Data Recovery Case File · NAS & Network Storage · A Good Test With a Catch
A Replacement Appliance Refusing an Array Does Not Establish That the Drives Have Failed
This enquiry describes a properly designed test, which is rare and worth crediting. An appliance holding archived data began erroring, so "we tried putting the drives into a box we know works, but that errors as well, so we believe there is a drive issue." The reasoning is sound and rests on one assumption: that a working appliance shown a healthy array would accept it — and appliances are considerably less accommodating than that.
| Media | Multi-drive network storage appliance holding archived content — original unit erroring; members transferred to a known-working unit of the same class which also errored |
| Reported situation | Network storage appliance holding archived content · unit reporting errors · members transferred to a second unit known to function · second unit also reporting errors · drive fault inferred from the substitution · recovery sought |
| Fault class | Array not accepted by either appliance — refusal potentially arising from configuration mismatch rather than member failure; per-drive condition undetermined |
| Equipment used | Substitution result assessed for alternative explanations before drive failure was accepted · no further insertion into any appliance permitted · each member imaged individually write-blocked before any interpretation · array metadata read per member and compared · volume assembled offline independently of any appliance |
The decode: what the test established, and the assumption inside it
Why the test was the right one to design: it separates the variables. Moving the members to a different appliance is exactly how you distinguish a failed unit from failed drives.
What it would have proved had it worked: the original unit was at fault. An array assembling in a second appliance would have settled the matter cleanly.
Now the assumption: that a second unit would accept a sound array. Appliances impose conditions well beyond whether the drives are readable.
What those conditions include: firmware version, model compatibility and configuration records. A unit running different firmware, or of a different generation, may decline an array it can read perfectly.
Why bay order alone can produce a refusal: position is recorded. Members reinserted in a different order may not be reconciled against the configuration the array expects.
Why the system volume matters as well: appliances store themselves on the drives. These units keep their own operating system across the members, so a second appliance encounters another installation rather than plain storage.
What follows for their conclusion: it may be right and it is not established. Errors in both units are consistent with failed drives and equally consistent with two appliances declining an array they will not adopt.
How the question is settled instead: per member, outside any appliance. Each drive is read individually, which establishes its condition without reference to what any unit thinks of the set.
Why the volume is then assembled offline: it removes the appliance entirely. The array is reconstructed from images according to what the metadata describes.
What must not happen now: no further insertion. Each attempt in an appliance risks it writing configuration or beginning a rebuild.
On the bench
The substitution result was assessed for alternative explanations before drive failure was accepted — the test being correctly designed to separate a failed unit from failed members, but resting on the assumption that a second appliance would accept a sound array, when units impose conditions on firmware version, model generation, configuration records and bay order, and store their own operating system across the members. Each member was read individually outside any appliance.
The outcome
The substitution assessed for alternatives, further insertion stopped, and the volume assembled offline from individual images. Free assessment, one fixed written figure including VAT, charged per drive, with 50% of parts and labour upfront where a drive has to be opened. The decode: the test was well designed and rests on one assumption. A working appliance shown a healthy array may still decline it, so errors in both units do not yet establish that your drives are the problem.
After moving drives to a second appliance
Stop inserting them anywhere — each attempt risks a unit writing configuration or starting a rebuild. The test was the right one to design, but it carries an assumption worth examining: appliances impose conditions beyond whether the drives read, including firmware version, model generation, recorded configuration and even bay order, and they store their own operating system across the members. So errors in both units are equally consistent with two appliances declining an array they won't adopt.
That may not be the drives — call Cambridge Data Recovery on 01223 655015; substitution assessed for alternatives, each member read individually outside any appliance, volume assembled offline.
Request a quote online →
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.