Specialist work · Buffalo NAS
Buffalo announces trouble with a blinking count and the letters EM, and neither tells you which layer failed: it may be the boot partition after a failed update, a member that has dropped out of the array, or a data volume that will not mount. The remedies differ, and the wrong one writes over what you came for. Until the disks have been imaged, leave the unit powered down and decline every offer to restore or reinitialise; a Buffalo NAS can be posted in from Maidenhead complete, disks left in their bays.
Describe the fault to an engineer
0800 6890668
| The symptom | The likely cause | Start here |
|---|---|---|
| The LEDs are blinking a fault code | One code, three candidates: the array, a disk, or the firmware | Power it down; note the blink count |
| The console shows an array degraded | A member has dropped out; the unit carries on with nothing spare | The clock is against you now |
| The LinkStation shows no sign of life | The chassis holds nothing; your shares are on the disks inside it | One of the better outcomes |
| A firmware update ran and it never came back | The boot partition took the damage, not the data partition | Refuse the reinitialise prompt |
| Windows offers to format the disks | As expected: Windows cannot read XFS or ext4 | Say no to it |
| Empty shares, with a ransom note left behind | A campaign that hunts NAS boxes left on the internet | Photograph the note, pull the power, see the ransomware page |
Nothing becomes chargeable until you have said yes to it. Each item takes a case number of its own the moment it arrives, an engineer establishes what has actually failed, and you are given a candid view of which files stand a realistic chance of returning. The single quote that follows is fixed, written down, and costs nothing.
No analysis begins until each member has a sector-by-sector image behind it. The chassis stays unpowered while that runs, so no firmware routine can start itself and nothing is written to the originals.
Everything from here runs on the copies: partitions mapped, Buffalo's array metadata decoded for member order and chunk size, the array assembled away from the unit, and the XFS or ext4 filesystem mounted read-only.
Array, partition table or filesystem — whatever needs mending is mended on the assembled copy and never on your disks. A damaged XFS log or ext4 journal is dealt with there, and then every share is opened and checked.
Nothing is billed until you have seen the recovered file list in full and told us to go ahead. What comes back arrives on new media, with the return journey at our cost, and the case stays open at this end until you have opened every file yourself.
The detail that decides most of these cases: a Buffalo array is put back together from its own metadata rather than a standard superblock, so member order, chunk size and start offset all have to be established from the images before a single share appears. Get that wrong and the disks look blank; get it right and a unit the firmware wrote off usually gives everything back.
Two 8TB disks and two 4TB, one member already down, and the TeraStation rebuild ran overnight only to halt with the array marked failed. Underneath sat Buffalo's own RAID descriptors and an XFS volume the firmware would no longer mount. We imaged all four, worked out member order and offset from the images, and every share opened exactly as it had.
Rarely. EM stands for emergency mode, and it means the firmware would not start — not that the array has gone. The disks are read away from the chassis, so what the front panel is blinking rarely settles the outcome. Leave the unit switched off and every option stays open.
Rarely, no. Older LinkStation and TeraStation units are formatted XFS and newer ones ext4, but the array is described by Buffalo's own metadata rather than a standard superblock, so a Linux machine usually lists partitions it will not assemble. Work from images, establish member order and chunk size, and the volume mounts read-only on plain hardware.
Almost always they are still there. An interrupted update damages the small system partition the unit boots from, and that is not where your shares live; the data partition is normally untouched. The risk comes next, from the recovery routine that offers to put the firmware back and reinitialises the disks while it does so.
It does not. RAID carries you through one dead disk and nothing else: deletion, ransomware, a failed enclosure, a botched update or a unit pulled off a desk all take the whole volume with them. Most Buffalo jobs that reach us were the only copy of something. If it is easier, a NAS can travel to the Guildford lab whole, drives still in it.
Each further power-up asks more of a drive that has already begun to fail. Open the case first and let the free diagnosis tell you what is left.