RAID arrays are built for redundancy, but multiple failures, a bad rebuild, or a controller fault can still take the whole array down. We reconstruct RAID arrays of every level — priced per case after a free evaluation.
A RAID array spreads data across several drives according to its level. RAID 0 stripes data with no redundancy, RAID 1 mirrors it, RAID 5 and RAID 6 stripe data with one or two sets of parity so the array can survive one or two drive failures, and RAID 10 combines mirroring and striping. Every array also depends on configuration details such as drive order, stripe size, parity rotation, and start offset, stored as metadata on the drives or the controller. Recovery depends on getting those details exactly right.
Arrays usually fail in stages. One drive drops out and the array keeps running degraded, often unnoticed, until a second drive fails. Rebuilds are especially risky: a rebuild reads every sector of the remaining drives, and a weak drive can fail partway through. Rebuilding onto the wrong drive, re-creating the array with different settings, or initializing it can overwrite parity or data, which makes recovery harder and sometimes impossible.
Because an array is only as recoverable as its weakest members, each drive is handled individually first. The array is then reassembled virtually from copies of the drives, never by writing to the originals, so a wrong guess about the configuration costs time rather than data. This is also why the original drives should arrive exactly as they were when the array failed.
What we recover
✓RAID 0, 1, 5, 6, 10 and nested levels
✓Hardware and software RAID
✓Rebuilt-gone-wrong and degraded arrays
✓Business file shares, VMs, and databases
Common failures we handle
•Two or more drives failed
•A rebuild failed or made things worse
•Controller or backplane failure
•Array won’t mount or shows as foreign
Before you send it in: what not to do
✕Don't start a rebuild or force a failed drive back online when more than one drive is affected — rebuilding from stale or failing drives can overwrite good parity.
✕Don't re-create or initialize the array to make it mount — creating a new array writes new metadata and can destroy the configuration recovery depends on.
✕Don't clear the foreign configuration or reset the RAID controller — that can erase the record of how the array was built.
✕Don't run file system repair on a degraded or partly rebuilt volume — repairs written to a misassembled array can corrupt data across every drive.
✕Don't remove drives without labeling their slot positions — drive order is essential to reconstructing the array.
How we recover it
1
Document and evaluate
During the free evaluation we record each drive's slot position and the controller configuration, and assess the health of every member drive.
2
Image every member
Each drive is cloned individually, with failing drives repaired in the cleanroom first, so no work is done on the original disks.
3
Determine the geometry
From the images, we work out the RAID level, drive order, stripe size, parity layout, and offsets, using controller metadata where it is intact.
4
Reconstruct virtually
We rebuild the array in software from the images, using parity to fill in data from failed or missing drives where the RAID level allows.
5
Extract and verify
We mount the reconstructed volume, recover file shares, virtual machines, and databases, and check critical files before returning them.
Multi-drive cases are quoted per job after a free evaluation — the price depends on the configuration, the failure, and the work involved. There's no charge for the evaluation, and you pay nothing unless we recover your data.
Do I need to send all the drives, including the ones that still work?+
Yes. Every member drive, plus any hot spares, should be sent together and labeled with its slot position. Data and parity are spread across all of them, so a healthy drive is just as important to the reconstruction as a failed one.
Should I send the RAID controller too?+
Usually the drives are enough, because arrays are reconstructed from drive images rather than through the original controller. Noting the controller's make and model, and any configuration details you have, can still help with unusual layouts. We will tell you during the free evaluation if the controller itself would help.
Can you recover a RAID 5 with two failed drives?+
Often, but not through parity alone, since RAID 5 can only rebuild one missing drive. Recovery usually means repairing at least one failed drive well enough to image it. If one drive dropped out long before the other, its data may be out of date, which has to be accounted for during reconstruction.
Is my data still recoverable?+
In most cases, yes. A device that won't boot or mount usually has a hardware fault while the data itself is intact. The key is to stop using it — don't reformat, rebuild, or keep powering it on — and let our cleanroom handle it.
How do I send in a RAID system?+
Start online for a free evaluation and we'll arrange secure, insured shipping. Your system is assessed in our cleanroom, you approve a fixed quote before any work, and you only pay on a successful recovery.
How long does it take?+
Most cases are evaluated within a couple of business days of arrival, with a turnaround estimate provided alongside your quote. Emergency service is available when time is critical.