Devices · RAID arrays & servers
Most arrays survive the first dead disk. What kills them comes next: a RAID 5 rebuild grinding a tired survivor into the ground, a member forced back online, drives swapped between bays. RAID 0 has no cushion at all — one failure cuts into every file. We work from images alone, so the harm stops the moment your set, bare disks labelled by bay, arrives from Exeter.
Talk it over with an engineer
0800 6890668
| What it says | What it usually means | Step one |
|---|---|---|
| Foreign Configuration Found (Dell PERC) | The card's record and the disks' own metadata no longer match | Neither Import nor Clear — image |
| 1786 — Drive Array Recovery Needed (HP) | Parity cover is gone — a rebuild is waiting, or quit half-done | Shut down, image after |
| 1784 — Drive Array Drive Failure (HP) | A member disk has failed | Don't swap drives on spec |
| 1788 — Drives Improperly Attached (HP) | The drives are in the wrong slots | Order matters — stop |
| Virtual Drive is Degraded / Offline | The volume has lost redundancy, or dropped out entirely | Power the set down |
| 1720 — SMART drive detects imminent failure (HP) | A working drive is telling you it won't last | Get it imaged today |
Your device is logged under its own case number the day it arrives. An engineer finds the fault, tells you plainly what stands a real chance of coming back, and writes you one fixed quote. Nothing is charged for that look, nothing obliges you, and paid work starts only when you say so.
Every drive — the ones the card rejected included — is read on purpose-built imaging rigs. All later work happens on those copies, and not one write ever goes back to your originals.
Stripe width, drive order and the way parity rotates are worked out from patterns in the data itself. We then stitch the set together across the images on lab machines — your controller stays out of it, and no rebuild is ever run.
On that rebuilt set we put the file system, VM containers and databases straight, then check the lot against a full listing before any of it leaves the lab.
Before any fee falls due you see a full listing of everything recovered and approve it. Your files travel back on brand-new media, return postage on us, and the case stays open until you've confirmed everything opens on your own machine.
Why RAID 5 gets its bad press: consumer SATA drives are rated at roughly one unreadable sector in 1014 bits read — call it one every 12.5TB. Rebuild a big single-parity set and every survivor has to hand back more data than that in one continuous read, which is exactly how rebuilds die. The figure sits on the spec sheet, not in folklore — though how much it matters in practice remains a live argument.
Mid-rebuild on a Friday, a second drive gave out and the server's production data was stuck. Every one of the four disks was imaged, the stripe rebuilt from three sound images and part of a fourth, and by Monday the factory was back at work.
Yes — every drive should come in. Even one flagged 'failed' may hold the freshest copy of some stripes, and when we rebuild we choose the best-read source for each block across the whole set.
No. Disk order, parity rotation and block size can all be recovered from the patterns sitting in the data. That's routine analysis, not guesswork.
Yes. Once the array is back together, the VMDK and VHDX files and database stores go to the top of the list — and each is tested to mount, not just ticked off as present.
Multi-drive jobs usually take four to seven working days. If the outage has truly stopped the business, the work can run without pause — say so when you ring and the intake gets shaped to match.
Each power-up of a failing device takes more data with it. Open a case first — the look-over is free whichever way it goes.