale-out repository - backups are working as expected. In Azure, I have another Veeam-10 server running and have added the Azure BLOB storage as a backup repository.

What I have noticed using Azure BLOB storage:

(1) The list of backups (in the Object Storage disk listing) on the Veeam server running in the cloud does not get updated when we run backups on-prem. We must disconnect the BLOB storage device then click "import backup" to get the storage repository refreshed with most recent backups.

(2) We are unable to use recovery points when restoring VMs to Azure stored the BLOB storage. When recovering VMs to Azure, we are not given a choice to pick any recover points.

Are these two limitations expected using Veeam with Azure BLOB storage? Or, have I misconfigured something?

Thanks.

[Gostev]

Hello!

Why do you have another Veeam server in Azure in the first place? This is your main "misconfiguration". You don't need to have one to restore VMs to Azure from backups stored in the blob storage - you should just do this from your on-prem install.

First issue is definitely because of that. Your cloud Veeam install does not "own" the object storage repository in question, which is why it is not "seeing" any recent backups there until you explicitly import them (which should only be done in a DR scenario, when your on-prem data center is lost).

I'm not too sure what is causing the second issue, but it is unexpected for sure. Assuming you do actually have multiple restore points in the blob storage, you should be able to see and pick from them after you import backups.

Thanks!

[rkelley]

Hello Gostev.

Reason for having another Veeam server: If the DC link fails, our on-prem Veeam server becomes unavailable and we can't recover to the cloud. We were told by the Veeam professional services team to have a second copy running in Azure just in case the DC site fails. If we switch from BLOB storage to Azure CIFS storage, we just hit the RESCAN button on the CIFS repository and Veeam automatically updates the list of backups as expected. This seems to be a limitation of backing up directly to BLOB storage.

As for the second issue: it seems Veeam will only allow us to recover from the most recent backup. This is confirmed by both the on-prem and Azure installs of Veeam-10.

[Gostev]

OK, this makes sense then. It is totally fine to keep a copy of Veeam in the cloud for DR scenario. Think about Import Backup wizard for a standalone object storage repository (not a part of SOBR) as the RESCAN button of a sort - no real difference there. In fact, if you create SOBR for your cloud Veeam server like the one you have in your on-prem Veeam server, then RESCAN button on SOBR will also be updating the list of backups automatically.

As for the second issue, your should investigate this with support.

[rkelley]

Thanks again. Unfortunately, hitting the RESCAN button on SOBR does not automatically update the list of backups. We have to detach the SOBR then re-import to get the list updated. Seems like a bug to me. We will work with support on this issue as well as the second one.