Data Recovery Case File · NAS & Network Storage · Both Drives Were in the Fall
The Mirror Covered One Failure and They Had Two
This enquiry describes a sequence in which the remedy was the second failure. A mirrored pair in a network unit: "these were dropped and one drive failed right away — a head crash, I would imagine. They then tried to rebuild the array, which caused the second drive to fail." A mirror protects against one drive failing. It does not protect against both being dropped, and the rebuild asked the survivor for the heaviest read of its life within hours of the same impact.
| Media | Two 3TB drives from a mirrored network unit — both subjected to the same impact; first member failing immediately; second failing during array reconstruction |
| Reported situation | Network unit in a mirrored configuration · unit dropped · one member failing immediately with mechanical symptoms · array rebuild initiated · second member failing during that rebuild · both members required |
| Fault class | Impact damage to both members with rebuild-induced failure on the survivor — mirror redundancy addressing single-member failure only |
| Equipment used | Both members assessed independently as separately damaged devices · no further rebuild attempted · laminar flow bench inspection of each before any spin attempt · imaging under strict per-sector timeouts · surviving content reconciled across both images |
The decode: what the mirror covered, and what the rebuild demanded
What redundancy actually promised here: that one drive could fail without loss. That promise was kept — when the first member failed, the volume remained available on the second, exactly as designed.
What it never promised: that both members would not fail together. Two drives in one enclosure share everything — the same shelf, the same power, the same room, and in this case the same fall. A mirror is protection against an independent failure, and an impact affecting the whole unit is not independent.
So the position immediately after the drop: one drive obviously damaged, and one drive that had taken the same impact and happened to still be working. The second was not undamaged; it was undamaged so far.
What a rebuild asks of a survivor: a complete read. Every sector, sequentially, at maximum sustained load, for as long as the capacity takes — which is the single most demanding thing that drive will ever be asked to do, and it was asked within hours of being dropped.
Why that is the worst possible moment for it: a drive with impact damage may have displaced heads, marginal surfaces, or a mechanism that is holding together at rest and not under sustained work. Ordinary use touches a small fraction of a drive; a rebuild visits all of it, so a drive that seems fine will meet its damage within the first pass.
So the second failure was not bad luck: it was caused by the response to the first. The instinct to restore redundancy immediately is exactly what the arrangement trains people to have, and it is the wrong instinct after a physical event, where the correct move is to stop and image both members before anything is rebuilt.
Why the position is nevertheless workable: a mirror's members each hold a complete copy. Two damaged drives holding the same content are two chances at every sector — a region unreadable on one may read on the other, and the images can be reconciled against each other. That is a materially better position than a single damaged drive.
What must not happen: no further rebuild, and no attempt to bring the array up. Both members are now damaged devices and each is assessed on its own.
On the bench
Both members were assessed independently as separately damaged devices — a mirror protecting against an independent single failure while two drives in one enclosure share the same shelf, supply and impact. No further rebuild was attempted, reconstruction requiring a complete sequential read of every sector at sustained maximum load, which is the most demanding operation a drive performs and was requested hours after the same fall. Inspection ran under the laminar flow bench before any spin attempt, with surviving content reconciled across both images.
The outcome
Both members treated as separately damaged, imaged under capped timeouts and reconciled against each other. Free assessment, one fixed written figure including VAT per drive, 50% of parts and labour upfront with the balance only on successful recovery. The decode: the mirror kept its promise — one drive failed and the volume survived. What it never covered was both being in the same fall. And a rebuild asks the survivor for a complete read at maximum load, which is the worst possible demand hours after an impact.
Mirrored pair damaged in the same accident
Don't rebuild — image both drives first. The instinct to restore redundancy immediately is exactly what the arrangement trains you to have, and it's the wrong one after a physical event. A mirror protects against one drive failing independently; two drives in one enclosure share a shelf, a power supply and any impact, so a survivor that still works after a drop isn't undamaged, it's undamaged so far. A rebuild then asks it for a complete sequential read of every sector at sustained maximum load, which is the most demanding thing it will ever do. The encouraging part is that both members hold the same content, so you get two chances at every sector.
Don't rebuild — call Oxford Data Recovery on 01865 593000; both members assessed independently, inspected before any spin attempt, imaged under capped timeouts and reconciled against each other.
Request a quote online →
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.