Lab accepting new cases ·  Mon–Fri, 9am–5:30pm Urgent? Call 01273 964902
BDR Brighton Data Recovery 01273 964902 Start a case
BDR / Specialist / Synology data recovery

Specialist · Synology recovery

“Volume Crashed” is DSM giving up. It isn’t the data going anywhere.

Synology's scariest message covers three completely different failures — a RAID losing members, LVM losing its map, or btrfs losing its tree — that all look identical on screen and need opposite treatment. Which is why the only universal first move is the undramatic one: power the DiskStation down and keep your finger off Repair.

No fix, no fee on most jobs Free diagnosis & written quote Post-in from anywhere in Sussex

Talk it through with an engineer
01273 964902

What you're seeing, decoded.

Something else? Run the triage →
The situationWhat it meansFirst move
DSM: Volume CrashedRAID, LVM or btrfs has failed — three suspects, one messagePower down; don't click Repair
DSM: Storage Pool DegradedA member is out but the pool limps on — redundancy already spentEvery running hour raises stakes
The DiskStation itself is deadThe array lives on the drives, not the boxOne of the happier diagnoses
A rebuild onto a new drive stalled or 'crashed'The rebuild's marathon read found a second marginal memberStop; images decide now
Drives moved to a PC, which offers to format themWindows can't read Linux NAS structures — normalDecline everything
Ransom note where the shares used to beA NAS strain has been throughPhotograph, power down, our ransomware page
Getting it to us: post your device tracked and fully insured to our secure intake lab — free return postage — or start by phone and we'll walk you through packing it. Sending details are on the contact page.

What SHR actually is — and why generic tools miss it.

Slices, not drivesSHR carves each disk into slices sized to the smallest member, so mixed-capacity drives all contribute — clever provisioning that means one volume can hide several separate arrays underneath.
Ordinary parts, unusual stackEvery layer is standard Linux — mdadm arrays across the slice groups, pooled by LVM, formatted btrfs or ext4. No Synology hardware is needed to recover it; the exact order of unwinding is everything.
Why 'flat RAID' tools come back emptyA four-drive mixed SHR-1 can be three or four mdadm arrays stacked under one LVM volume. Point a single-array recovery tool at it and the layout simply isn't there to find.
The two forbidden repairsDSM's Repair button, which can eject healthy members and overwrite metadata mid-crisis — and the btrfs repair command whose own documentation warns it can finish a damaged tree off. Both replaced, here, by read-only assembly from clones.

How we recover it, stage by stage.

See recent recoveries →
01

Booked in, diagnosed free Free

Your device is logged with its own case reference the moment it arrives. An engineer assesses the fault, confirms what's actually recoverable, and you get a fixed price in writing — no diagnosis fee, no obligation, and no paid work until you say go.

Free diagnosisFixed written quoteNo obligation
02

Image every member, read-only

Each drive is imaged with bad sectors mapped — the DiskStation never gets the chance to resume a rebuild, and the btrfs tree's historical roots are preserved exactly as the failure left them.

All members imagedHistorical roots preserved
03

Unwind the stack in order

On the images: slices identified, each mdadm array assembled, LVM's map read, and the btrfs or ext4 volume opened read-only — layer by layer, in the only order that works.

mdadm → LVM → filesystemAssembled read-only
04

Identify which layer lied

Crashed-at-btrfs, crashed-at-LVM and crashed-at-RAID get their specific repairs on the reconstruction — including walking btrfs back to an earlier intact tree where the newest one is damaged — and the shares verified before sign-off.

Layer-specific repairShares verified back
05

Verified, returned, signed off

Before you pay the recovery fee you approve a full listing of what came back. Your data returns on new media with free return postage, and the case only closes once you've confirmed everything opens on your side.

File listing approvalNew media includedFree return postage

What the lab checks first

  • One message, three diseases — btrfs tree damage, LVM metadata loss and mdadm member failure all present as Volume Crashed. Diagnosis is reading the layers, not the headline.
  • btrfs keeps its own undo history — copy-on-write means earlier tree roots often survive on disk, and stepping back one generation recovers volumes the newest tree declared lost.
  • The Repair button has a body count — mid-crisis it can eject healthy drives and write over exactly the metadata reconstruction needs. It's for healthy-ish pools, not crashed ones.
  • Exposure, not Synology, is the ransomware story — years of NAS campaigns share one profile: boxes reachable from the internet on old firmware. Off the internet and updated is most of the defence.

The architecture fact that decides these cases: a mixed-capacity SHR volume is not one array — it's several mdadm arrays across per-drive slices, glued by LVM, wearing one filesystem. Recovery that doesn't unwind those layers in exact order finds nothing; recovery that does routinely brings back volumes the DiskStation itself pronounced crashed.

From the casebook.

EX · BDR-2026-0857VERIFIED ✓

A four-bay DS with mixed drives, crashed by its own good intentions

Two 4TB and two 8TB drives, one failure, and a rebuild DSM attempted overnight that ended in Volume Crashed. Under the hood: three separate mdadm arrays, two healthy, one degraded-but-derivable. Assembled off-box in order, the btrfs tree stepped back one generation, and every share returned.

All shares restored5 days in lab

Before it reaches us.

Do

  • Power the unit down at the first crashed or degraded warning
  • Label drives with their bay numbers before removal
  • Send the drives — or the whole unit unopened
  • Say whether SHR or classic RAID was chosen, if known

Don't

  • Click Repair on a crashed or degraded volume
  • Run any btrfs repair command, ever, on the originals
  • Reinitialise drives when a PC offers
  • Reorder bays to 'test' which drive failed

Asked on this bench, answered honestly.

DSM says Volume Crashed — is my data gone?

Very unlikely. Crashed means DSM can no longer assemble the stack — because the RAID lost members, LVM's metadata broke, or the btrfs tree corrupted. Those are three different, mostly recoverable failures; the message just can't tell you which. Powering down preserves all three possibilities.

Is SHR proprietary — do I need another Synology to recover it?

No. SHR is standard Linux underneath — mdadm software RAID across partition slices, pooled with LVM, formatted btrfs or ext4. It reassembles on an ordinary workstation from drive images; what it punishes is tools and people who expect one flat array.

Should I let DSM rebuild onto the replacement drive?

Not before the members are imaged. The rebuild reads every sector of the ageing survivors — the exact stress that produces second failures — and can overwrite the historical filesystem roots that recovery falls back on. Image first; then the same rebuild becomes risk-free.

Can you recover after ransomware hit the NAS?

Often partially or fully, depending on the strain — NAS-targeting campaigns have hit internet-exposed boxes for years, and snapshots, replication targets and free-space remnants all get checked. Photograph the note, power down, and start with our ransomware page's routes.

Whatever's failed, don't power it on again.

Every restart of a damaged device costs data. Open a case first — the diagnosis is free either way.

01273 964902