Multiple member failures, a controller that died, a rebuild started on the wrong disk, a NAS that will not mount its volume. We image every member independently, work out the real geometry from the data rather than trusting the controller, and rebuild the array virtually.
An array almost never fails all at once. It usually fails twice: the original fault, then the recovery attempt made on top of it.
A member dropped out weeks ago without anyone noticing, and the array only stopped when the next one went.
A rebuild started against a stale member writes parity over live data. Stop the controller before it finishes.
The disks are healthy but the metadata that describes the array is gone with the card.
Disk order, stripe size, parity rotation and offset are worked out from the data itself, not from labels on the tray.
QNAP, Synology and Netgear boxes running mdadm, LVM and Btrfs or ext4 on top, with the volume marked inactive.
VMFS datastores, deleted LUNs and thin-provisioned VMDK or VHDX files that the hypervisor no longer sees.
Evaluation and quotation are free. You pay only after we show you a verified list of what was recovered, and only if you approve it. Nothing recovered means nothing invoiced.
Every case follows the same audited path. You get a case ID at step one and can ask for status at any point.
An engineer, not a call centre, works out what failed and whether the media should be powered on again at all.
We collect the device from your location, logged and sealed, or you send it in by courier.
A full diagnostic report: the fault, the recoverable scope, the turnaround time and a fixed price.
Sector-level imaging in a Class 100 environment, then reconstruction from the image. The original is never written to.
Online or on site, you check the recovered files before any invoice is raised.
Data is delivered on fresh encrypted media. Our working copies are sanitised to standard and the NDA stands permanently.
Members are imaged read-only and the array is assembled virtually from the images, so a wrong assumption costs a retry rather than the data.
The first hour decides how much is recoverable.
Not before the members are imaged. A rebuild writes to the array, and if it uses a stale member or the wrong order, it overwrites the data you are trying to get back. Power the array down and call first.
It matters that we know the original order, so photograph the tray before removing anything. If the order is already lost, we can determine it from the stripe pattern in the data, it just adds analysis time.
Yes. The array lives on the disks, not in the enclosure. We pull the members, image them and assemble the volume in the lab, so a dead QNAP or Synology chassis is not a problem.
It is the most common case we see. A degraded array has no redundancy left, so the next fault takes the volume down. Get the data imaged first, then rebuild the hardware.
Free consultation, free pickup, a fixed quote before any work starts, and an invoice only if you get your data back.
Data Recovery Lab