Data Recovery Case File · NAS & Network Storage · Keep the Disks, Not the Box
The Enclosure Is the Replaceable Part
His enquiry asks for something reasonable that turns out to be unnecessary. A network drive "with a damaged shell. Not sure if the data on the drive is still okay — are you able to check to see if the data is okay, and if so supply a new shell for the drive?" The request assumes the enclosure and the data are connected, and mostly they are not. The disks are the storage; the box is a computer that serves them over the network — and while damage to a shell can matter, the route to the data does not run through replacing it.
| Media | Network storage unit with physical damage to the enclosure — member disk condition unestablished; contents required |
| Reported situation | Enclosure physically damaged · condition of contents unknown · owner requesting verification of the data and supply of a replacement enclosure |
| Fault class | Enclosure damage with member condition to be established separately — data readable independently of the unit |
| Equipment used | Members removed, labelled by bay and assessed independently of the enclosure · imaged individually write-blocked (Atola TaskForce 2) · volume assembled offline from images · no initialisation or rebuild permitted in any replacement unit |
The decode: what the box actually does, and what damage means
What a network unit is: a small computer with disk bays. It runs its own operating system, manages the array, and shares the contents over the network. The data lives entirely on the disks, together with the metadata describing how they fit together. So the enclosure is a service that presents the storage rather than part of the storage itself — and the data can be read without it, by imaging the members individually and assembling the volume offline.
Why the shell is not worth replacing for this: checking whether the data is intact does not require a working unit at all, and putting the disks into a replacement carries a real risk that reading them separately does not. A different unit — even the same model — may not recognise an existing array and may offer to initialise the disks or create a new volume so that it can use them. Both write to the members, and both destroy the array. Never put existing disks into a new unit to see whether they work. If the enclosure genuinely needs replacing afterwards, that happens after the data is safely captured, not as the means of getting to it.
When enclosure damage does matter, and it is worth establishing: if whatever damaged the shell also reached the disks. An impact hard enough to break a casing can knock a drive that was running, which is the fault class that needs inspection before power rather than testing. So the useful questions are what happened, whether the unit was powered at the time, and whether it made any noise afterwards. Damage that is purely cosmetic — a cracked panel, a broken mount, a bent bracket — means nothing for the disks.
What to do practically: remove the disks, label them by bay, and keep them together. Bay order determines member order in the volume, and a set of unlabelled disks from a multi-bay unit is a solvable problem that need not have existed. Then have them assessed independently of the unit.
What not to do: not connect a member to a computer, which will offer to initialise a filesystem it cannot read. Not power the damaged unit repeatedly to see if it still works. And not buy a replacement enclosure as the first step.
On the bench
Members were removed, labelled by bay and assessed independently of the enclosure, since the data lives on the disks together with the metadata describing the array and the unit is a service that presents them rather than part of the storage. Each was imaged individually write-blocked on the Atola TaskForce 2, with attention to whether the impact that damaged the shell had also reached a running drive. The volume was assembled offline from the images, and no initialisation or rebuild was permitted in any replacement unit.
The outcome
The disks assessed independently of the damaged unit, imaged individually and the volume assembled offline from the copies. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode, for anyone with a damaged network unit: the box is the replaceable part. Your data and the metadata describing the array live on the disks, so the contents can be read and verified without a working enclosure at all. Do not put the disks into a replacement unit to test them — a new unit may offer to initialise them or create a fresh volume, and both destroy the array. Label them by bay.
Network unit with a broken casing
Don't buy a replacement enclosure and put the disks in it — that's the one thing that can turn a non-problem into a real one. A new unit, even the same model, may not recognise an existing array and may offer to initialise the disks or create a fresh volume so it can use them. Both write to your members and both destroy the array. You don't need a working unit to find out whether your data is intact, because the data and the metadata describing how the disks fit together live on the disks themselves; the box is just a small computer that serves them over the network. Take the disks out, label them by bay so the member order is preserved, keep them together, and don't connect any of them to a PC. Worth mentioning whether the unit was powered when it was damaged.
Don't put the disks in a new box — call Newcastle Data Recovery on 0191 406 1051; members assessed independently of the enclosure, imaged write-blocked, volume assembled offline with no initialisation permitted.
Request a quote online →
Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.