Data Recovery Case File · Portable Drives · Check the Source

The Drive That Failed While Protecting You

Her enquiry described a failure with a particular irony to it. An external drive that "shut down part way through a backup. Now my Mac cannot recognise the drive or access any files on it." The device whose job was to hold the safety copy failed while making it. Before any technical assessment, that fact raises a question worth answering first, because it frequently makes the whole recovery unnecessary: a backup is a copy of something, and the something is usually still there. The Mac that was being backed up did not fail — the destination did.

The decode: the free question, then why backups kill drives

Ask what the backup was of. If the Mac it was protecting is still working, then nothing has been lost — the safety net has failed, not the data, and the correct response is to get a replacement drive and start backing up again today rather than to commission a recovery. That is obvious written down and it is missed constantly, because a failed backup drive feels like losing everything. It only becomes urgent in specific circumstances: where the source machine has since failed or been replaced, where older material was moved onto the backup and deleted from the Mac to free space, or where the backup contained versions of files that have since changed. Establishing which applies costs nothing and should come first.

Why backups are when drives die: the pattern behind this case and a great many others. Most of the time an external drive sits idle or handles brief transfers. A backup is completely different — sustained writing across large regions of the surface, for a long period, often while the drive sits warm on a desk with poor airflow. It is the most demanding thing the drive ever does, and it is therefore when a marginal drive fails. That is not the backup's fault; it is simply the load that exposes weakness. It does mean, though, that a drive used only for backups gets little exercise between the events most likely to break it.

What an interruption partway leaves behind: a backup that stopped mid-write leaves the destination in an inconsistent state — a partial set with structures mid-update. Even if the drive were readable, that partial backup would be of limited value compared with the live source. Another reason to look at the Mac first.

What the failure mode is likely to be: a drive that stops being recognised at all, having previously worked, has failed below the filesystem — at the mechanism, the board, or the drive's readiness sequence. That is established by assessing it on native equipment rather than by reconnecting it to the Mac, which has already given the only answer a host can.

The practical lesson worth taking: a single backup drive is a single point of failure doing the job of a safety net, and this is exactly how that arrangement ends. Two destinations — a second drive rotated occasionally, or an off-site service alongside the local copy — turn this into an inconvenience rather than an enquiry.

On the bench

The status of the source machine was established before any work was proposed, since a backup drive failing while the original is healthy calls for a replacement drive rather than a recovery. Where the backup was genuinely the only copy of something, the drive was assessed on native equipment rather than through the host — which had already given the only answer a computer can — and imaged under per-sector timeouts where reads proved unreliable. The contents were verified by opening before delivery.

The outcome

The source checked first, the drive assessed on native equipment and the contents recovered where genuinely required. 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, for anyone whose backup drive has failed: ask what it was a backup of, because if the machine it protected still works then nothing has been lost and the answer is a new drive today rather than a recovery; it only becomes urgent if the source has since failed, or if you moved older material onto the backup and deleted it from the source. And note why it happened when it did — a backup is sustained writing across the whole surface, which is the most demanding thing the drive ever does.

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.