Taking work now — the first look is freeServer and RAID drives posted in from anywhere in the UK, or handed in at ten drop-off pointsQuicker still, give us a ring:0800 6890668
RDDRAID Drive Data Recovery 0800 6890668 Price my job
RDD / Every kind of member drive / NVMe U.2, U.3 and M.2 members

U.2 · U.3 · E1.S · M.2 · PM9A3 · P4510 · CM6 · SN840 · Micron 7450

NVMe U.2, U.3 and M.2 members. PCIe drives in the front of the server, and nothing in a desktop to plug them into.

NVMe members sit in the front of a server on a PCIe backplane, in U.2 and U.3 carriers, E1.S rulers, or as M.2 sticks on a BOSS card mirroring the boot volume. They are Samsung PM9A3 and PM1733, Intel and Solidigm P4510 and P5520, Kioxia CM6 and CD6, WD SN640 and SN840, Micron 7450, and they fail like any other solid-state drive: a controller stops, a firmware image trips, a power event leaves the mapping table inconsistent. HPE's NVMe SSDs on one firmware line failed after 4,700 hours until fix 4ICRHPK3, which is the same clock story on a faster bus. They need a PCIe adaptor for their form factor to be read outside the server, and a bench that knows not to reset the cluster to see whether they come back. One NVMe member imaged is £300 + VAT after the free look, fixed in writing, 3–4 days at the bench.

Free first lookOne fixed figure in writingNo data, no bill on most jobsReturn postage paid

Rather talk it through? An engineer answers the bench line
0800 6890668

Power down. Label every slot before a drive comes out. Do not rebuild while a second drive is flagged. Do not clear or import a foreign configuration. Do not initialise, reformat or PSID-revert anything. A rebuild reads every sector of every surviving member, and on a set whose drives were bought together it is where the second failure is found. Nothing on the drives gets worse while the server is off.

NVMe member symptoms, and what each means.

Not listed? Describe it on the form →
Missing after a rebootController or firmware. Do not reset the cluster or the host to see.Do not delay
Several members gone after a firmware updateA batch trigger. Stop updating.Do not delay
Read-only, or capacity wrongMapping table or controller. Image now.Do not delay
Failed after a power eventMapping table inconsistent. Imaged at controller level where supported.Everyday job
BOSS M.2 mirror degradedOne stick failed; the other is the boot volume. Image both.Everyday job
vSAN disk group downA cache or capacity NVMe gone. Send the group, labelled.Do not delay
Packing it and posting it: power the server or shelf down, write the slot number on each drive with a marker before it comes out of its carrier, and keep it in the carrier. Send every member from the set where the set is the job, or the one failed member where the set is healthy and you want it imaged for the rebuild. Each drive travels in an anti-static bag inside its own padding, in a box with nothing able to move. Send the controller only if we ask for it. Insure the parcel for what the data is worth rather than the price of the drives, and use a tracked service. The posting address is not printed anywhere on this site; it arrives by email in reply to the form, with a booking sheet to print; the sheet inside the parcel is what matches it to your enquiry when it is opened. Or hand the sealed parcel in at the nearest of ten drop-off points, your name on the outside and the sheet inside; say where you are on the form and it comes by email. We pay the postage home either way. The whole of it is written up on the guide to packing and posting.

Where it lives, and what fails.

Form factors and adaptorsA U.2 drive needs an SFF-8639 to PCIe adaptor; a U.3 drive a tri-mode one; an E1.S ruler its own carrier; an M.2 stick a slot with the right key. None of them plugs into a SATA port, and the bench has adaptors for all of them.
Controllers and mapping tablesAn NVMe controller keeps a table that maps logical blocks to NAND pages. A power event that leaves it inconsistent gives a drive that is present and unreadable; recovery is at controller level, model permitting, and the free look says whether the model permits.
The 4,700-hour clockHPE's NVMe advisory a00111900 named drives failing after 4,700 hours of operation, fixed in firmware 4ICRHPK3. Drives fitted together fail together, on NVMe as on SAS.
vSAN and Storage Spaces DirectA failed cache NVMe takes its disk group down; a failed capacity drive takes the components on it. The drives are imaged individually and the object layout read from the images; the hypervisor's own repair is not run.

Model strings, and what they tell us.

