Time to restore from backup data stored in Azure Blob after VBR server corruption

[Asahi]

Hi,

Backup of VMs to Azure Blob using VBR.
There are about 300 VMs and the backup data size is about 50TB.

In case the VBR server is damaged and unavailable, we would install Veeam on a new Windows machine and do a configuration restore.
https://helpcenter.veeam.com/docs/backu ... ml?ver=120

Can I restore from Azure Blob data immediately after a successful configuration restore?
Or do I have to wait for synchronization of metadata information, etc.?

If a repository rescan operation is performed and there is a wait, how long will it take?
https://helpcenter.veeam.com/docs/backu ... ml?ver=120

Kind Regards,
Asahi,

[veremin]

One clarifying question: how are you using Azure Blob Storage? As a standalone object storage repository or as a capacity extent of a Scale-Out Backup Repository?

[Asahi]

Hi Veremin,

Thank you for reply.

Use as standalone object storage, not SOBR.

Kind　Regards,
Asahi

[veremin]

When restoring the configuration in "Restore" mode, the repositories will be rescanned automatically.

However, even if you choose the "Migrate" mode for configuration restoration that doesn't imply automatic rescan, we recommend running a rescan afterward, as there are no guarantees that the data in the repository hasn’t been modified since the configuration was backed up.

Thanks!

[Asahi]

Hi Veremin,

Thank you for reply.

What I would like to know is the approximate time it takes to rescan.

Kind Regrds,
Asahi

[veremin]

Honestly, we cannot say or predict how long the rescan procedure will take, as it varies from one infrastructure to another, and there are a whole bunch of parameters affecting it: whether immutability is enabled or disabled, the number of checkpoints created (if immutability is enabled), the total number of restore points, the number of machines present in them, the presence or absence of encryption, bandwidth, latency, etc.

Probably the best option would be to simulate a DR scenario, temporarily stop writing to the object storage repository on the first backup server, connect it to a new backup server, and rescan it (or restore the configuration) to get an idea of the time required.

Thanks!