Case record · NAS & RAID · SDR-2025-0642
Two Dead Drives in One RAID 5 Array.
A business storage shelf with a familiar injury: two drives failed in the RAID 5
. The IT contractor had a couple of spares
to hand and had already slid one in to see whether the array would just rebuild itself
. It stalled a few per cent along. The trap was half sprung already.
Seeing the same thing yourself?
0800 6890668
The translation.
RAID 5 can survive the loss of one disk and no more. If two disks fail then there is not enough information available to perform a rebuild, and a rebuild started in that state will write new parity across the same stripes that need to be read by the reconstruction. In addition, business shelf arrays present additional problems with vendor sector sizes and proprietary metadata formats that standard desktop based utilities are unable to read. As such, each of these spare disks was eventually put to use doing a proper job, but not the job they had originally been given at the start.
Kit used on this job.
What happens in a case →| Which platform | Its role in this case | Why we use it |
|---|---|---|
| PC-3000 SAS/SCSI | Spoke to the SAS members directly, which no ordinary desktop controller can manage | For SAS and SCSI disks out of business shelves, which normal kit cannot reach |
| Atola TaskForce 2 | Imaged the whole set side by side, turning a week of queueing into days | Images several drives in parallel — it turns a week on an array into a few days |
| UFS Explorer RAID Recovery | Worked out the layout and built the volume back up from the copies | Understands how NAS volume managers really work, instead of a flat array |
In the lab.
Image the lot, healthy members included
The entire set of disks was loaded onto the imagers as a group (with the two dead disks being placed alongside the four good working disks). SAS disks need hardware that can communicate properly with their interface, and the vendor sector format of the shelf had to be handled correctly at this stage. If it hadn't been, then everything built upon that foundation would have been nonsense.
Bring the failed pair back far enough to read
Each of the two dead units was examined and coaxed into a readable state. As is typical, neither unit was completely dead. Most of each of the surfaces was still readable. Having images of both provided the reconstruction process the opportunity to select the better copy of any block instead of having to rely solely on parity for all blocks.
Derive the layout, assemble it in software
For this example, the stripe size, disk order, parity rotation and delay were taken directly from the on-disk metadata instead of being guessed from defaults. Then the volume was put together virtually across the images and the file system read out of that reconstruction.
The result.
After this, the array was brought back and verified, then written onto the company's own spare disks as plain storage, the only practical use the spares were ever going to have. The stripes that the aborted rebuild had overwritten cost part of one project folder.
Similar cases on the index.
Also from NAS & RAID.
Does this sound like your own drive?
This is true in all cases mentioned previously: shut it down, and let a free diagnosis be performed before you make any decisions.