Data Recovery Case File · Desktop Externals & Aging Drives · The Copy Found It
Ordinary Use Touches a Fraction of a Drive and a Copy Touches All of It
Her enquiry describes something going wrong during the one action that was meant to make everything safe. Having removed a drive from a tower herself to save money, she "connected it to my laptop no problem and began to copy the files off. Then something happened and it paused and everything" stopped. The drive was not fine before and broken during — it was already failing, and a full copy is what found the part that no longer reads.
| Media | Internal hard drive removed from a desktop tower — reading normally on initial connection; sequential copy stalling partway; access degraded since |
| Reported situation | Drive removed from a desktop machine by the owner · connected to a laptop successfully · content visible and copying commenced · transfer stalling partway through · access not restored since · contents required |
| Fault class | Read failure encountered during sequential access — fault pre-existing and surfaced by full-surface traversal rather than caused by it |
| Equipment used | Stall interpreted as a pre-existing fault surfaced by sequential access · no further copying permitted · Atola Insight Forensic error-rate assessment across the surface · imaged write-blocked under strict per-sector timeouts with difficult regions deferred · losses mapped per file |
The decode: why copying finds faults that daily use never meets
What ordinary use actually reads: a small and repeated fraction. Opening documents, browsing folders and running applications touches the same regions over and over and leaves the great majority of a drive untouched for months or years at a time.
What a copy reads: everything. Every file, every region, in sequence, at sustained speed — which is the first time in the drive's life that regions nobody has visited since they were written are asked to produce their contents.
So what the stall means: the copy reached a region that will not read. That region was already unreadable and nobody had asked it — the transfer did not create the fault, it arrived at it. This is the single commonest way developing drive faults are discovered.
Why that reframing matters to her: people who lose a drive during a backup conclude the backup caused it, and become reluctant to try again. The opposite is true — the copy is what turned an invisible problem into a visible one, at the moment she was trying to protect herself rather than at some worse moment later.
Why she was reading fine beforehand: because she was reading the easy parts. A drive that opens folders and shows filenames is reading its filing structures, which are small, heavily used and generally in good condition. Content is elsewhere.
What the continued stall does if left running: a file-by-file copy through the operating system inherits its retry behaviour, so each unreadable region consumes the full timeout before it moves on. Hours can be spent on a handful of sectors, all of it sustained work on a surface that is failing.
Why imaging differs and it is not a matter of better software: per-sector timeouts are capped deliberately, difficult regions are deferred rather than allowed to consume the session, and the healthy expanse is taken first at speed. The bad areas are revisited afterwards, when everything else is already safe.
What she did right: removing the drive and connecting it elsewhere was sensible and it cost nothing. The instinct to get the data off was correct — only the method was the ordinary one.
On the bench
The stall was interpreted as a pre-existing fault surfaced by sequential access — ordinary use reading a small repeated fraction of a drive while a full copy traverses every region in sequence, frequently asking areas untouched since they were written. No further copying was permitted, a file-by-file transfer inheriting the operating system's full retry timeout at each failure. Imaging ran write-blocked under strict per-sector timeouts with difficult regions deferred.
The outcome
The stall read as a pre-existing fault, no further copying permitted, and the drive imaged with difficult regions deferred. 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: the copy did not break it. Ordinary use reads a small repeated fraction of a drive, and a full transfer reads all of it — so a migration is where existing faults are found, not where they are made.
Transfer that stalled partway through
Stop the copy rather than letting it grind — a file-by-file transfer inherits the system's full retry timeout at every failure, so hours get spent on a handful of sectors, all of it sustained work on a failing surface. And don't conclude the copy caused this. Ordinary use touches a small repeated fraction of a drive, mostly filing structures that are small and in good condition, while a full transfer reads every region in sequence, including areas nobody has asked for since they were written. So a migration is where an existing fault surfaces. Yours surfaced while you were trying to protect yourself, which is better than later.
Stop it there — call Oxford Data Recovery on 01865 593000; stall read as a pre-existing fault, error rates measured across the surface, imaged under capped timeouts with difficult regions deferred.
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.