An Azure backup service that provides built-in management at scale.
Hi Jordi Pera & thx for join me at Q&A platform,
seems like an Azure portal request-construction issue rather than a MARS backup failure. Your backups are completing successfully, recovery points exist, and the same protected items are accessible through az rest and az backup item list. That strongly suggests the Recovery Services backend is working. One detail stands out in the InvalidBatchRequestUrl error. The URI you've posted appears to go directly from /subscriptions/{subscriptionId} to /providers/Microsoft.RecoveryServices/, without the /resourceGroups/{resourceGroupName} segment. If that's the actual request generated by the portal rather than a shortened version you've posted, it could explain why the ARM batch endpoint rejects it.
I'd compare the exact URL from the failed browser request with the working REST request, particularly the resource group segment and encoding of the protection container and protected item identifiers.
I wouldn't reinstall the MARS agents, re-register the protected machines, or recreate the vault. There's no evidence that backup configuration or recovery data is damaged. At this point, I'd escalate this to the Azure Backup portal team with a sanitized HAR trace, the failing batch request, the equivalent working REST request, the vault resource ID, and the exact UTC timestamp. Microsoft needs to investigate why the portal generates a request that the ARM batch endpoint rejects.
You can continue checking backup items and recovery points through Azure CLI while the portal issue is investigated. The blank details page alone doesn't indicate that your backups are failing.
rgds,
Alex