Data Recovery Case File · NAS & Network Storage · A Clone Is Not an Image

Copying a Failing Drive Onto Another One Is the Wrong Shape of Operation

This trade enquiry states its objective precisely, which is useful because the objective needs adjusting. A client server built on a single drive with no redundancy, failing a manufacturer diagnostic and refusing to start, where the request is to retrieve the data "in a way that it can be cloned to another hard drive." The instinct is right and the operation named is not: on a failing drive a clone and an image behave very differently.

The decode: why the order of operations matters more than the destination

What a clone does: copies a failing drive directly onto a replacement, sector by sector, from beginning to end. It proceeds in physical order and does not choose what to read first.

Why that ordering is wrong for a degrading drive: difficult regions are encountered in whatever position they occupy. A clone that meets an unreadable area early can stall there for hours while everything beyond it remains uncopied, and the drive's remaining life is spent on the hardest part.

What an image does differently: the same capture, with control over sequence and time. Per-sector timeouts prevent any region absorbing the available reading, healthy areas are taken first at full speed, and marginal ones revisited across later passes — so the bulk is secured before anything difficult is attempted.

Why that difference decides outcomes rather than merely tidiness: on a drive with limited reads left, order determines how much is recovered. Securing ninety per cent quickly and then working on the remainder is a different result from stalling at five per cent.

Why the destination he wants is still available: an image can be written to a replacement drive afterwards. The outcome he described is achievable — it is the intermediate step that should change, and writing a completed image to new media is fast and safe.

Why that intermediate step is worth keeping regardless: the image remains after the transfer. Anything found to be missing later can be revisited without touching the failing drive again, which on a single-copy system is worth a great deal.

What the absence of redundancy means for the urgency: everything depends on one device. There is no second member to fall back on and no parity to reconstruct from, so the drive's condition is the entire margin.

Why the diagnostic result should be taken seriously: a manufacturer test reporting failure has communicated with the drive and run defined checks. It is a device-level finding rather than an inference, and it means further use is not advisable.

What should be checked alongside, since this is a business system: whether any backup exists at all. A server built on a single drive frequently has none, and establishing that changes both the urgency and the conversation with the client.

What must not happen meanwhile: no further start attempts and no cloning software run against it. Each attempt reads the drive in physical order and stops where it struggles.

On the bench

The clone request was converted to a block-level image before any copying began — cloning proceeding sector by sector in physical order without choosing what to read first, so an unreadable region encountered early absorbs hours while material beyond it remains uncopied, whereas imaging imposes per-sector timeouts, takes healthy regions first at full speed and revisits marginal areas across later passes. The image was written to replacement media once capture was complete, and retained.

The outcome

The clone request converted to a block-level image, healthy regions captured first under capped timeouts, and the image written to replacement media afterwards. 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: the destination you want is right and the operation should change. A clone reads in physical order and stalls where it struggles — an image controls the sequence, so the bulk is secured first.

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.