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

Specialist · Synology recovery

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

The Synology "scary" failure message describes three totally separate types of failure: a RAID losing some members, LVM losing its layout map, or btrfs losing its tree. They are indistinguishable visually and require opposite treatment. This is why the only common initial action is the mundane one: turn off the DiskStation and do NOT press Repair. The next step should be to obtain a quote, and there are two ways to do this: you may call us, or use the quote request form on this website. If you can remove the disks from your Brighton DiskStation individually, they are sent labelled by bay; if you cannot, it is sent as a complete system. Either way, post it tracked and insured or drop it off at our Guildford intake lab, Monday to Friday, 9am to 5:30pm (no appointment needed). Both methods arrive at the same workbench, from £500 + VAT.

Only pay if we recover your data, 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 problem as describedWhat it meansFirst thing we do
DSM: Volume CrashedYour RAID, LVM or btrfs has failed: three possible causes of the failure but only one error message displayed on screen.Turn off your DiskStation and do not attempt to Repair the system.
DSM: Storage Pool DegradedOne member has gone out, but the RAID still runs, with its redundancy already used upWith each hour it keeps running, you increase the risk of losing additional components of your RAID configuration.
The actual DiskStation diedThe array configuration remains intact on the drives, regardless of whether or not the DiskStation is operational.One of the better diagnoses
A rebuild on a brand new drive failed to finish (or crashed)Marathon read from the rebuild: a second marginally bad member was detectedStop; images decide now
Drives were taken to a PC and offered for formatting by WindowsWindows does not recognise Linux NAS data structures: normal behaviourDecline everything
Ransom note where shares used to beA NAS strain has been through thisPhoto, power down, our ransomware page
Getting it to us from Brighton: get a quote first by phone or with the form. And we will walk you through packing it. Then post your device tracked and fully insured to our intake laboratory at Guildford (free return postage), or drop it off there Monday to Friday, 9am to 5:30pm (no appointment needed). Details for sending are located on the contact details page.

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

Slices, not drivesEach disk is divided by SHR into slices that match the size of the smallest member. This means that even mixed capacity drives will be utilised in this configuration. The SHR provisioning method allows one volume to have several separate RAID arrays hidden underneath it.
Ordinary parts, unusual stackAll layers are standard Linux. The disk slices are grouped together using mdadm software RAID and then combined into larger pools by LVM. These large pools are then formatted as either btrfs or ext4. Since no Synology hardware is needed, recovery can take place on a standard workstation. However, the process of recovering data from these disks involves a great deal of detail and understanding regarding how to unwind the original storage setup in exactly the right order.
Why 'flat RAID' tools come back emptyA four-disk mixed SHR-1 configuration can be three or four different arrays created by mdadm and stacked within a single LVM volume. Therefore, if you try to use a tool designed to recover a single array against this setup, it will never be able to determine the correct layout.
The two forbidden repairsThe "Repair" option within the DSM software can remove healthy disks from an array and rewrite the meta-data while the job is still running. The btrfs repair command has its own documentation warning users that it could potentially destroy a damaged file system. In both cases, we replace the use of those features with read-only assembly from cloned disk images.

How we recover a Synology, stage by stage.

01

Booked in, diagnosed free Free

Your device is logged under its own case reference as soon as it arrives. An engineer assesses the fault and confirms what is actually recoverable. And you receive a fixed price in writing: no diagnosis fee. No obligation. And no paid work until you say go.

Free diagnosisFixed written quoteNothing to sign
02

Image every member, read-only

All hard drives are imaged with all bad sectors marked: DSM will no longer be able to attempt to rebuild any failed drives; the btrfs tree will retain its original configuration as well as its history at the time of the failure.

All members imagedHistorical roots preserved
03

Unwind the stack in order

Drive images contain identifiable sections (slices) that are assembled into each mdadm array; LVM’s map is read; and either the btrfs or ext4 volume is opened in a read-only mode: each layer must be built in the correct order in order to assemble properly.

mdadm → LVM → filesystemAssembled read-only
04

Identify which layer lied

The "re-assembly" for each of crashed-at-btrfs, crashed-at-LVM and crashed-at-RAID includes its specific repairs, such as using tools to walk btrfs back to an intact tree at some previous point in time, where the current (newest) tree was damaged, and the shares are verified prior to signing off.

