Unable to delete Microsoft.DataReplication replicationVaults – force delete fails – backend cleanup required

Backend cleanup of Microsoft.DataReplication/replicationVaults is not covered in the provided documentation and cannot be performed directly. However, there are a few documented checks that should be completed before engaging Microsoft support for backend removal:

1. Verify and remove any locks
   • Resource group deletion requires that all resource locks and backup data are removed first.
   • Check the macewanu-prd-uw2-azmigrate resource group and each replication vault for management locks. If any exist, remove them and retry deletion of the vaults and then the resource group.
2. Confirm there are no dependent resources
   • Resource Manager cannot delete a resource group if resources are still dependent on each other across groups.
   • Ensure that no other resources (for example, networks, subnets, or other managed resources) reference these replication vaults. If there are dependencies, remove or update them first.
3. Retry deletion at the resource group level
   • Once locks and dependencies are cleared, attempt to delete the entire macewanu-prd-uw2-azmigrate resource group again. Resource Manager will determine the deletion order and retry failed deletes for transient errors.
4. When deletion still fails
   • If the replication vaults remain in Succeeded state and block resource group deletion after all locks and dependencies are removed and delete operations have been retried, this indicates a platform-side issue.
   • At this point, open a support request with Microsoft Azure Support, include:
     • Subscription ID and resource group name
     • Full resource IDs of the three replicationVaults
     • CorrelationId from failed delete attempts
     • Timestamps and regions (westus2, centralus)
     • Confirmation that locks and dependencies have been removed and that delete operations with the latest API version (including force=true) have been attempted.
   • Request backend cleanup of the orphaned Microsoft.DataReplication/replicationVaults so that the resource group and provider registration can be unblocked.

These steps align with the documented behavior for deletion failures and resource group deletion ordering; when those are exhausted, only Microsoft Support can perform backend cleanup.

References:

• Azure Resource Manager resource group and resource deletion
• Microsoft.DataReplication replicationVaults 2024-09-01 (deployment-language-arm-template)