Data Recovery Case File · NAS & Network Storage · Degraded for a Year

"Degraded, Unmounted": a RAID 5 that spent a year without its safety net — the owner's own analysis confirmed, every member imaged first, and the array reassembled around the freshest consistent set on the TaskForce 2

This enquiry read like a well-kept incident log. An eight-bay QNAP, RAID 5: "Drive 3 was marked as failed approximately one year ago and has not been participating in the array since. Drive 6 failed a couple of weeks ago. The array is now showing 'Degraded, Unmounted.' Drives 1, 2 and 5 appear normal; drive 4 shows a yellow warning but is still accessible." And his own reasoning, which deserves confirming up front because it's exactly right: "as drive 3 has been offline for around a year, I believe its data is likely outdated — my primary objective is to recover or rebuild using drive 6, the most recent failure." Correct on both counts. This page decodes what his year of degradation actually meant — an array running with zero redundancy, one failure from exactly this outcome — why stale drive 3 cannot rejoin the story, and the discipline that reassembled the array: every member imaged first, reconstruction on the copies, in-place rebuilds refused.

The decode: the year without a net, the stale member, and the right order of work

What a degraded year really meant: RAID 5's arithmetic is unforgiving and worth stating plainly. The array stores one drive's worth of parity spread across the members, and that parity buys survival of exactly one failure — no more. The day drive 3 dropped out, that one allowance was spent: the array kept serving files, as RAID 5 does, but from that moment it was a tightrope walker whose net had been quietly removed. Every day of the following year, the data's survival depended on no other drive failing — seven aging disks, spinning in the same box, sharing the same heat and hours, and any one of them a single point of total failure. Drive 6's failure a year later wasn't misfortune arriving; it was probability catching up. The doctrine his year illustrates: a degraded array is an emergency wearing a working face — the moment redundancy is spent, replacement and rebuild become the most urgent task the system has, not a chore for someday.

Why drive 3 is out, and drive 6 is everything — his analysis, confirmed: his reasoning holds precisely. A RAID member that left the array a year ago holds a year-old snapshot of its stripe data — coherent with nothing that's been written since, and worse than useless in a reconstruction, since mixing stale blocks into a rebuild corrupts the result silently. Drive 3 is history, not help. The reconstruction therefore rests on the current generation: healthy drives 1, 2 and 5; warning-flagged drive 4, which must be treated as fragile; and drive 6, whose recent failure means its platters hold up-to-date stripe data behind whatever fault stopped it — making its recovery the keystone. RAID 5's parity offers one further grace: for any stripe where drive 6 stays unreadable, the remaining current members can compute the missing blocks — provided, and only provided, they themselves are read completely and consistently. Which dictates the order of everything.

The discipline — images first, rebuilds refused: the single most dangerous move now would be the obvious one: replacing drives and letting the NAS attempt an in-place rebuild. A rebuild is the most punishing workload an array ever runs — every sector of every member read end-to-end — aimed here at a fragile, warning-flagged drive 4 and a set of survivors of unknown condition, with any mid-rebuild failure compounding the damage on the originals. The professional order inverts it entirely: image every member individually first, write-blocked, each drive handled per its own condition — then perform the reassembly virtually, on the copies, where nothing can be lost twice.

On the bench

Every member went onto the Atola TaskForce 2 individually — its parallel ports imaging the healthy drives at full speed while drive 4 was handled on the gentle settings its warnings demanded, multipass with tight timeouts, and drive 6 received the full damaged-drive discipline, its recent-failure fault worked until the maximum of its current stripe data was banked. With every member fixed as a hashed image, the array was reassembled virtually: the TaskForce's RAID engine detecting the QNAP's configuration across the images, drive 3's stale image excluded exactly as his analysis prescribed, and parity computation filling any stripes drive 6's fault had withheld — the missing blocks calculated from the surviving members rather than hoped for. The volume mounted from the reconstruction, and the filesystem was verified and extracted — the year on the tightrope ending with everything carried across.

The outcome

The array reassembled from member images — drive 3 retired as stale, drive 6 recovered and parity closing its gaps — and the data verified and delivered. Free assessment, one fixed written figure including VAT, no recovery, no fee. His log-precise enquiry earned a log-precise verdict: the analysis was right, the year of degradation was the real failure, and the doctrine stands for every array owner — RAID 5 buys you one loss, a degraded array is an emergency in working clothes, and when the second failure comes, the road back is images of every member first, reassembly on the copies, and no rebuild ever attempted on the only originals you have.

Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.