RAID & NAS · bench notes · PLY-2025-0914
The Disk Came Out and Windows Saw Nothing.
In Plympton a Seagate Personal Cloud dropped off the family's network. A neighbour, an IT man by trade, worked through it sensibly enough: the disk came out, went into a USB dock, and Windows was handed whichever driver the forums say reads Linux partitions
. Nothing changed. You can see the disk perfectly well. What you cannot see is anything to open.
An offer to initialise came up and he refused it, which was right. Older material had gone elsewhere years before; the last eighteen months had not.
Sounds like yours? Talk it through.
0800 6890668
Why it happens.
One suspect had already been eliminated by his own work: the filesystem was in order. Two others were not. Take the disk first. Certain sectors answer instantly. Others hang, and go on hanging, and an ordinary desktop machine has no plan for that — it waits, gives up on the read, and announces an empty disk. Then take the way the appliance stacks things. A home NAS puts nothing as simple as ext4 directly onto the platters. Partitioning of its own design goes down first, a volume manager sits above it, and the filesystem lives inside the pair. Good sectors can therefore be arriving at a Windows driver that can make no sense of any of them.
The kit this job called for.
How the job runs →| The kit | What it did here | Why we keep it |
|---|---|---|
| DeepSpar Disk Imager 4 | Kept reading through the stalls that had defeated both of his computers | Scores every head before it starts, works one surface at a time, and holds resets, timeouts and power in check |
| UFS Explorer RAID Recovery | Took apart every layer that stood between the platters and the folders this family used | Rebuilds the set, then works back through whatever volume layers the NAS added on top |
| R-Studio Technician | Set the rebuilt shares against the family's own account of what belonged there | Reads nearly every file system, and rebuilds RAID sets from images |
How it ran.
Run his test again, this time on lab hardware
Everything he had found was found again on the bench, this time with numbers on it. A few milliseconds for some reads. Close to thirty seconds for others. Fast and slow lay mixed right across the surface in no order anyone could work out. Two or three attempts is all an operating system will make at a disk behaving that way, and both his machines stopped there.
Imaging hardware that will not walk away
Timeouts were cut short on purpose, and stalls answered with a reset instead of a wait. Most of the surface was sound, and that let it move across at a decent rate. Awkward ground was noted and left for the moment, then picked up at the end, by which time nothing readable remained exposed.
Take the layers apart in the order the box built them
From there the work was all done on the image, in whatever order the appliance itself had used: partition table, volume manager, ext4. Taken that way round, the shares came back looking as the household knew them, with eighteen months of photographs kept nowhere else sitting where they had always sat.
How it ended up.
Those months travelled home on an ordinary external drive, to sit beside the older years. No harm whatever had been done by the neighbour; each step he took was read-only, and it narrowed the ground we had to cover. This one wanted imaging hardware and a decode taken a layer at a time. Software by itself could never have got there. Two copies are made of everything off the phones in that house now.
Pages people open after this.
Other jobs on RAID & NAS.
Seeing the same thing?
Leave it switched off and post it in. Nothing happens until the diagnosis, which lists what can still be read and what cannot.