Specialist · Synology recovery
One red banner on an Exeter DiskStation, three different troubles underneath: RAID members gone missing, LVM metadata in a muddle, or a btrfs tree that no longer adds up. DSM shows them all identically, yet each needs its own treatment — and treating the wrong one makes things worse. Until somebody has read the disks properly, the safe course is simple: power the box down and leave the Repair button alone.
Talk it over with an engineer
0800 6890668
| What's wrong | What it usually means | Step one |
|---|---|---|
| DSM has flagged Volume Crashed | A single message hiding three suspects — btrfs, LVM or the RAID | Power down; don't touch Repair |
| Storage Pool shown as Degraded | One member has gone — everything still works, minus the safety net | The clock is now against you |
| The DiskStation itself won't start | Nothing moved — the data lives on the disks, never in the box | One of the kinder faults |
| A rebuild onto a new disk stopped partway | The rebuild's heavy reading showed up a second tired disk | Halt it; imaging comes first |
| A PC now offers to format the NAS disks | Expected — the Linux layout means nothing to Windows | Say no, every time |
| Your shares replaced by a ransom note | A strain that hunts internet-facing NAS boxes | Snap a photo, power off, then read the ransomware page |
Your device is logged under its own case number the day it arrives. An engineer finds the fault, tells you plainly what stands a real chance of coming back, and writes you one fixed quote. Nothing is charged for that look, nothing obliges you, and paid work starts only when you say so.
Each drive gets a sector-mapped image before anything else happens. The box itself stays powered off, so no rebuild can restart and the older btrfs roots survive exactly as the failure left them.
All of it on the images: partition slices identified, the mdadm arrays put back together, LVM decoded, and finally the btrfs or ext4 volume brought up read-only. Order matters here, and we keep to it.
Whichever level failed — RAID, LVM or btrfs — the remedy goes onto the rebuilt copy, never the originals. Where the current btrfs tree is ruined, an earlier healthy generation is used instead, and every share is opened and checked.
Before any fee falls due you see a full listing of everything recovered and approve it. Your files travel back on brand-new media, return postage on us, and the case stays open until you've confirmed everything opens on your own machine.
The fact that decides these cases: mix drive sizes under SHR and you don't get one array at all — you get several mdadm arrays laid across slices of each disk, gathered into a single filesystem by LVM. Approach that stack in the wrong order and nothing appears; take it apart correctly and a 'crashed' volume usually proves to be all there.
Two 4TB disks, two 8TB, one failed member — and by morning the DSM rebuild had ended in Volume Crashed. Under the hood lay three separate mdadm arrays; two were sound, the third degraded yet still workable. We put them back together away from the box, stepped the btrfs tree back one generation, and every share returned.
Very rare. 'Crashed' only means DSM couldn't reassemble the stack: missing RAID members, broken LVM metadata or a corrupted btrfs tree. Three different faults, most of them fixable — and the message can't say which you've got. Keep the box switched off and every route stays open.
No. Strip the branding and SHR is standard Linux: mdadm software RAID built over partition slices, LVM pooling them, and btrfs or ext4 formatted on top. Given good images of the drives, it reassembles on an ordinary workstation — problems only arrive when a tool, or a person, expects one flat array.
Only once each member disk is safely imaged. Rebuilding reads the worn survivors end to end — exactly the load that finishes off a second drive — and it may overwrite the older filesystem structures a recovery depends on. With images taken first, the rebuild can do no harm.
Some of it usually, occasionally all of it — the strain decides. Internet-facing NAS units have been targeted for years, so the hunt covers snapshots, replication copies and what's left in free space. Take a photo of the ransom note, power the box off, then follow the ransomware page's routes.
Each power-up of a failing device takes more data with it. Open a case first — the look-over is free whichever way it goes.