Data Recovery Case File · Milestone · Backups Are Not Binary
A Sector Every Five Seconds: the Backup With One Gap
The eight hundred and fiftieth case in this archive is not about someone who failed to back up. It is about someone who did — carefully, and almost completely. A 2TB Seagate: "Windows notified me of drive issues and prompted a restart, during which it began a check disk process. That was going well, then the process ground to a halt, scanning a sector every 5 seconds and estimating days to finish." Then the sentence that makes this a milestone: "While most of the data on the drive is expendable and/or backed up, it contains my PhD thesis in its current state, which doesn't have a backup." A deadline is approaching. And during disconnection, the drive's SATA connector snapped, its plastic lodging in the cable. Four things went wrong. The one worth eight hundred and fifty case files is the third.
| Media | 2TB Seagate hard drive — reported as faulty by the operating system; filesystem check stalled at approximately one sector per five seconds; SATA connector subsequently fractured during disconnection with contacts remaining attached |
| Reported situation | Drive faults reported and a check disk begun automatically · process degraded to a crawl with a multi-day estimate · doctoral thesis in current form held without backup while remaining data was backed up or expendable · academic deadline imminent · physical connector damage sustained during removal |
| Fault class | Progressive media degradation compounded by a write-based repair attempt and by physical interface damage — two faults, one original and one caused by the rescue |
| Equipment used | Connector footprint repaired at board level before any read attempt · Atola Insight Forensic surface and error-rate assessment · PC-3000 Express with Data Extractor imaging under per-sector timeouts, sequenced to capture the thesis first · weak regions revisited on later passes · written timescale issued against the deadline |
The decode: the crawl, the connector, and the gap
A sector every five seconds is the diagnosis: he watched his drive fail in real time and the tool reported it as progress. A healthy drive reads thousands of sectors per second; one managing a single sector every five is spending essentially all of its time retrying reads that never complete. The multi-day estimate was not a big job being worked through — it was a failing drive being waited on, with the remaining time recalculated from a rate that reflected the failure. Any repair utility quoting days is describing a drive that cannot read, not a task that is large. Worse, a check disk writes: it was attempting to repair a filesystem on a drive that could not reliably return data, marking sectors, relocating what it could and modifying structures throughout — a sustained write-and-retry session on precisely the weakest ground. Stopping it was the right call. Starting it was never his decision; the system began it automatically after prompting a restart, which is worth knowing, because that is how most people meet this.
The connector was caused by the rescue: and it deserves saying without blame. SATA connectors are brittle plastic, drives that have been in service for years develop tight fits, and a connector removed at any awkward angle can snap — leaving the plastic in the cable with the drive's contacts still in place. That last detail is the fortunate one: contacts still attached means the footprint is repairable rather than torn away. But it means the drive now carries two faults, one original and one created while trying to fix the first. That pattern runs through this entire archive: the second injury is almost always inflicted during the rescue.
And the gap, which is the milestone. Read his sentence again: most of the data is expendable and/or backed up. He has a working backup regime, and a genuinely good one — he can distinguish, in a single clause, between what does not matter, what is already safe, and the one thing that is neither. That precision is rarer than any technical knowledge and it is the whole point of this page. Backups are not binary. The question is never whether you have one; it is where the gap is. And the gap is always the same shape: the thing being actively worked on. The file that is different every day. The version newer than the last sync. A thesis in its current state is the definitive example — the one document that changes continuously, that is open right now, and for which a weekly or nightly copy is structurally the wrong instrument. Periodic backup protects things that sit still. Work in progress needs continuous protection: a syncing folder, a versioned cloud store, something that captures the file as it changes rather than as it was. That is a different tool for a different problem, and it is the one thing that would have made this a footnote instead of an emergency.
The outcome — and the volume closes
The fractured connector footprint was rebuilt at board level before the drive was asked to read anything, because a damaged interface on a failing drive gives you one attempt and no second chance. The Atola Insight Forensic then mapped surface condition and error rates, converting his five-seconds-per-sector experience into a measured picture of where the drive still read. Imaging ran on the PC-3000 Express under Data Extractor with per-sector timeouts enforced — the decisive difference from the check disk's unlimited retries — and, crucially, sequenced by his priority: the filesystem structures read first so that the thesis could be located, then its regions swept before anything else, then the remainder. It was verified by opening and delivered against a written timescale, ahead of his deadline. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour upfront with the balance only on success.
And with case 850 this volume closes, and site four passes its first hundred. The milestones behind it trace one argument, refined each time. At 550, a repair shop that knew not to touch a failing drive. At 600, two shops that tried, met their ceiling and pointed onward. At 650, institutions each doing their job correctly while the data question belonged to nobody. At 700, an executor with every legal right and no way to open a single file. At 750, the honest no. At 800, the customer who did everything right. And here, at 850, the one that completes it — because after eight hundred and fifty case files, the advice was never back up. Everybody has heard that, and most people do something. The useful advice is know where your gap is. He knew. He could name it in one sentence, under pressure, at the worst possible moment: everything else is expendable or safe, and this one file is neither. That is what a real backup regime buys — not immunity, which nothing buys, but a gap small enough to name. And a gap small enough to name is a gap small enough to close, on the day you notice it rather than the day the drive stops.
Working out where your own gap is
Assume you will lose a drive and ask one question: what would I be unable to replace? Not what's valuable — what's unrecoverable from anywhere else. For most people the answer is short, and it's almost always the same shape: the thing being worked on right now. Photographs already uploaded are safe; installed software is replaceable; downloaded media is expendable. The thesis, the manuscript, the client project, this week's accounts — those are the gap, because they change daily and a periodic backup always lags behind them. That's the crucial distinction: scheduled backups protect things that sit still, and work in progress needs continuous protection instead — a syncing folder or a versioned cloud store that captures the file as you change it. And if a repair tool ever starts crawling, stop it: any check quoting days is telling you the drive can't read, not that the job is big.
Name it when you call — call Glasgow Data Recovery on 0141 404 0294; imaged under per-sector timeouts and sequenced to capture your priority first, with a written timescale against your deadline.
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.