RAID failure recovery

RAID failure recovery starts with the state you have now.

Start with the condition you can see now. ADR separates controller and server presentation problems from the condition of the member data, then selects a recovery path that does not depend on bringing the failed array online.

Call a RAID recovery technician Start emergency triage

If possible, photograph the controller screen and bays, label every member, and record the server, controller, RAID level, drive count, capacities, messages, and actions already attempted.

Failure-state firstWork from what the system shows now

Members identifiedPreserve bay order and history

Sources protectedImage unstable media before reconstruction

Output separatedRecover to different destination storage

Choose the current symptom

The missing array does not need to be repaired before ADR can begin.

Each path explains what the condition may mean, what evidence to preserve, what actions can change the state, and how ADR evaluates recovery away from an uncertain controller.

RAID 5 failureOne or two members failed, the array is offline, or a rebuild stoppedPreserve every original and replacement member before another parity operation.Open RAID 5 recovery → RAID 6 failureMultiple members failed or the virtual disk disappearedDual parity does not make another rebuild safe when the surviving set is uncertain.Open RAID 6 recovery → Degraded and unstableMembers are dropping, timing out, or changing statusThe surviving set may not be healthy enough to support a rebuild.Review the unstable state → Multiple-member failureMore than one drive is failed, missing, or unreadableThe controller's fault threshold is not a recoverability verdict.Review multiple-member failure → Failure during rebuildAnother drive failed while the rebuild was runningStop before the controller reads or writes through additional uncertainty.Protect the original and rebuild members → Rebuild failureA rebuild stalled, failed, or made the array disappearKeep both the original member and the partial rebuild target.Review the failed rebuild → Configuration lostThe controller reports foreign, missing, or unconfigured storageDo not replace the original metadata with a guessed configuration.Review configuration loss → Drives visible, virtual disk missingThe virtual disk disappeared after a reboot or power eventThe logical volume can be reconstructed without recreating it on the controller.Review the missing virtual disk → Server storage missingThe operating system cannot see or mount the RAID volumeThe failed server is not required as the recovery environment.Review the server failure → SQL Server affectedThe database is in Recovery Pending after a storage eventProtect the RAID and database files as separate recovery layers before repair.Review Recovery Pending → VMware outageThe datastore is missing and the virtual machines will not startPreserve the underlying storage before rescans or repair attempts alter it.Review the missing datastore → RAID 50 failureOne lower stripe group is missing and the array is offlineRecoverability depends on the members within that group, not the surviving group alone.Review the missing RAID 50 group →

Separate presentation from data

A controller message tells you what it will do—not everything the members contain.

ADR reviews member identity, readability, metadata generation, order, offset, stripe geometry, and parity away from the failed array. Candidate reconstructions are validated before recovered files are written to separate storage.

See ADR's recovery method Open controller and system triage

Get started now

Tell ADR what changed and what the system shows.

The server does not need to boot, the volume does not need to mount, and the array does not need another write before an ADR technician reviews the next safe step.

Call 1-800-228-8800 Start emergency triage Review RAID recovery services →