Data Recovery Case File · Desktop Drives · The CHKDSK Trap

"Disk unknown, not initialized": the WD Black that CHKDSK fixed on Tuesday and finished on Wednesday

This sequence is so common we could set our watch by it: a single folder throws "corrupted or unreadable", CHKDSK runs and apparently cures it, everything works for a few hours — and the next boot, the whole drive is gone. The customer's clear timeline made this a textbook case file for why that happens.

The enquiry

“I couldn't access one directory — 'the file or directory is corrupted or unreadable' — which seemed odd as it was just a single sub-directory. I ran CHKDSK and it fixed the problem; I could access everything again. I rebooted the next day and now Disk Management says 'disk unknown, not initialized'. The disk still spins, and it does the same in another PC.”

Why the "fix" was the fault talking

A single unreadable folder on an otherwise fine drive is rarely a filesystem accident. It usually means the physical sectors underneath that folder's metadata have started to fail — the corruption is the symptom, and the surface is the disease. CHKDSK, which is a repair tool and not a diagnostic one, responded the only way it knows: by writing — rebuilding structures, moving records, exercising precisely the weakening region and its weakening head. The evening of restored access was the repair holding briefly on degrading media. By morning, the decay had reached something the drive cannot work without, and it stopped presenting itself as a usable disk at all. "Not initialized" on a previously-full drive is not an invitation — initialising writes a fresh, empty structure over the old one, and to Windows' credit it at least asks first.

The recovery

On the PC-3000 the drive's state matched the story: mechanics running, electronics healthy, but reads failing across a concentrated region — a surface problem with a weak head's fingerprints on it, now including territory the filesystem couldn't function without. The drive's background self-maintenance was disabled before anything else; a drive in this condition busily remapping sectors is a drive spending its last strength on paperwork.

Imaging in Data Extractor ran the degraded-drive playbook: the healthy expanse of the surface banked first in fast, gentle sequential passes, the damaged region deferred to the end and then worked patiently — head-aware, tight timeouts, bounded retries — extracting what the failing zone would still yield. From the completed image, the NTFS volume was reconstructed: the drive-wide structures rebuilt from their redundant copies, the file tree recovered essentially intact, and — with a certain symmetry — the losses concentrated in and around the original troublesome folder, whose underlying sectors had been the first casualties all along. The customer received his file set with a precise list of what the dead region had taken.

Outcome

Well north of 99% of the volume recovered and verified, losses itemised rather than estimated. The honest moral isn't "never use CHKDSK" — on a healthy drive with a genuine software-side inconsistency, it's the right tool. The moral is sequence: when corruption appears out of nowhere on a drive full of things you care about, the first move is a copy, not a repair. Repairs write. Writing is the one thing a failing drive can't afford.

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.