RAID Recovery Case Study
The wrong drive was pulled. Every file still came back.
A three-drive RAID 5 server went down mid-repair, when a technician replacing a failed disk removed a healthy one instead. Our laboratory reconstructed the array and recovered the client's data in full.
The incident
A routine drive swap that took the whole server down.
The client's in-house IT technician had spotted a failing Maxtor drive in a three-disk RAID 5 server and set out to replace it — a straightforward job that went wrong at the last step.
Instead of removing the failed Maxtor drive, the technician pulled one of the two remaining healthy disks. With two drives now out of the array, the server crashed immediately. Putting the healthy drive back in did not bring the server back up — a RAID 5 array can tolerate one missing member, not two, and the array itself was now unreadable.
The client brought all three drives to our Belfast laboratory for SATA RAID recovery, using our same-day service to get the business back online as quickly as possible.
In the laboratory
Rebuilding the array from three individual drives.
- 01
Individual drive analysis
Each of the three drives was examined separately to establish its condition and how current its data was, before any attempt was made to rejoin them.
- 02
Repair and imaging
One drive — the originally failed Maxtor unit — had developed bad sectors. Our engineers carried out the necessary repairs and were able to read 99.8% of that drive. The remaining two drives were healthy, and 100% of their data was imaged.
- 03
RAID controller emulation
With clean images of the current data from all three drives, our engineers used custom RAID controller emulation to string them back together in their original configuration.
- 04
Full array reconstruction
The complete RAID 5 array was reconstructed from the three images, restoring the volume and file structure exactly as it existed on the server.
Outcome
100% of the client's data, fully functional.
Despite one of the three drives being only 99.8% readable, the redundancy built into RAID 5 meant nothing was actually lost. Every file on the array was recovered and returned in fully working order.
Because RAID recovery of this kind involves analysing, repairing and reconstructing multiple drives rather than one, the work came to over £650. The client chose our same-day service specifically to minimise how long the server stayed down.
Stop touching the array and get the drives to us untouched. Most RAID failures — including this kind of human error — are fully recoverable when the remaining drives aren't pushed further.