NAS & RAID · case record · MHD-2025-7103
RAID Five and a Monday Deadline.
A leisure-centre tender closed at ten on Monday, and the Windsor practice had already done its own triage. The QNAP left the network on Thursday, and by Friday it would not power up: the enclosure has gone, the disks were never at fault
. All four were out and marked with their bay order before anything was touched
. They wanted a fixed price and an honest answer to one question: is Monday realistic
.
Much the same fault? Call us.
0800 6890668
In everyday terms.
There is one honest way to make a weekend of this, and it is to overlap the stages rather than remove any of them. All four members go onto imagers at the same moment instead of queuing, and the geometry of the array is worked out from whichever copies complete first while the remainder are still filling, so no stage sits waiting on another. The shortcut we did not take was assembling the array directly on the practice's own disks. That would have saved a few hours by staking the tender on a faultless read, and had a member stumbled halfway there would have been nothing underneath to catch it.
What the bench used here.
How a job is handled →| Tool | Why we used it | What it brings |
|---|---|---|
| Atola TaskForce 2 | Imaged the four members at the same time, which is what bought back the weekend | Parallel imaging across the whole set, so the copying stage runs in days, not weeks |
| PC-3000 Portable PRO | Interrogated every member for health and firmware state where it sat | Manufacturer-level access from a case that sits beside the imagers |
| UFS Explorer RAID Recovery | Proved member order, stripe width and parity direction against the completed images | Understands the layers a NAS stacks on top of its disks, not just the array |
The lab work.
Four members reading at once, nothing queued
Within an hour of the courier arriving all four members were on imagers and copying, and they stayed there until the final sector. Handing disks to an imager singly is what turns a three-day array into a week-long one. Taking the set together is what separates Friday-to-Monday from Friday-to-an-apology.
Establish the geometry while the images are still filling
Metadata on the first images to complete supplied the candidate layouts: the order of members, the stripe size, and the direction in which parity travels. Each candidate was then measured against genuine directory structures on the volume, rather than accepted on the grounds that most controllers work that way. When the fourth copy finished, the assembly was already written down.
Assemble in software, verify the tender first
Built in software across the four images, the volume presented a filesystem that opened at the first attempt. Verification then followed the practice's order of importance rather than the disk's: the tender set, meaning the drawings, the specification, the cost plan and the appendices, went through first and was confirmed clear by telephone while the remaining folders were still validating.
How the job closed.
The verified files reached the practice before Sunday was out, and on Monday the tender was filed from their own drawings. What made it quick was overlapping the stages and reporting plainly where each of them stood, not omitting any of them.
More of the jobs we've closed.
Other RAID & NAS jobs.
Does any of that match your case?
Nothing needs deciding until the diagnosis is back: switch the device off, send it in, and let the findings settle it.