Account for disks added in Azure before Site Recovery failback

Source facts

During modernized Site Recovery reprotection to on-premises, only disks previously replicated from the source are copied back. Disks newly attached to the failed-over Azure VM are not included. The on-premises VM, when present, is shut down during reprotection for consistency. After failback, protection toward Azure must be enabled again; returning the workload and restoring its protection direction are separate stages. Microsoft Learn.

Applicability

Review workloads that changed their disk layout while operating in Azure after an on-premises outage. Compare the original protected disk inventory with the current Azure VM rather than relying on its name or the apparent completeness of its application.

DSE recommendation

Build a return manifest that assigns an owner and destination to every newly attached disk and its data. Resolve the separate transfer and application-consistency plan before approving failback. Keep the original replicated disks visibly distinct from cloud-added storage, and agree which application checks must pass before the Azure-side recovery work is considered complete.

Verification

During an approved rehearsal, reconcile source-era disks, current Azure attachments, and the expected on-premises result. Verify that the separate data-return plan covers every new disk, including application dependencies that refer to it. After the workload returns, check replication toward Azure independently. Do not accept an accessible on-premises VM as evidence that all cloud-era additions were transferred or protection was restored.

Official references

Microsoft Learn: About failover and failback in Azure Site Recovery – Modernized.