Taking work now — the first look is freeSpinning disks in a machine Quicker still, give us a ring:0203 0868633
ADR Advanced Data Recovery 0203 0868633 Price my job
ADR / Media and machines / Synology, Netgear NAS recovery

Specialist · Synology, Netgear and D-Link NAS

Synology and Netgear NAS recovery, London. A volume the unit will not mount belongs on a bench, not in one more rebuild.

RAID, NAS and server work is priced from £500 + VAT, and the assessment in front of it is free. Four shapes account for nearly all of it. A Synology comes up to DSM with the storage pool no longer listed. A rebuild halts because a second bay has gone amber. A ReadyNAS suggests removing inactive volumes, which is a suggestion that it erase them. Or an old D-Link ShareCenter goes quiet, years beyond the last firmware it will ever see. Whichever it is, the disks are cloned read-only before anything else happens, the layout of the array is read back from those clones, and file system work is done on clones as well — never on what you posted. Sets arrive from Bromley, Hammersmith, Richmond, Camden and Westminster.

On most jobs: no data, no bill A free look first, then one figure in writing Media posted in from Croydon, Ealing and Camden

Put it to an engineer — the first look costs nothing
0203 0868633

Synology and Netgear: what the symptom usually means.

Different fault? Try the finder →
What is going onThe usual reason for itWhere that leaves you
The unit starts and there is no storage poolWhat the array says about itself is damaged. What lies beneath it is notPower it down, then post the disks
A rebuild stopped with a second bay amberEvery ageing disk in the box was told to read every sector of itself at onceOff at the wall, now
The volume comes up read-only, then disappearsA file system fault above a set that is otherwise soundComes back, usually
One bay is amber and the shares are still openOne member is out, and there is no protection left on the restHave an image taken today
The firmware updated and the volume never returnedThe enclosure, not the disks — the members behind it are intactPost the drives, numbered by bay
A drive put in a PC reads as unformattedA Linux array means nothing to Windows, so it offers to initialiseRefuse, and leave it alone
Packing it and posting it:a damaged drive is finished off by movement, so use a rigid box and pad it out. Declare the parcel at what the files are worth to you, not at what the disk cost, then send it tracked to our intake lab. We meet the cost of the trip home. Rather have the packing checked over before you tape the box up? Ring first. It is all set down on theguide to packing and posting.

Where these units actually stop.

Synology DS and RS unitsTwo bays beneath somebody's desk, or eight and twelve in a comms cabinet: between them these carry most of the work. What DSM presents as a pool is an ordinary Linux array underneath, a volume manager over that, ext4 or Btrfs on top — which is why images alone rebuild a set. Hybrid RAID adds a wrinkle, laying several arrays across disks of unequal size.
Netgear ReadyNAS and D-Link ShareCenterReadyNAS OS 6 lays Btrfs over an expanding X-RAID array, and the older NV and Pro units used ext3 instead. D-Link's ShareCenter boxes are simpler still, usually a mirrored pair with ext underneath, and long past any firmware. Both write the layout onto the members, so both go back together here from clones.
Pools, snapshots, iSCSIA good deal gets stacked over the raw disks in these boxes: pools, thin volumes, Btrfs snapshots, iSCSI LUNs served out to other machines. One of those layers can break with everything beneath it perfectly whole, and separating them is where most of the effort goes on a unit that starts to an empty pool.
Never done hereNobody here starts a rebuild, imports anything, makes a fresh pool, or lets firmware near what you sent us. Every one of them writes to a disk, and a write shuts a door. A member is read once and then left alone. Repairs are made to an image, so a wrong turn costs this bench an afternoon and costs you not a penny.

From the box arriving to your files going back.

Work we have closed →
01

Logged the day it lands, and the first look costs nothing Free

A case number goes on it the day the parcel lands, and an engineer settles what has truly failed before anything else happens — free of charge, and first in the order of work. Back to you come two things together: a straight note of what is liftable and what is not, plus one figure, fixed and written down. Accept it, or decline and owe us nothing.

Nothing to pay for lookingA single figure, put in writingNothing owed at this stage
02

Enclosure and disks are told apart

Every drive is lifted out and put on a write-blocked port, so it answers for itself and not through a failed board. The one the unit rejected goes on as well. That decides what no front panel ever can, which is whether the enclosure or the members are at fault. Where the board or the firmware is the culprit and the disks behind are sound, the job is far smaller than the lights would have you believe.

Each disk read separatelyThe failed part is named
03

Cloned before a thing is assembled

Copies come first. A surface that reads badly is set aside while the straightforward ground goes down, then worked in short runs. Everything after the cloning runs against an image. Your own drives are never written to, at any stage, by a single byte.

Images before anything elseYour drives are never written to
04

The array first, the file system after

The unit records all of it on the members themselves — which bay, how wide the stripe, the offset the data starts at, the volume layer over the top — so nobody needs the enclosure to reassemble anything. Over the images the set is rebuilt, mounted read-only, and the file system put right there. After that, shares and mail stores and databases are opened up on the bench. A name in a listing proves nothing about whether the file under it will open.

Built over the imagesOpened here first of all
05

You see the file list before you pay

What was recovered is listed for you first, and only then does a bill exist. Approve the list and it is invoiced; turn it down and it is not — and where nothing has come back, most jobs carry no charge whatever. Recovered data travels home on fresh media bought in for your job, with the postage at our end. Your case is not closed until you have read those files on a machine of your own.

