Portable drives · bench file · HUL-2025-0319
Slower Every Week, Then It Stopped Showing Up.
Six weeks, beginning to end. The man who looks after the archive for a village between Hornsea and Beverley watched it slide week after week — a minute to come up, then five
, then open a folder and all you get is the spinner
, and finally the light comes on and that is it
. The village's whole record sat on it: minute books scanned from the 1890s onwards, the school log books, photographs collected at open days. He had been advised at some point to defragment the drive and had never bothered, which is the luckiest part of this case.
Sounds like yours? Ring us.
0800 6890668
What causes it.
Slow, here, is arithmetic. Nothing inside is stiffening with age. A sector that needs a second try costs you almost nothing; one that needs fifty costs plenty, and the total climbs as the bad patch spreads. Thumbnails go first, because drawing a preview means reading a whole file end to end, retries and all. None of it is the drive running slower. It is the drive being asked for the same sector again and again. And the day it does not show up at all is that sum performed at mount time: the handshake a computer wants before it will list a drive has come to take longer than the computer will wait.
The tools this one needed.
The steps we follow →| The gear | What this one did | Why we have it |
|---|---|---|
| DeepSpar Disk Imager 4 | Finished the hard passes over SATA, with the timeouts set for the slow head | Grades each head first, reads one surface at a time, and keeps resets, timeouts and power under control |
| DeepSpar USB Imager | Made the opening pass through the drive's own USB bridge, with no operating system in the loop | Normal imaging speed, straight through the USB bridge |
| UFS Explorer Professional Recovery | Put the folder tree back together from the images, then wrote that out to the new disk | Opens the volume formats most tools cannot: APFS, ReFS, XFS, ZFS, Btrfs |
The stages.
Read the numbers first, open nothing
Nothing was opened on arrival. A dying drive has only so many reads left in it, and one spent on a look round is one wasted, so the measurements came first. The reallocated count was high and still climbing, and a head-by-head survey found one of the four answering late, and later on each pass. That head's zone was exactly where the thumbnails had stopped drawing.
Take the good areas while they are still good
Imaging ran in order of confidence rather than sector order. The three good heads worked at full speed and had their zones off before the end of the day, which put the bulk of the archive beyond the fault before the drive was leaned on at all. The limits came off after that — longer waits, resets under supervision — and the passes moved across to the region that six weeks of waiting had been about.
Work the failing head, then rebuild on a good disk
The late head's zone was crossed slowly, halting and power-cycling each time it stalled. What came off went onto a new disk, laid out the way the archive had always been laid out, so it could go straight back into use and the filing carry on from where it stopped.
How it finished.
Losses were small. Back came the minute books and the school log books, and nearly all the open-day photographs. A single batch, scanned in the earliest sessions, lay in the zone the late head served; part of it came back damaged, part not at all; the report lists each one affected. Those minute books can go under a scanner a second time, as the originals are in the parish safe. The photographs were only ever lent for the day and went home to their owners years ago.
What people read after this.
More from the Pocket-size drives file.
Is yours doing the same?
Turn it off. Send it to us. The diagnosis comes first. It says which files are still readable and which are not.