Data Recovery Case File · Desktop Externals · The Telemetry Enquiry

Worked seven times harder than its class: a nine-year backup drive's own statistics, decoded — and the selective recovery its owner designed

This enquiry arrived as a dataset. The patient: a nine-year-old Seagate Backup Plus Desk 2TB, now clicking and screeching, visible in Explorer with a drive letter but no name and no access, failed on both of the vendor tool's read tests. And then the numbers he'd pulled before writing: lifetime bytes read 425TB; lifetime bytes written 805TB; power-on hours 8,087; annualised workload rate 1,333TB per year. Partly used, he noted, as an automated backup destination — and his request was correspondingly precise: the drive is almost full, but much of it is duplicated elsewhere, so "I'll need to decide on a selective recovery depending on what's available once a file list is possible." A customer who reads his drive's own telemetry and designs his recovery around a file listing deserves both things: the decode his numbers contain, and the workflow he asked for. Delivered, in that order.

The decode his numbers deserve: drives are rated in terabytes per year, and nobody reads the rating

Here is the specification almost no owner ever meets: hard drives carry a workload rating — the total data traffic, read plus write, they're engineered to handle annually. Desktop-class drives like his are rated in the region of 55–180TB per year depending on generation; heavier-duty NAS and enterprise families climb to 180–550TB. His annualised figure — 1,333TB/yr — sat at roughly seven times a desktop drive's engineering envelope, and the composition explains how: 805TB written against 425TB read is the signature of an automated backup target, a machine whose scheduled jobs pour data in relentlessly, year after year, at volumes the owner never watches because automation's whole point is not watching. Meanwhile his 8,087 power-on hours — under a year of cumulative runtime across nine calendar years — completes the decode's punchline: this drive wasn't old in hours; it was old in tonnage. Wear tracks work, not birthdays, and a drive can be gently used by the clock while being worked to death by the terabyte. The clicking and screeching, the failed vendor tests, the lettered-but-nameless limbo in Explorer: a mechanism that had honestly delivered seven lifetimes of duty, reporting completion. His telemetry didn't just describe the failure. It explained it, to the decimal.

The workflow he asked for — list first, choose second, extract third

His selective-recovery request gets the concrete answer it deserves, including the honest economics. The sequence: the drive was made readable the only way its state allowed — donor head-stack under laminar airflow, firmware secured on the PC-3000 — and then imaged in full regardless of the selection to come, because stabilising and reading a wounded mechanism is one careful pass whether you want ten files or two million; the surgery is the cost's core, and honest quoting says so up front rather than implying selection halves the bill. What selection genuinely buys: from the completed image, a full recovered-file listing — every path, name, size and date — went to him as a browsable document, and he did exactly what his enquiry planned: cross-checked it against his other backups and marked the population that existed nowhere else, a few hundred gigabytes of the almost-full 2TB. That selection was extracted, verified by opening files across it, and delivered on new media — with the remainder held on the image briefly against second thoughts, then securely disposed of on his confirmation. Time saved, media right-sized, verification focused where value lived: the real dividends of choosing, honestly located.

Outcome

The unique data home, the duplicated bulk consciously left behind, and a case that upgrades two pieces of public knowledge. For every automation owner: find your backup drive's workload, not just its age — the telemetry he supplied is readable on most drives with free tools, and a destination absorbing multiples of its class rating is scheduled for exactly his ending; rotate the role across drives, or buy the workload class the job actually demands. And for every enquirer who's ever wondered: yes, file-list-first is a service you can ask for by name — the listing is producible once imaging completes, the choosing is yours, and a lab that quotes honestly will tell you, as this page just has, which part of the bill selection trims and which part the drive's condition already wrote.

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.