RAID 5 · one disk of parity
These come in most weeks. A member leaves the set, the alert goes unread, a spare is fitted months later, and the rebuild stops on a second disk. We receive sets from all over Plymouth, including engineering companies working into the dockyard, boatyards on the Cattewater and holiday-let offices out in the South Hams. Often the failed units are better than the front panel suggests, since failed is a decision made by a controller, not a diagnosis. What we send back is read off copies, and your original disks remain untouched.
You will be talking to an engineer
0800 6890668
| What turns up | What that points to | Where to start |
|---|---|---|
| The controller says a member is missing | Once the parity is used up, a second failure ends the volume | Power it down before a second fails |
| A rebuild ran half the night and halted | The full read found a soft patch on another disk of the same make and age. That is how these usually end | Leave it powered off. No second attempt |
| Two members read as failed | Typically one finished disk alongside one holding a few sectors it will not give up, rather than two similar disks | Send them all; the images will separate them |
| A rejected member was forced back online | Stale blocks now sit among the current stripes | Touch nothing more |
| A new controller cannot find the set | The dispute is about metadata; the files themselves have not moved | An ordinary reassembly |
| Recovery software has already been pointed at the members | Something has written to your original disks | Send them regardless, and say which tool ran |
Whatever is delivered is booked in using a unique reference (case number) on the date it arrives at the bench. The engineer will identify the actual fault at no charge. This gives a straightforward answer on what can be recovered from the drive and what cannot, plus one written quotation. There will be no further activity until you have read and accepted the quotation.
Each disk is copied in full to our own storage, whether or not the card has marked it as failed. Slow reads are recorded as they occur and the worst areas are left for last. No bytes are written back to the disks you posted.
Three things need to be settled: which bay each disk sat in, how big each stripe is, and how the parity travels around the set. The answers come from what each disk wrote about itself and from how your own data is laid out. None of it is estimated.
The array is built above the images. Each block is taken from whichever image gave the cleanest read, and if no image gave one, parity supplies it: that is how files that lived on a written-off member still come back. After the file system is repaired, the volume is mounted and checked.
Give yourself time to consider it. It costs you nothing. We list all of the files we recover for you before sending out an invoice, and we only ask for payment once you have authorised it. We also buy media specifically for your recovery and send it back to you, with the return postage paid by us. The case stays here on our bench until we have your confirmation that the files open properly on your own machine.
Do the maths once and put it down: makers rate a consumer disk at 1014 bits between unrecoverable read errors, roughly one bad sector in every 12.5TB, and an enterprise disk at 1015. So with a set of four 16TB drives, the rebuild needs about 48TB of perfect reads from the other three. The risk is real and certainly isn't just theoretical. At the same time, disks generally outlast their rating. Both are true, and both prove the same point: make copies first, and the rebuild afterwards costs you nothing.
Disk two appeared to die at the same time as disk four. In reality disk two had left the set in June and no one noticed. Disk four then stopped part way through the rebuild in August. Imaging found 19 bad sectors on the fourth disk, so three good disks and one nearly sound were enough between them to restore the stripe. The year-end filing went in on schedule.
Probably not. Cards reject disks for their own reasons, and a few sectors answering slowly is enough to cause it. A rejected disk and a finished disk are different things. With every member copied to our own storage, whatever one image will not yield is recovered from the parity held by the others. Two disks ejected by a card is fairly common here; two disks that genuinely cannot be read is quite rare.
Not until copies have been made. A rebuild requires each surviving disk to read itself right through, which is exactly what a tired disk fails to do. It also puts new parity over the existing parity that would otherwise restore your volume. Once the images are taken, rebuild the set as many times as you like, because the copies remain as they were.
A URE is a sector the drive's own error correction could not repair, so the drive gave up on it. Manufacturers publish about one per 12.5TB read for a consumer disk and one per 125TB for an enterprise disk. Use these as the floor a manufacturer will stand behind, rather than a clock running down. Modern firmware meets a URE by marking that stripe and carrying on.
No, the members are enough. Each disk in a set holds a description of that set, and where the description is damaged, your own data shows the same thing. Assembly happens in software above the images. What you post is read and nothing more.
Diagnosis costs nothing. We give you a report listing which files can be read, which cannot, and what is needed to get them off. Do not connect the drive until then.