Hard drives · health, honestly

Hard drive health: listen first, then look.

A mechanical drive announces trouble through two channels — the sounds it makes and the statistics it keeps. The full check-up costs nothing and takes five minutes: sixty seconds with your ear, then a free tool reading the drive’s own black-box recorder. Here’s both halves, and the responses that don’t make things worse.

Free 48-hour diagnostic
Everything in-house
No fix, no fee · most jobs
// order matters

Ears before software.

A drive that clicks, scrapes or repeatedly spins up and sighs back down has already failed the health check — no download required, and running one anyway only adds mileage to failing mechanics.

Step 1
The 60-second listen
Step 2
Read the SMART log
Judge by
Sectors & trend
Never by
A percentage
// step one

The sixty-second listen.

In a quiet room, power the machine on and pay attention through the first minute — the spin-up is where mechanical truth comes out. A healthy drive gives a smooth rising hum and settles into a whisper with soft, irregular ticks under load. The vocabulary of trouble is small and unmistakable: rhythmic clicking (heads sweeping and failing to find their bearings), scraping or grinding (contact where none should exist — power off now), beeping or whining from a drive (a motor fighting to spin), and the spin-up-sigh-silence cycle of a drive giving up and retrying. Any of these overrules everything a utility will tell you afterwards: the drive has declared itself, further running is wear, and the sensible destination is a clicking-drive bench rather than a download page.

// step two

Read the black box — and judge it like an engineer.

Quiet drive? Now open the log. CrystalDiskInfo on Windows, Disk Utility’s SMART line or smartctl on a Mac, or the maker’s tool — SeaTools for Seagate, WD’s Dashboard — all read the same self-maintained ledger. Skip the headline verdict and go to three sector counters: Reallocated Sectors (patches of disk that died and were swapped for spares), Pending Sectors (areas the drive currently can’t read — quite possibly your files), and Uncorrectable Sectors (reads that failed after every retry). Healthy is zero across all three; a small static number is a scar; and movement is the verdict that matters — a count that grows between checks is a drive scheduling its own funeral. Spin Retry deserves a glance too: any value above zero says the motor itself has struggled, which is rarer and graver than a bad sector. Age and temperature are context, not conclusions — and on external drives, note that some USB enclosures block SMART entirely, so ‘no data’ through a caddy means unreadable, not healthy.

// the wrong responses

Warnings are for evacuating, not for examining harder.

Faced with a caution, instinct reaches for more tests — a full surface scan, CHKDSK, the vendor’s ‘repair’ mode. Understand what those are: hours of sustained, punishing reads (and in CHKDSK’s case, writes) applied to a machine that just reported an injury. CHKDSK in particular is routinely mistaken for a health check when it’s a file-system correction tool — it can ‘complete successfully’ on a dying drive while adding exactly the load that finishes it. The professional order is the reverse: evacuate first — copy the folders you can’t replace onto other storage while the drive still cooperates, most precious first — and let diagnosis wait until the data is safe. If the evacuation itself stumbles — copies hanging, the machine freezing, any new sound — that’s the drive’s last message: power down and hand it to the hard drive recovery bench, where imaging hardware reads failing drives on gentler terms than any operating system.

// questions

Asked before you ask, answered.

Twice a year for a drive you rely on, plus after any event — a knock, a house move, an unexplained freeze. The check takes two minutes, so frequency isn’t the burden; remembering is. Pair it with something you already do on a schedule, and screenshot the table each time so you’re comparing, not guessing.

No — and be wary of anything claiming otherwise. Reallocated and damaged sectors are permanent scar tissue; ‘repair’ utilities merely hide bad areas from the file system, and a drive that has started reallocating is on a one-way road. The professional response is recovery and retirement: get the data onto new hardware, and let the patient rest.

Nobody honest will give you a number. Some drives limp for months on a caution; some are gone by the weekend, and the deciding factor is usually workload — every hour of use is another spin of the wheel. The practical translation of ‘caution’ is not ‘monitor it’ but ‘copy the irreplaceable things off today, while reading is still cheap’.

// heard something? seen a caution?

Two minutes on the phone beats another scan.

Describe the sound or send the SMART screenshot — a Bristol engineer will tell you honestly what it means. Free 48-hour diagnostic, written quote before any work.

Call us — 0117 332 1137
Mon–Fri · 9am–5:30pm · No fix, no fee
Start a free diagnostic →