Case notes · NAS & RAID · SDR-2025-0928
My Cloud Went Silent, Then Turned Odd.
For four years the box had simply sat there — on the router in the corner, never touched
— and then it began making an odd noise
and disappeared from every route at the same moment: nothing over the network, nothing over USB. The warranty spelled out the modern bind neatly: the hardware was covered, the contents were not.
Seeing the same thing yourself?
0800 6890668
The translation.
Strip away the cleverness and a network drive is one plain disk in a smart box, and smart boxes fail in a particular way: once the disk inside stops replying, every path to it closes at the same moment. That togetherness is the diagnosis. A networking fault would spare USB; a bridge fault would spare the network. Both gone at once, with a new noise for company, points hard at the mechanism — and a lone disk, unlike a RAID 5 set, has no redundancy behind it.
Kit used on this job.
What happens in a case →| Platform | Its role in this case | Why we use it |
|---|---|---|
| DeepSpar Disk Imager 4 | Head-mapped imaging, begun once the mechanism held steady | Head-map-driven imaging, with per-head timeouts, resets and power-cycle control |
| PC-3000 Data Extractor | Ran the read passes with reset control on a disk long past its best | The PC-3000 platform's imaging unit, commanded by its firmware unit |
| UFS Explorer RAID Recovery | Read the My Cloud's Linux volume arrangement out of the image | Understands how NAS volume managers really work, instead of a flat array |
In the lab.
Lift the disk clear of the box and assess it as a disk
The shell came apart and the disk was judged on its own merits. That odd noise turned out to be a mechanical failure in the head stack, which to the NAS firmware simply looked like storage that had stopped replying — hence its vanishing from both routes at once.
Repair it enough to read, then map the platters
Matched parts settled the mechanical fault, and a head map was drawn before any bulk reading started. The surfaces had suffered during the stretch when the NAS went on hammering a disk that was already failing — network storage nobody is watching will retry for weeks after a person would have given up.
Read it as Linux, because that is what it is
The image was then read for what it actually was: a Linux volume layout that WD's firmware wrote, and nothing like an NTFS disk. Shares came out with their structure intact — the difference between NAS data you can use and NAS data you merely hold.
The result.
Everything the agency named came back on fresh media; a few photos in the worn band did not. The empty shell went off to make its warranty claim — both halves settled, in the only workable order.
Similar cases on the index.
Also from NAS & RAID.
Does this sound like your own drive?
The rule holds as in every case above: switch it off, and let a free diagnosis come before any decision.