Two drives failed or went offline
The controller has exceeded RAID 5 fault tolerance, but a failed label does not prove both members are physically unreadable.
Review this condition →RAID 5 data recovery
RAID 5 data recovery after one or two drive failures.
Stop writes and preserve every member. ADR can identify the usable drive set, protect unstable sources, determine the array layout, and reconstruct the data without asking the failed controller to rebuild it again.
The server does not need to boot, the volume does not need to mount, and the RAID 5 does not need to be brought online before ADR can begin.
Members firstIdentify every source and replacement
Read-only accessProtect unstable drives before reconstruction
Layout verifiedOrder, offset, stripe, and parity checked
Output separatedRecovered data goes to other storage
Start with the symptom
What is the RAID 5 doing now?
The number of failed labels is only part of the story. What the controller changed, which members remain readable, and what happened during any rebuild determine the safest next step.
One drive failed and the array is offline
RAID 5 should normally remain available after one member drops. An offline volume means another condition must be identified before rebuilding.
Review this condition →The rebuild failed or stalled
Keep the original member and the partial rebuild target. Both may contribute to recovery.
Review this condition →Another drive failed during the rebuild
Stop the rebuild and preserve the original member, replacement member, and every surviving source before the failure state changes again.
Review this condition →The virtual disk disappeared
Do not recreate or initialize it. Preserve the member order and controller state before testing the layout.
Review this condition →The configuration is foreign or missing
Controller metadata can be evaluated without importing, clearing, or replacing it on the original drives.
Review this condition →SQL Server or a database is affected
Protect the RAID and database files as separate layers before attach, repair, or log rebuilding.
Review this condition →The recovery path
ADR does not use the hardware rebuild as the recovery method.
The failed server and controller are evidence, not prerequisites. If the drives can be connected independently to a controlled recovery system, ADR can begin while replacement infrastructure is prepared.
- 1
Record the state
Capture the controller, server, RAID level, bay order, drive identities, errors, and actions already attempted.
- 2
Protect the members
Access sources read-only and image drives whose stability or coverage is uncertain.
- 3
Reconstruct virtually
Determine member order, data offset, stripe size, and parity rotation away from the failed controller.
- 4
Recover and verify
Validate files, databases, virtual machines, or other priority data and write the result to separate storage.
Historical client account
Two members were reported failed.
“We lost two drives out of three in a RAID 5. ADR recovered our data and kept the process professional from start to finish. I recommend ADR to anyone facing the same position.”Bryan Neely · January 2012 · Historical client account
We can get started now
Tell ADR what failed and what has already been tried.
Begin with the controller or NAS model, RAID level if known, drive count and capacities, current messages, rebuild history, and the business data needed first.