Data Warehouse Snapshot vs Backup

[alexp01482]

Hi

I'm looking at DR solutions for our dedicated SQL pool. Snapshots are taken 4-5 times a day and the automatic geo-redundant backup happens once per day.

What is the difference between the snapshots and the backup?

If the primary region becomes unavailable can I recover the DW in hte secondary region with a snapshot or do I need to use the backup?

Cheers

Alex

[v-ssriganesh]

Hi @alexp01482,

May I ask if you have resolved this issue? If so, please mark the helpful reply and accept it as the solution. This will be helpful for other community members who have similar problems to solve it faster.

Thank you.

[v-ssriganesh · Accepted solution]

Hello @alexp01482,
Thank you for reaching out to the Microsoft Fabric Forum Community. Below, I’ve clarified the differences between snapshots and geo-redundant backups and addressed your recovery scenario.

• Snapshots are point-in-time restore points taken 4–5 times daily and stored in the primary region. They’re used for recovery or copying within the same region (e.g: for data corruption or testing), with an 8-hour RPO. Snapshots are inaccessible if the primary region fails.
  
  
• Geo-redundant backups are full backups created once daily and stored in a secondary region. They’re designed for disaster recovery, allowing restoration in another region with a 24-hour RPO. They incur separate storage costs.
  
  
• You cannot recover in a secondary region using snapshots, as they’re stored in the primary region. You must use the geo-redundant backup, restored via the Azure portal, PowerShell, or REST APIs.

Use snapshots for primary region recovery and geo-redundant backups for cross-region DR. Create manual restore points before major changes to reduce data loss. For more details, see the Backup and Restore documentation.

If this information is helpful, please “Accept as solution” and give a "kudos" to assist other community members in resolving similar issues more efficiently.
Thank you.