Data Recovery Case File · NAS & Home Servers · Names Without Contents

An array answering the register but not the shelves: the half-present NAS decoded, the two-drive fork mapped honestly — and the shares brought home from the drives he wisely pulled

His enquiry ended on the sentence every NAS case should start with. The unit: a Netgear ReadyNAS four-bay running two 2TB Seagate Barracuda drives. The presentation: it "connects but only displays the folder names" — and any attempt to open them fails with the network's least helpful poetry: "You might not have permission to use this network resource... The specified network name is no longer available." The physical clue: one of the two drives is making a loud buzzing noise. And the conduct: "I have removed both HDDs but would like a preliminary diagnosis before taking further action." Drives out, hands off, question asked — the exact sequence this archive begs of every NAS owner, performed unprompted. The diagnosis he asked for is below, fork and all.

The decode: names without contents, the buzz — and the two-drive fork answered honestly

Why the folder names still show: a NAS keeps its shares' top-level structure close at hand — cached, cheaply served, the register at the front desk — while opening a folder demands real reads deep into the array. His unit was a shop with the till working and the shelves unreachable: the interface serving what it already held, every substantive request dying against a storage layer that had stopped answering, and the "network name no longer available" error is simply how Windows reports a share that greeted it and then went silent mid-conversation. The buzz names the culprit: a loud buzzing drive is a mechanism in real distress — motor or heads labouring — and one failing member is exactly enough to produce the half-present array. The fork his two-drive unit sits on, mapped honestly because he asked for a preliminary diagnosis: a two-bay-populated NAS of this class runs its pair in one of two very different marriages. Mirrored (the common default for two drives): each member carries a complete copy — the healthy drive alone may hold everything, the buzzing one becomes the backup question rather than the blocker, and prospects are excellent. Striped or expanded: the data interleaves across both — the failing member must be imaged before anything reassembles, and the recovery runs through the bench first. Which marriage his unit performed isn't guessable from the outside; it's read from the members themselves at assessment — and this is precisely why his drives-out-hands-off conduct matters so much: the catastrophic NAS mistakes all happen at this fork before it's identified — re-inserting drives, accepting rebuild prompts, letting a degraded array resync onto a failing member — every one of them a write aimed at the exact structures the answer lives in. He took the drives out and asked. The fork got answered on images instead of gambles.

The recovery — per member, then the marriage

Both members were imaged individually — the healthy drive at pace; the buzzing one first, and first-class: its distressed mechanism treated at the laminar flow bench with donor attention, then imaged under management at convalescent tempo, coverage closing high. The configuration read from the pair resolved the fork (the assessment's whole point, delivered as promised in the preliminary report before any further work was authorised), the array was reconstructed from the images — the originals never re-married, never resynced, never asked to perform again — and the shares mounted whole: the folder names his NAS had been advertising finally opening onto their contents, verified across the set. Delivery went out on new media in duplicate, with the NAS doctrine attached for the unit's next life: redundancy is a running repair-window, not a backup — a degraded array is an emergency precisely because it still half-works — and the alert email his ReadyNAS could have been sending deserves configuring before any drive goes back in a bay.

Outcome

The shares recovered from a half-present array — and the NAS filings his model conduct earns. Names-without-contents is a degraded array announcing itself: the front desk runs on cache while the storage fails beneath — stop browsing, stop rebooting the unit (each boot re-exercises the failing member), and treat permission-flavoured errors on a previously-fine NAS as storage news, not account news. Pull the drives and ask — exactly his sequence: label the members by bay, keep them out, and let the mirrored-or-striped fork be answered from images; never re-insert, rebuild or resync a degraded set. And carry the structural lesson: a NAS's redundancy buys time to replace a failing drive, not immunity from this page — the array needs a backup that isn't in the same box, because "one of the two is buzzing" should only ever be an errand. He asked for a preliminary diagnosis before taking further action. That sentence, adopted nationally, would empty half this archive's NAS shelf.

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.