Azure Storage Mover's Azure Files-to-Azure Files copy capability is in preview, with recurring transfers between shares across accounts, subscriptions, and regions in the same tenant. It offers another way to prepare a secondary dataset. Microsoft's preview guide.
The engineering opportunity is useful, but the design question comes first: what will the application actually be able to recover, and how will the team prove it?
Check the Preview Boundary First
The guide specifies SMB, SSD provisioned v2 shares, subscription-level preview enablement, and Azure portal-only configuration for this scenario. NFS and CLI/PowerShell support are not included. Copy scope is the whole share, not a selected subdirectory. Preview prerequisites and limits.
My recommendation is to record the actual source and destination configuration before proposing a pilot. Include the workload owner, tenant, region pair, expected data volume, identity dependencies, and the business purpose of the secondary copy.
Do not turn a preview announcement into a promise that every existing share can participate unchanged. Resolve eligibility and the intended operating model before building a delivery date around it.
A Timetable Is Not a Recovery Result
Microsoft explicitly provides no RPO guarantee: the documented recovery point is the last successful run's completion time, not the configured interval. A scheduled run fails if the previous run is still active. Scheduling limits.
For a pilot, I would record the latest successful completion, unresolved failures, and whether a new run is making progress. Agree in advance when an overdue or failed run requires an operator to investigate.
Exercise the copy during representative change activity, not only against a quiet sample. Keep the same test dataset and validation criteria when comparing schedules so that a shorter interval does not disguise a less complete result.
Ask the application owner to define acceptable data currency and recovery time separately. Neither should be inferred from a job's name or a green configuration screen.
Decide Who Can Write to the Destination
The guide says to treat the target as read-only between runs because there is no conflict resolution. Its mirror mode deletes destination items absent from the source. Target and copy-mode behavior.
I would document the chosen mode as part of the workload design and review it with the data owner. Use disposable test data to establish what happens when a file is changed, renamed, or removed.
Keep recovery exercises from accidentally becoming a second production writing location. Before a drill, name the operator, the permitted actions, and the process for returning the environment to its normal state.
Do not treat a synchronized copy as proof of historical retention. Ask a separate question about how the team would recover an earlier good version after an unwanted application change.
Validate Through the Application
My proposed acceptance check would compare representative files, directory structure, permissions, and timestamps, then test access using the intended recovery identity.
Include a small file, a large file, a restricted directory, and a workload-specific example that exercises the application's real expectations. Record discrepancies rather than accepting a plausible folder count.
Then rehearse the application handoff in an approved test environment. Confirm how clients find the destination, how operators judge that it is usable, and who authorizes the switch. Copy validation and application recovery should have distinct evidence.
Practical Cloud Engineer Takeaway
Confirm preview eligibility before committing to a pilot.
Define business recovery requirements before choosing a timetable.
Monitor completed, validated outcomes rather than configuration alone.
Review destination write ownership and deletion behavior explicitly.
Rehearse application access and recovery without risking the primary workload.
Who Should Care?
Azure storage administrators, application platform teams, and continuity planners evaluating a second location for Azure Files data.
Bottom Line
The preview creates a useful copy-and-sync option. Its value depends on a complete recovery design: a suitable destination, visible job outcomes, controlled writes, and an application-level rehearsal.
These are proposed evaluation practices. No storage job, subscription, or file share was created or changed for this article.
Sources
Stay radical, stay curious, and keep pushing the boundaries of what is possible in the cloud.
Chriz
Beyond Cloud with Chriz
Comments