Case record · NAS & RAID · SDR-2025-1103
RAID Five, Racing the Clock.
The telephone call was all logistics: RAID 5 on a four-bay NAS
, disks already removed from the unit, and we can drive them over this afternoon
. They needed their tender to be in by Monday morning. Before committing, they wanted a straight answer on cost and how long
, with no sales pitch.
Seeing the same thing yourself?
0800 6890668
The translation.
While a deadline may establish your timeline for completion, it will not affect how you complete the task; it simply determines when you will finish. The imaging stage stays the same: rebuilding an array directly from the customer's original disks to save a few hours puts the array at risk, and if a disk fails partway through there is no second run. In addition to saving time, imaging each disk at once versus sequentially allows us to maintain momentum on a job against a deadline, because we can work out the configuration of the data while all of the imaging takes place so that no work sits idle.
Kit used on this job.
What happens in a case →| Which platform | Its role in this case | Why we use it |
|---|---|---|
| Atola TaskForce 2 | Every member imaged at the same time — the largest single saving on offer | Images several drives in parallel — it turns a week on an array into a few days |
| PC-3000 Portable PRO | Firmware-level work taken to the disks instead of the disks to the bench | Identical firmware-level reach, in a box that goes to the disks |
| UFS Explorer RAID Recovery | Worked out stripe size, disk order and parity rotation from the copies | Understands how NAS volume managers really work, instead of a flat array |
In the lab.
All of them on imagers together
Each member disk went onto imaging equipment as soon as it arrived. Imaging a multi-disk array one disk at a time is what stretches three days into a week. That is why if you have enough imagers to work on each disk simultaneously the Friday deadline will be made instead of being missed.
Work out the geometry as the copies run
The array's parameters came off metadata in the first images to finish, rather than waiting on the rest — stripe size, disk order, parity rotation and delay, each one checked against real data patterns instead of trusted to whatever the controller normally does.
Rebuild it, then check the things the firm needs soonest
The virtual volume was put together using the images taken and the file system in that virtual volume was read. Then validation of the virtual volume occurred with priority based upon what our customer told us their needs were. We validated the shares that they needed for Monday first, so we could let them know those were safe before the rest of the validation was completed.
The result.
The new volume passed validation and returned that weekend. On Monday this practice ran on their own data. It was completed in a very short amount of time (with no steps being skipped to manage it), and status was reported at the completion of each step rather than promised at the beginning.
Similar cases on the index.
Also from NAS & RAID.
Does this sound like your own drive?
Like normal, the rule is identical to the others from our previous examples. Turn it off first. Then let the free diagnosis come before you make any decisions.