Your RAID Array Failed: Can the Data Be Recovered?
Dieser Artikel ist bisher nur auf Englisch verfügbar.
RAID failure recovery isn't one problem — it depends heavily on which drive(s) failed and what's been done to the array since.
First: don't guess at a rebuild
Rebuilding a RAID array with drives in the wrong order, or with the wrong array parameters guessed from memory, is one of the most common ways to turn a recoverable failure into an unrecoverable one. If you're not certain of the exact original configuration, don't attempt a rebuild from scratch.
Single drive failure: usually fine, act promptly
If your array is designed to tolerate one drive failing (RAID 1, RAID 5, RAID 6, RAID 10 — see our RAID levels comparison) and exactly one has failed, the data is intact. The priority is replacing the failed drive and rebuilding before a second failure happens — which is a real risk, since rebuilding puts unusually high read load on the remaining drives, right when you can least afford another failure.
Two or more drives failed (beyond what the array tolerates)
This is a genuinely harder recovery. It usually means:
- Imaging each surviving drive individually first — never working directly against the only remaining copies.
- Reconstructing the array's logical structure (stripe size, drive order, parity algorithm) from what's recoverable, since this metadata is sometimes what's actually damaged rather than the data itself.
- Extracting files from the reconstructed virtual array once the structure is rebuilt correctly.
This is meaningfully more specialized than single-drive recovery and is where DIY attempts most often make things worse — guessing at stripe size or drive order during reconstruction can produce garbled output that looks like corruption but is actually a reconstruction error.
Common causes worth knowing about
| Cause | What happened |
|---|---|
| Controller failure (not drive failure) | The RAID controller card itself died — drives may be completely fine and readable once moved to compatible hardware |
| Multiple drive failures from the same batch | Drives bought together often have similar wear, so failing close together isn't a coincidence |
| Failed rebuild attempt | Wrong order/parameters used during a DIY rebuild attempt |
| Accidental array deletion/reconfiguration | Reconfiguring a NAS or RAID controller without realizing it would wipe the array definition |
If the RAID controller itself failed rather than the drives, the drives may hold completely intact data — the fix in that specific case can be as simple as getting compatible replacement hardware, not data recovery at all.
Dealing with a failed RAID array right now? Stop before attempting a rebuild if you're not certain of the original configuration — get in touch first.
Häufige Fragen
One drive in my RAID 5 failed — is my data safe?
Yes, for now — RAID 5 tolerates exactly one drive failing. Replace the failed drive and let the array rebuild as soon as reasonably possible, since a second failure during that window would cause data loss.
Can data be recovered if two drives fail in a RAID 5 array?
Sometimes, but it's a specialized recovery, not a standard one — it typically requires imaging the surviving drives and reconstructing data through the array's parity math manually, which is meaningfully harder than a single-drive failure.
I rebuilt the array and now data is missing — what happened?
A rebuild attempted with drives in the wrong order, an incorrect array configuration, or a rebuild started with an already-marginal second drive are the most common causes — this is a common way a recoverable situation turns into a harder one.