Data Recovery Case File · Desktop Externals & Aging Drives · Image It Before It Joins the Other
The Drive That Still Works Is the One With a Deadline
His enquiry concerns equipment older than most of the people who will read this. Two drives from a machine running an operating system from the early 1990s: "one drive I can access and is functional, but due to bad sectors I am unable to clone. The other drive is inaccessible, lots of noise when it runs." The instinct is to worry about the noisy one — and the urgent one is the drive that still reads, because that will not last and the other is not getting worse in a drawer.
| Media | Two hard drives from a machine running a legacy operating system — one accessible with unreadable regions preventing cloning; one inaccessible with abnormal acoustics |
| Reported situation | Two drives from a machine running an operating system of early-1990s vintage · first drive accessible and functional · cloning of that drive failing due to unreadable regions · second drive inaccessible · second drive producing abnormal noise when running · access to both required |
| Fault class | One degrading but readable drive and one mechanical failure — capture priority determined by remaining readability rather than by severity |
| Equipment used | Readable drive prioritised and imaged first with unreadable sectors logged rather than substituted · period-appropriate interface adapters identified before any handling · imaging under strict per-sector timeouts with multiple passes over weak regions · second drive inspected under laminar flow bench before any spin attempt · each assessed independently |
The decode: why the working drive goes first, and why cloning fails on it
Why the priority inverts: the failed drive is stable. It is not deteriorating in storage, and whatever is wrong with it will be exactly as wrong next month. The working drive is deteriorating while it is being used, and every access it gives is one of a shrinking number.
Why the noisy drive nevertheless feels more urgent: because it is obviously broken and the other is obviously not. Attention follows the visible fault, and in this case that is the wrong way round.
Why his cloning keeps failing: cloning copies data sector by sector, and where a sector cannot be read the tool must put something in its place, because the destination has to remain positionally identical. Many tools stop entirely at that point, and those that continue write a substitute silently.
What is needed instead: imaging that logs unreadable regions rather than stopping or silently filling them, with configurable timeouts so a difficult sector is given a short budget and deferred. The healthy expanse is captured at speed first and weak areas revisited on later passes, which recovers substantially more from a degrading drive than any cloning tool can.
Why knowing what could not be read matters as much as the data: a logged list tells him whether the gaps fall in material that matters. A silent clone gives no such information, which is why his attempts have produced nothing usable.
What the age raises before anything is dismantled: the interface. Drives of that period use connections no current machine provides, so the correct adapter is identified in advance rather than discovered — and drives of that vintage are also physically larger and heavier than modern equivalents.
What the context implies, since nobody keeps a machine that old for sentiment: it is almost certainly running something that cannot be replaced — controlling equipment, or holding records in a format nothing current reads. Which raises the same question as any legacy installation: what does this depend on that is not on the disk? A licence, an application nobody sells, a peripheral with no modern driver.
What must not happen: no further cloning attempts on the working drive. Each is a full-surface read on a drive with a diminishing number of them, spent on a method that cannot succeed.
On the bench
The readable drive was prioritised and imaged first with unreadable sectors logged rather than substituted — a failed drive being stable in storage while a degrading readable one deteriorates with every access, which inverts the intuitive priority. Cloning tools stop or silently substitute where a sector cannot be read, because the destination must remain positionally identical. Period-appropriate interface adapters were identified before any handling, and imaging ran under strict per-sector timeouts with multiple passes over weak regions.
The outcome
The readable drive imaged first with failures logged, adapters identified before handling, and each drive assessed independently. Free assessment, one fixed written figure including VAT per drive, 50% of parts and labour upfront with the balance only on successful recovery. The decode: your working drive is the urgent one. The failed drive is stable in a drawer and will be no worse next month, while the readable one is deteriorating with every access — and cloning cannot succeed on it because a clone must write something where a read fails.
One drive failing and one already failed
Image the working one first — that inverts the instinct and it's the right order. A failed drive is stable in storage and will be exactly as broken next month, while a drive that still reads is deteriorating every time you use it, and each access is one of a shrinking number. Stop trying to clone it, too: cloning must put something in every sector it can't read, because the destination has to stay positionally identical, so tools either halt or substitute silently. What you want is imaging that logs what it couldn't read, with short timeouts so difficult sectors are deferred and the healthy expanse is captured first.
The working one is urgent — call Oxford Data Recovery on 01865 593000; readable drive imaged first with unreadable sectors logged, period adapters identified before handling, each drive assessed independently.
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.