Data Recovery Case File · Solid State & Flash · Present, and Delivering Nothing

Portable SSD Will Not Mount: Present and Delivering Nothing

His enquiry listed three failures in one line, and the list is the diagnosis. A 1TB SanDisk Extreme portable SSD that "won't mount, fails to run First Aid, and can't be scanned by recovery software." The instinct is to read that as three separate problems — a mounting problem, a repair problem and a scanning problem — and it is not. All three are the same thing observed by three different tools: the drive is present enough to be addressed and unable to sustain being read. Each tool tries in its own way, each requires the drive to keep answering for a period, and each fails at the same underlying obstacle. That single finding is more useful than any of the three messages, because it points at where the fault is and, importantly, at what will happen if the attempts continue.

The decode: one finding, three symptoms, and a closing window

Why mounting fails: mounting requires the operating system to read the drive's partition and filesystem structures cleanly and completely. A drive that stalls, resets or returns inconsistent data partway through that read cannot be mounted — not because the structures are necessarily damaged, but because they cannot be read reliably enough to be parsed.

Why First Aid fails rather than reporting a problem: a genuinely informative distinction, and one people misread constantly. A repair utility that runs and reports filesystem errors is telling you about the filesystem. A repair utility that fails to run at all is telling you it could not get far enough to examine anything. That is a hardware-level statement dressed as a software one, and it means the drive is not the patient the tool was designed for.

Why scanning software cannot scan: the same again, at greater length. A full scan is the most sustained read anybody asks of a drive — the entire surface, continuously, for an extended period. A device that cannot hold still for a filesystem parse certainly cannot hold still for that. So the scanner's failure adds no new information; it merely confirms the finding at a larger scale, and at some cost.

The window, and why it is closing: the practical point. A portable SSD in this state typically still responds — it enumerates, it identifies itself, it answers briefly — and that partial responsiveness is a resource. Every mount attempt, every repair run and every scan spends it: powering an unstable controller repeatedly, forcing resets, and in some failure modes provoking further deterioration. The right use of that remaining responsiveness is not another attempt by another tool but one image, taken through hardware that manages the instability rather than inheriting it — timeouts enforced so a stall cannot hang the pass, resets absorbed and the read resumed rather than restarted, and no mount or repair ever requested. Everything afterwards happens on the copy, where the filesystem can be examined and rebuilt at no risk.

If the controller gives out entirely: the honest fallback for SSDs is narrower than for cards and sticks — the mapping and often the encryption live in the controller, so reading the memory chips directly is not the dependable route it is elsewhere. Controller-level access through vendor modes is attempted, it frequently works, and where it does not the limit is stated plainly.

On the bench

No further attempt was made through ordinary tools, since his own testing had established what those can establish. Access ran behind the DeepSpar USB Stabilizer 10Gb, write-blocked, with the inline hardware holding the conversation the drive could no longer sustain — timeouts enforced, resets absorbed, and the device never asked to mount, repair or serve a folder listing. A single prioritised image was taken while the drive still answered, healthy regions first and completely. The filesystem was then reconstructed on the copy, where the structures that would not parse under load could be examined at leisure, and the contents were verified by opening before delivery on fresh media.

The outcome

The drive imaged in one hardware-managed pass while it still responded, the volume rebuilt on the copy and the contents verified and delivered. Free assessment, one fixed written figure including VAT; where a chip has to be removed, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode, for anyone whose drive has defeated three tools: that is one finding rather than three faults — the device is present and cannot sustain being read, and mounting, repairing and scanning all require exactly that; a repair utility that fails to run, rather than reporting errors, is making a hardware statement in software's language; and the partial responsiveness you still have is a resource being spent by every further attempt. Stop, and let it be imaged once through hardware that manages the instability.

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.