Three Drives, One of Them SCSI

The decode

Drives out of network boxes carry a trap that catches careful people. Network enclosures format their disks in Linux-family filesystems a desktop does not read. Plugged into a Windows machine, such a drive looks blank or "needs initialising" — and accepting that helpful prompt writes fresh structures over the old volume. The absence of visible files on a network-box drive means nothing until the right tools have looked; the initialise prompt means everything and must always be declined.

The non-spinner divides into two families, told apart in seconds under controlled power. Silence with no current draw is electronics — a failed board, the calibration-chip work that goes with modern boards. A twitch, hum or buzz against stall current is mechanical — heads stuck to a platter or a seized bearing, which is head-swap bench work with donors on standby. Either way, a drive that never spins has usually protected its data by refusing to run; non-spinners image surprisingly well once the specific obstacle is removed.

SCSI is an appointment with the bench's older toolbox, not a problem. The interface predates today's plug-and-play, with its own connectors, addressing and termination rules — and a laboratory that has run long enough keeps those host adapters alive precisely for drawers like this one. Age shifts the risk profile more than the interface does: older mechanisms deserve gentler first spins and imaging that assumes fragility, because a drive's first power-up in fifteen years should also be treated as potentially its last.

The unremembered third drive is the easiest to promise things about. Identification is what intake is for — make, capacity, interface and origin are read off the drive itself, and it joins whichever family it belongs to. Not remembering is normal; the drawer remembers for you.

One report, three verdicts and honest triage across them. Mixed intakes are evaluated individually and reported together: what each drive holds, what each recovery will cost and, because network-box pairs sometimes mirrored each other, whether two of the drives turn out to hold one set of data, which halves a bill nobody wants doubled.

On the bench

Each drive was identified and, before being connected to any host, had its spin-up current profiled on a bench power supply. The non-spinning drive went to PC-3000 UDMA, where its fault was localised, with donor parts and head-swap bench work held ready. The SCSI unit was run through the bench's legacy host interface and imaged gently under a Tableau write-blocker. The third drive was identified on arrival and handled according to its family. The network-box volumes were mounted from the images in UFS Explorer, the data extracted per drive and verified by hashing, and any overlap between the drives was reported honestly.

The outcome

Three independent diagnoses, imaging matched to each drive's age and failure mode, the network box file systems read from their images, and all results contained within one report. A free assessment, one fixed written quote, plus VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the remaining amount due only if we recover your data. The decode: a drawer of old drives represents several individual patients, not a group. The older SCSI drive communicates using a protocol the bench still supports. A non-spinning drive has likely protected itself. And although the network box pair may appear to be two separate recoveries, they could turn out to be one recovery under two labels.

Our case files are written up from genuine enquiries our lab has handled for customers across Sheffield and South Yorkshire, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery approach our engineers apply to that fault, using the equipment listed.