Devices · NAS and network drives
Most NAS failures are quiet ones. Everything still mounts, everything still opens, and a disk can be dead in bay three for a month before anyone looks. The damage comes at the next step: DSM or QTS puts up a scrub, a repair, a rebuild; somebody accepts; the pool goes. Power the box down and leave it down, wherever in Hull it stands. They reach us from a freight office on Hedon Road, a solicitor's practice in Beverley, a fabrication shop outside Scunthorpe.
An engineer will take the details
0800 6890668
| What you see | What the fault is | What to try |
|---|---|---|
| Synology: Volume Crashed | Too many disks gone for parity to cover | Cut the power before DSM suggests a fix |
| Synology: Storage Pool Degraded | One disk is out; the layers above carry on regardless | Every hour it runs costs you |
| QNAP: system volume not active; RAID unmounted | QTS built the volume and can no longer mount it | Put it together off the box |
| DeadBolt — WARNING: Your files have been locked | Ransomware; .deadbolt on the end of every file | Photograph the note; change nothing |
| Buffalo TeraStation E14 — Cannot mount the RAID array | Nothing wrong with the array; the box will not start it | Write the E-number down, then unplug it |
| Buffalo TeraStation E16 — The hard disk was not found | Either the disk is dead or the box has lost sight of it | The files are generally still on it |
| Buffalo TeraStation E30 — The hard disk may be broken | A disk has been thrown out of the set | Refuse the rebuild |
Each item that arrives gets a case number the day it lands. An engineer looks at it and finds the real fault, and that costs you nothing. You are told in plain terms which files can come off the drive and which cannot. A price follows — one figure, written down. No work starts until you have seen it and agreed to it.
Every disk gets a full copy on imaging hardware built for the job, weak areas included. After that, all the work is done against those copies. The enclosure sits powered off on our bench, so it never picks up where it left off.
A NAS volume is almost never a single object. On Synology it is matched partitions, an md array over each group, LVM over all of it. Buffalo, Netgear and QNAP arrange theirs differently again. Every tier goes back up in software, bottom first.
Once the stack holds, btrfs or ext4 is repaired. Shares come back named as you named them, with the folder tree as it was. You get the full listing and check it yourself; nobody here closes the job until you have.
You are not billed while you are still thinking. The list of what came off the drive goes to you first, and the invoice waits on your word. Files go home on media bought in for the job, with the postage back on us. We keep the job open until you have opened them on your own machine.
The lesson of 2022: the first DeadBolt wave, on 25 January 2022, caught roughly 3,700 QNAP boxes that answered from the open internet, and more waves came through that year. QLocker and eCh0raix had already been over the same ground. Two things were true of nearly every victim: the box could be reached from outside, and its firmware was whatever shipped in the carton.
Two red lights, and the box was still savable. It was the rebuild someone kicked off next that finished it: the unit's own metadata got written over. Each disk was then imaged past its bad sectors. The mirror went back together a block at a time, every one taken from whichever disk had read cleanly there. The shared volume came out complete.
Rarely. The disks carry their own description of the volume: which bay each one sat in, the stripe size, every layer stacked above. None of that lives in the chassis. Good disks in a dead box is among the easier work we get.
Just the disks. Bare drives are what the Leeds lab wants. Take them out one bay at a time and write the bay number on each before it is packed. Post them fully insured, or drop them in yourself. No disk is read for data until it has been imaged.
Synology Hybrid RAID is there to stop odd-sized disks wasting capacity. Every drive is sliced into partitions that find partners across the set; an md array sits over each matched group; LVM ties those arrays into the single volume you see. We take that apart layer by layer, starting at the bottom.
No. Volume Crashed describes the limits of the box, not the state of the data. Copy every drive, put the pool together in software, and a volume DSM has written off usually reads right through.
Leave it switched off. A failing drive has a limited number of starts in it, and each one spends a little more. The free diagnosis will say which files can be read.