Restore from Capacity Tier with Failed Performance Tier

[EricinIT]

I've got a DR planning scenario I'd like some input on. Suppose there's a scale-out backup repository with a Windows direct attached storage performance tier and capacity tier set to tier data after 30 days to object. The backup jobs have been configured with GFS to keep enough restore points that some fall into capacity tier. Then for some reason, the site with the performance tier fails that it is totally offline for many days. The backups in capacity tier are still fine and should be able to be restored.

The reason for asking this is I just tried restoring from capacity tier in this type of scenario. In the Veeam console, I went to Backups > Capacity Tier > found my restore point and did the restore process. I got a message the gateway wasn't available. This is true since the capacity tier gateway was the server with the performance tier at the site that failed. So, I checked the gateway on the object storage and tried to change it from the server hosting the direct attached storage. Veeam would show "This operation is taking longer than expected" message and not finish. I let it sit for 30 minutes and finally cancelled it. I checked the object storage properties again and it didn't change. So, it seems the only way to change the gateway is to have the gateway server up for the capacity tier storage.

Does anyone have input on a different way to do this or recommendations?

[HannesK]

Hello,

Veeam would show "This operation is taking longer than expected" message and not finish. I let it sit for 30 minutes and finally cancelled it.

So, it seems the only way to change the gateway is to have the gateway server up for the capacity tier storage.

both sounds more like a bug than a feature... I recommend checking with support because that's not how the software is supposed to work. Please post the case number for reference (and install the latest patch from last week, assuming that you use V12)

Best regards,
Hannes

[EricinIT]

Hi Hannes,

Thanks for pointing out the latest patch v12.0.0.1420 P20230412. I just updated and tried it again. I was able to change the Connection Mode selection on the object repository to an online server and click Next without it hanging. I also had to adjust the Mount server in the settings. The Edit Object Storage Repository settings let me click all the way through to save it.

I then tried the restore but got "Failed to create backup mount source. Failed to start an agent on one of the extents." error. I created a new temporary empty performance tier on an available server and added that into the SOBR as another performance tier. However, I still received this error.

So, I edited the SOBR again. This time removing the original failed performance tier. I had to click Yes to "Remove the extent without evacuating its backups?" message and click through the settings to Save the SOBR without that failed performance tier in it. Now, only the new temporary empty performance tier along with the capacity tier object storage was in the SOBR. After all this, I tried the restore from Capacity Tier again and was able to restore data. It does work but it isn't straight forward on the process should the site with the original performance tier is lost.

Eric

[Gostev]

@EricinIT it's very straight forward actually: you just install Veeam in a DR site and import backups. There's no need to create SOBR like you did.

[EricinIT]

@Gostev The import backups option from object storage isn't available if the object repo is part of a SOBR according to the limitations at https://helpcenter.veeam.com/docs/backu ... ml?ver=120. I just confirmed the option doesn't show in the menu on my server when it's in the SOBR.

[Gostev]

@EricinIT that's exactly why I said that you don't need to create SOBR in your DR site's Veeam install in the first place. You should proceed directly to importing backups into that DR backup server, without creating a SOBR. If you want a simple and straight forward process, that is

Now, if you did not actually lose the entire site with your production backup server and all, but only your Performance Tier storage - then it is even easier: you just need to initiate the restore process from the Backups > Object storage (copy) node of the management tree, in which case the restoration will be performed directly from the Capacity Tier storage, ignoring the fact that the Performance Tier is unavailable.

[EricinIT]

@Gostev Ok, so not importing the config (so SOBR config won't be there) and just add in the object storage repository would likely work as well to restore it.

[Gostev]

Correct.

[towen00]

...Now, if you did not actually lose the entire site with your production backup server and all, but only your Performance Tier storage - then it is even easier: you just need to initiate the restore process from the Backups > Object storage (copy) node of the management tree, in which case the restoration will be performed directly from the Capacity Tier storage, ignoring the fact that the Performance Tier is unavailable.

Hello,

I tried this, but it don't work for me. I have this popup : "Failed to create backup mount source. Failed to start an agent on one of the extend". and then "Faild to mount backup".

Capacity Tier is an Amazon S3 storage.

Any Idea ?

Thanks

[Gostev]

These errors would indicate some backup infrastructure issue, please work with Support to troubleshoot. Perhaps a host specified as the Mount server is down.