Describe yours to us →
Model string What it is What it means on the bench
MZQL2…HBLASamsung PM9A3 U.2, 1.92–7.68TBCommon in Gen10 and R750 backplanes
SSDPE2KX…Intel P4510 U.2Now Solidigm; firmware from Intel's era
KCM61RUL…Kioxia CM6-R U.3Tri-mode backplanes
WUS4BB…WD Ultrastar DC SN840 U.2Dual-port on some
MTFDKCC…TFRMicron 7450 PRO U.3vSAN nodes

From the parcel 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 the parcel and a number on every drive the day it is opened, matched to the slot you wrote on it. An engineer settles what has actually happened before anything spins: the interface, the sector format, the firmware, the security state, and what the controller was saying when it stopped. 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 writingNo rebuilds, no imports, no resets
02

The drive on its own bench

An NVMe member goes onto the PCIe adaptor its form factor needs, and its controller is asked once. If it answers, it is imaged immediately; if it does not, the drive is examined for what has failed, and the free look says whether the model can be worked at controller level.

Read at its native interfaceFailing heads to the clean bench
03

Imaged once, at its native sector size

An NVMe member that answers is imaged in one pass with its weak blocks last, at whatever speed the controller will tolerate. One that does not is worked at controller level, model permitting, and the result is an image or an honest no.

Head by head, weak areas lastNative sector size preserved
04

The image, and the set

Where the set is healthy and one member was the job, the image goes back to you on a fresh drive, sector-identical, ready to rebuild from. Where the set is the job, every member is imaged and the array reassembled in software from the images: order, chunk size, parity rotation and the reshape point read from the drives' own metadata, the file system repaired on the virtual volume, never on the originals. The array work is covered in full on our RAID array site, and it happens on the same bench.

One member: a sector-identical image on fresh mediaA set: reassembled in software from the images
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 at all. 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 opened the files on a machine of your own.

No charge until you accept the figureFresh media, supplied with the job3–4 days at the bench

What arrives most often

  • Do not reset the host or the cluster to see whether the drive comes back. A controller that is failing answers fewer times each attempt.
  • Send the whole disk group where vSAN or Storage Spaces Direct is involved; the object layout spans the drives.
  • The adaptor matters. A U.3 drive on a U.2 adaptor, or an M.2 stick in the wrong key, reads as nothing and tells you nothing.
  • Check the power-on hours on the survivors when one NVMe member goes: the clock stories apply here too.

One job, followed all the way through.

UK · RDD-2026-0560JOB LOGGED ✓

A PM9A3 out of an R750's vSAN node, missing after a power cut, with the disk group down and the cluster one failure from losing an object

The owner did not reset the host, which is why the other drives in the group were still readable. The missing drive was imaged at controller level on a U.2 adaptor, its mapping table rebuilt from the image, and the disk group's components read from the images of all four drives. The objects came back and the cluster was resynced from them.

100% of the disk group recovered5 days here, and back by post
Illustrative example — replace with a genuine case

What helps, and what harms.

Do this much first

  • Label the slots and the disk group
  • Send every drive from the group
  • Tell us the form factor and the host
  • Say what firmware was applied, and when

What sets us back

  • Resetting the host or the cluster
  • Updating firmware on the failed drives
  • Running the hypervisor's repair on a degraded group
  • Plugging an M.2 boot stick into a laptop to see

Questions answered before you commit.

Can you recover a U.2 NVMe from a server?

Yes, on the PCIe adaptor its form factor needs, at controller level where it must be. £300 + VAT per drive after the free look.

My vSAN disk group is down. What do I send?

Every drive in the group, labelled by slot, and a note of which was the cache. The objects span the drives and are read from the images.

Can you recover a BOSS boot mirror?

Yes. Both M.2 sticks are imaged and the mirror reassembled from the images.

What does it cost?

£300 + VAT for one member after the free look; two to four from £500 + VAT; larger sets from £1,250 + VAT.

How long does it take?

3–4 days at the bench for one drive; 5–10 days at the bench for a group.

Begin here if yours is doing the same thing.

SAS and SATA SSD members →Enterprise SSD members →Several drives failed at once →Dell PowerEdge and PERC →

Nothing gets worse while it is powered down.

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 server off and the drives in their slots.

0800 6890668

Price my jobPacking and posting