No charge until you accept the figureFresh media, supplied with the jobThe post home is ours

What arrives most often

  • Enclosures fail more often than disks — a power board, a fan that gave up, an update that stopped halfway, and DSM will show you an empty pool while a shelf of perfectly sound members waits behind it. This is the cheap version of the job, and which version you have is established by the free assessment, before any figure is agreed.
  • Nothing punishes a set like a rebuild — to finish one, every sector of every remaining member has to be read. Those members were racked on the same day, from the same batch, and have run the same hours since. A second amber bay halfway through is not misfortune, then. It is the frailest disk left in the box being asked for the hardest work it has ever done.
  • Read the prompt before you agree to it — when a ReadyNAS proposes removing inactive volumes so the disk becomes usable, it is proposing to erase them, not to mend anything. When Windows proposes initialising a Synology disk, all it means is that Linux is beyond it. A great many people read either as proof the disks are finished. Neither is. Saying no to both is the single most valuable act described anywhere here.
  • Number the bays before a single disk is pulled — the order can be recovered from the drives at this end, but a note saying which one sat where, together with a frank account of what has been tried already, is what decides how fast the set goes back together. And send them all. The one marked bad is often the member still holding the newest blocks.

Four things decide a job like this: the proportion of each member that will still read; whether a pool was created afresh once the old one went; the length of time a rebuild ran; and whether encryption was ever turned on. Shut a unit down with every disk in place and your options are as wide as they get. Slot a spare in, kick off a rebuild, flash the firmware twice and leave the thing running all night, and they narrow a long way — you are still told which files land in the gaps, and hearing it still costs nothing. Where a volume is encrypted the key has to come from you. Storage is our whole trade: no phones, no tablets. The published band for RAID, NAS and servers begins at £500 + VAT, put in writing after the free assessment, and on most of it you owe us nothing at all unless your files are back.

One job, followed all the way through.

LDN · ADR-2026-3391JOB LOGGED ✓

After a firmware update, a four-bay QNAP with no volumes left

An Ealing pattern-cutter kept two decades of work — nesting files, grading sets, product photography — on a four-bay QNAP that sat beneath the cutting table. After an update the box restarted, and the storage pool had gone: four bays lit green, no volumes, no shares, nothing on the screen. It was unplugged that afternoon and the disks came to us numbered one to four. All four cloned through without a snag, which pointed at the enclosure rather than the disks. The layout of the array was read from the images, the volume brought up over them read-only, and every share was there. Parcel to parcel it took three days, and the patterns went home on a new drive, ordered as the cutting room had always ordered them.

100%imaged sector for sector3days here, and back to you

What helps, and what harms.

Do this much first

  • Switch the unit off as soon as a bay turns amber
  • Number each disk as it leaves its bay
  • Send every drive, the one marked bad as well
  • Say what has already been tried on the unit

What sets us back

  • Beginning another rebuild with a member absent
  • Updating firmware on a set that is already short
  • Creating the storage pool again to see if it returns
  • Letting Windows initialise one of the members

Questions answered before you commit.

The unit starts and the volume is not there. Are the files still on the disks?

Almost always. Every member in one of these boxes is inscribed with the layout: its own place in the order, the width of a stripe, the offset the data starts at, the layer built over the top. That inscription is fragile in a way the blocks below it are not. So a pool that has gone from a unit showing green across all its bays is telling you what the enclosure can no longer make sense of, not what the platters hold. Turn it off. Leave the disks where they sit. Copies are made before any of it is put back.

Do I send the NAS itself, or just the drives?

Whichever suits the box. Send a desktop Synology or a two-bay D-Link whole, switched off, with its adapter — nothing about the job is worse for it arriving that way. Rack units weigh too much, so draw the disks and write a number on each as it leaves, so bay four reaches us marked bay four. Any drive the unit had given up on comes too: very often that is the member holding the freshest blocks. Everything reaches this bench by post. Send it tracked, and declare it at the value of the data rather than the value of the box.

Windows offered to format the disk when I plugged it in.

Refuse it and unplug the disk. What these units put on a member is a Linux array, most often with a volume manager on top and ext4 or Btrfs above that, and Windows can make sense of none of it. Offering to format is the only response it has. Nothing is blank about your disk and nothing is broken about it. Say yes to that prompt, though, and a fresh structure gets laid across the opening sectors, which makes an ordinary job a fiddly one. If it has already been done, tell us — the work reorders itself around it, it does not stop.

What does a NAS job cost, and how long does it take?

From £500 + VAT. Anything holding a second disk falls in that band — RAID, NAS and servers together — and the price page publishes it, so there is no need to ring and ask. Order of events: free assessment, then one figure written down, then your yes before anything chargeable is begun. Should nothing come back at all, most jobs are not invoiced. Allow three to four working days on a multi-disk set, counted from the parcel landing to new media going out again. Anything genuinely urgent is moved up, and we pay the postage home.

Begin here if yours is doing the same thing.

The drive inside the machine → The noises it is making → Nothing detects the disk → Every device we take →

Nothing gets worse while the power is off.

Looking at it is free. Back comes a list of what opened and what did not, together with a single price to finish, set down in writing while you are still free to say no. On most jobs an invoice only follows the data. Until that list reaches you, leave the drive switched off.

0203 0868633

Get it assessed Sending it to Guildford