Data Recovery Case File · NAS & Network Storage · Count Them and Look Them Up
A Device Without a Screen Reports Faults by Number
His enquiry contains something unusually precise for a device that cannot display words. A two-drive network unit that "has worked fine until today, when it has started blinking red — seven short blinks — and I am unable to access any information." He counted them, which is the useful half of the diagnosis, because that number corresponds to a specific fault the unit believes it has.
| Media | Two-bay network storage unit with removable members — signalling a coded fault condition; shares not accessible; array configuration to be established |
| Reported situation | Two-bay network unit in normal service · unit signalling a repeating coded fault pattern · specific blink count observed and reported by the owner · no access to stored content · contents required |
| Fault class | Unit-reported fault of identifiable type — code interpretation determinative of response; member configuration governing scope |
| Equipment used | Blink code interpreted against manufacturer documentation before any intervention · array configuration established from the members · no reset or reinitialisation permitted · bay positions recorded before removal · members imaged individually write-blocked and assembled offline |
The decode: why the count is the diagnosis
Why devices signal in patterns: a network unit has no screen and no keyboard. Its only way of reporting a specific condition is a coded sequence — a number of flashes, a colour, or a combination — and manufacturers document what each one means.
So the first step costs nothing: look the code up. The manufacturer's documentation lists what each pattern indicates, and the answer distinguishes conditions with entirely different responses — a failed member, a filesystem that will not mount, a firmware fault, an overheating condition, a fan failure.
Why that matters more than it sounds: those five faults call for five different actions, and one of them is urgent while others are not. Without the code, everything is guesswork, and guesswork about a network unit usually ends in a reset.
Why he is already ahead of most enquiries: people report a red light. He reported a count, and the count is the entire content of the message the unit is sending.
What the code will not tell him: whether the data is recoverable. It reports what the unit has concluded, which may be accurate or may be the unit misreading a member that is actually fine — a drive dropping out for a moment can be marked failed permanently.
What must be established alongside it: how the two drives were arranged. Mirrored, each holds a complete copy and one surviving member is everything; combined, both are needed and the scope is entirely different. That is readable from the unit's interface or from the members themselves.
What must not happen, and the unit will suggest most of it: no reset, no reinitialisation, no acceptance of any offer to repair or recreate the volume, and no rebuild. A rebuild reads every sector of the surviving member at sustained load, which is the heaviest demand it will ever face and is being made at the worst moment.
What to do before removing anything: label the bays. Member order can matter to assembly, and drives that come out unlabelled are drives whose order has to be determined analytically afterwards.
On the bench
The blink code was interpreted against manufacturer documentation before any intervention — a unit without a display reporting specific conditions by coded sequence, with the count distinguishing member failure, mount failure, firmware fault and thermal conditions, which call for materially different responses. Array configuration was established from the members, no reset or reinitialisation was permitted, and bay positions recorded before removal.
The outcome
The code interpreted before any intervention, configuration established from the members, and bay positions recorded before removal. 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: you counted them, which is the whole diagnosis. A unit with no screen reports specific conditions by coded sequence, and the manufacturer's documentation says what that number means.
Network unit signalling with a blinking light
Count the flashes precisely and look the code up in the manufacturer's documentation — that's free, it takes minutes, and it's the entire content of the message the unit is sending. A device with no screen reports specific conditions by coded sequence, and the count distinguishes a failed member from a mount failure from a firmware fault from overheating, which need entirely different responses. Reporting "a red light" tells nobody anything. Meanwhile, don't reset it, don't reinitialise, and refuse any offer to rebuild — a rebuild reads every sector of your surviving member at maximum load, at the worst possible moment. Label the bays before removing drives.
Count them and check the manual — then call Easy Data Recovery on 028 9002 0144; code interpreted before any intervention, bay positions recorded, members imaged individually and 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.