Interrupted rebuild response
A rebuild reads surviving members and writes reconstructed stripes to a target. If another disk becomes unreadable, the controller resets, geometry is wrong, or power is interrupted, the array can contain a mixture of old and newly written state.
Understand the symptom
Controllers rebuild in an implementation-specific order. A target disk may contain valid reconstructed data up to one point and no useful data beyond it, while source members may include unreadable sectors or stale states. Restarting, forcing, or initializing can write new metadata and blur the boundary further. Offline reconstruction compares member generations, RAID metadata, file-system consistency, and known data patterns without committing changes to the originals.
What you may see
A source member may be repeatedly retrying unreadable areas or the controller may have encountered inconsistent metadata.
The array may exceed its redundancy while already writing reconstructed data to the target.
Wrong member selection, stale data, unreadable stripes, or prior configuration changes can produce a mountable but inconsistent volume.
The controller may no longer trust its RAID metadata; initialization can overwrite the state needed for analysis.
Protect the source
It modifies the target and may change array metadata, so another attempt is not a neutral diagnostic.
Parity checks and consistency operations may rewrite blocks based on an already incorrect member set.
Repairing above an incorrectly assembled array can convert stripe-level corruption into permanent logical changes.
Do this now
Prevent restart, auto-rebuild, consistency check, scrub, initialization, and file-system repair.
Record percentage, start time, failure time, member events, controller messages, and every disk replacement.
Preserve bay order, serial numbers, hot spares, removed disks, and the rebuild target.
Identify critical shares, databases, virtual machines, and dependency order before extraction begins.
NEPFOSYS process
Each drive is acquired separately, including members the controller considers failed, stale, spare, or partially rebuilt.
Metadata, generation differences, stripe consistency, and file-system evidence help identify the most coherent reconstruction.
RAID and file-system work proceeds on images, with priority workloads validated rather than trusting a simple mounted volume.
A clear next step
Start a case with the exact symptoms, incident timeline, prior attempts, encryption details, and the files that matter most. NEPFOSYS will route the case to the appropriate DFW intake or nationwide mail-in path.
Start with a case reviewQuestions, answered
Not when important data is at risk. A stalled rebuild may indicate an unreadable survivor or inconsistent state, and resuming continues to write. Preserve and evaluate the member set first.
Yes. It may contain partially reconstructed stripes that are newer or more readable than corresponding areas on another member.
A controller can complete writes using the wrong or degraded source set. Completion confirms the operation ended, not that every reconstructed stripe and file is correct.
DFW local · Nationwide mail-in