LinkStation and TeraStation boxes are Linux machines running conventional software RAID beneath a deliberately closed interface. When one stops working the disks are usually intact — the obstacle is the appliance, not the storage.
Buffalo uses standard mdadm arrays with XFS or ext file systems, so once the disks are imaged the reconstruction is well understood. What makes these jobs awkward is the appliance’s own recovery tooling, not the format.
A LinkStation or TeraStation flashing an error and refusing to serve shares has usually dropped into EM — emergency mode. It means the unit could not mount its system partition or its array, and it is a state that alarms people far more than it should: EM mode is the appliance reporting a problem, not the disks being destroyed. The array beneath is frequently intact.
The risk is what EM mode invites you to do. Buffalo’s NAS Navigator offers to restore firmware, and on several models that process rebuilds the system partition and reinitialises the data array along with it. People who lose Buffalo data usually lose it here, minutes after deciding to fix it properly.
The three recurring causes are a power cut during a write, leaving the array metadata inconsistent; a firmware update that did not complete; and an ordinary disk failure in a RAID 1 or RAID 5 set that went unnoticed until a second member followed. Buffalo units are often installed and then left alone for years, so a degraded array can run unprotected for a very long time before anybody looks.
External Buffalo drives — the MiniStation and DriveStation lines — behave like any other USB external: the bridge board fails at least as often as the disk, and the unit should be sent whole rather than opened.
Disks are removed, labelled by bay and imaged read-only. The mdadm array is then reassembled from the images by deriving its parameters directly rather than trusting superblocks that may be stale, and the XFS or ext file system is rebuilt from the assembled volume. Because Buffalo uses conventional Linux storage underneath, a unit whose hardware has died is often a more straightforward job than the closed interface suggests — the appliance is simply removed from the equation.
Either is fine. Label disks with their bay number before removing them, and tell us the model, the RAID level if known, and whether EM mode appeared or a firmware restore was attempted — that last detail changes the approach materially. NAS and RAID recovery is from £500 +VAT after a free 48-hour diagnostic and a written quote; a single external Buffalo drive is from £300 +VAT.
Emergency mode — the unit could not mount its system partition or its array and has stopped serving shares. It sounds worse than it usually is: the disks underneath are frequently intact. The danger is the firmware restore it prompts you toward, which on several models reinitialises the data array.
Usually yes. TeraStation units run standard Linux software RAID with XFS or ext file systems, so once each disk is imaged read-only the array can be reassembled outside the appliance and the file system rebuilt. A dead unit is often a more straightforward recovery than its closed interface implies.
Not if the data matters. Several Buffalo restore routines rebuild the system partition and reinitialise the data array as part of returning the appliance to working order — leaving you with a functioning NAS and nothing on it. Remove the disks instead.
Treat it like any other USB external: try a different cable and port, and listen. Silent and undetected usually means a failed bridge board with a healthy disk behind it. Send the unit complete rather than opening it.
From £500 plus VAT for a LinkStation, TeraStation or other multi-disk unit, and from £300 plus VAT for a single external drive. Fixed in a written quote after a free 48-hour diagnostic.