Layer-specific repairShares verified back
05

Verified, returned, signed off

You approve a full listing of everything that came back before you pay the recovery fee. Your data returns on new media with free return postage. And the case only closes once you have confirmed all items open on your side.

File listing approvalNew media includedFree return postage

What the lab checks first

  • One message, three diseases: A Volume Crashed message could be due to damage to a btrfs tree, or loss of LVM metadata, or failure of one of the members of an mdadm RAID array. The diagnosis depends on reading the layers, not on what the headline says.
  • btrfs keeps its own undo history: Because btrfs uses copy-on-write, there often still exist on disk older versions of the root of the btrfs tree that were created before the newest version that reported a volume was lost. By going back one generation of the tree it is possible to recover the volume that the newest version of the tree said was lost.
  • The Repair button has a record of damage: Before the end of the job the Repair button can remove good drives from the pool, and write over the very same metadata that re-assembly needs. This option is primarily designed for use with healthy-ish pools, not crashed pools.
  • Exposure, not Synology, is the ransomware story: For multiple years now we've seen a common profile for all NAS-based ransomware campaigns: the boxes could all be reached from the internet, and had outdated firmware running. The majority of defence against this type of threat lies in keeping your box off the internet, and keeping your firmware up-to-date.

The architecture fact that decides these cases: A mixed capacity SHR volume is not a single array: it consists of multiple mdadm arrays across per-drive slices, held together by LVM under one filesystem. A successful recovery will unwind those layers in the correct order; if they aren't unwound successfully, then nothing will be found; but if they are unwound correctly, then the recovery will typically bring back volumes that the DiskStation itself previously called crashed.

From the casebook.

EX · BDR-2026-0857CHECKED ✓

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

Two 4TB and two 8TB drives, one failed, and DSM did a rebuild attempt overnight that resulted in Volume Crashed. Behind the scenes: three distinct mdadm arrays, two active/healthy and one degraded-but-recoverable. Once rebuilt outside of DSM in the right order, the btrfs tree stepped back one generation and every share came back.

All shares restored5 days in lab

Before your Synology reaches us.

Do

  • Turn the device off when the first error message appears related to a crash or degradation
  • Mark each drive prior to removing with a label indicating which bay it was located in
  • Obtain a quote prior to sending the drives, or consider sending the entire unit sealed and unopened if they refuse to come out
  • If you know which type of RAID was used (SHR or traditional RAID), let us know

Don't

  • Repair your crashed or degraded RAID using the "Repair" option in the menu
  • Run any btrfs repair commands at any time on your original disks
  • Initialise the drives immediately when your computer prompts you to
  • Re-arrange the order of your disk bays to test which disk may have failed

Asked on this bench, answered honestly.

DSM says Volume Crashed — is my data gone?

This is highly unlikely. A "crash" means DSM was unable to reconstruct the disk stack: either due to the RAID losing members, corruption within LVM's metadata structure, or corruption of the btrfs file system. These represent three very different, mostly recoverable types of failure, and DSM simply doesn't provide enough information to allow you to identify which type occurred. Turning off power leaves all three options open.

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

No, because SHR uses standard Linux under the hood (mdadm software RAID on top of partitioned slices, pooled by LVM, formatted as either btrfs or ext4). It can also be reassembled on an ordinary workstation from copies of the individual disks. What it does punish is people and programs that assume they're dealing with one single flat array.

Should I let DSM rebuild onto the replacement drive?

The rebuild should not occur until after the members have been imaged. The rebuild reads every sector of the ageing survivors, which is the exact stress that creates secondary failures, and can write over historical filesystem root points that recovery will fall back on. Therefore, image the disk set first; once this has occurred, the rebuild process is risk free.

Can you recover after ransomware hit the NAS?

Many times, in part or completely, depending upon the version: NAS targeting malware has attacked internet exposed boxes for years, and snapshots, replication targets and left over space all get checked. Take pictures of the note, turn the box off, and start using our ransomware page's routes.

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

Each time a damaged device starts up again, data is lost. Open a case first; the diagnosis is free either way.

01273 964902