The network layer
The unit is fine and the volume intact. A changed IP address, a failed switch port, or an SMB version the two ends no longer agree on — Windows disabling SMBv1 orphaned a great many older units overnight.
Not a data fault
A NAS is a small computer with disks attached, and a failure anywhere between the network and the file system produces the same experience at your desk. Working out how far along that chain things still work takes a few minutes and changes everything that follows.
Network, then share, then volume, then array — in that order, because each is cheaper and safer to test than the one beneath it. Going straight to the disks is how a permissions problem becomes a recovery job.
“I cannot reach my files” describes four quite different situations, and three are considerably less serious than the fourth.
A NAS is a small computer with disks attached. Between your files and your screen sit a network, an operating system, a storage array and a file system — and a failure anywhere along that chain produces the same experience at your desk.
So the useful first move is not to assume the disks have failed. It is to work out how far along that chain things are still working, which you can do yourself in a few minutes and which changes everything about what happens next.
Two checks, ten seconds each, and between them they locate the fault.
If it replies, the NAS has powered up, loaded its operating system, initialised its network interface and is running its network stack. Power supply, board and networking are all fine, and the fault is below that — in the storage layer.
If the web interface loads, the unit is fully operational and the problem is narrower still: the volume will not mount, or the sharing service will not start against it.
Most units say so plainly once you are in. Degraded means a member has dropped and the redundancy is already spent — copy your data off before doing anything else.
Then it is not booting. That could be the power supply, the board, or — worth knowing — the operating system, which on most units lives on the disks themselves rather than in the enclosure.
With what each looks like and how serious it is.
The unit is fine and the volume intact. A changed IP address, a failed switch port, or an SMB version the two ends no longer agree on — Windows disabling SMBv1 orphaned a great many older units overnight.
The array assembles but the volume will not mount — often after a power cut interrupted a write, or a thin-provisioned pool quietly filled to capacity.
A member dropped and nobody noticed until a second followed. This is where most units arrive from, and it is array work rather than NAS work.
Power supply, board, or a firmware update that never completed. The disks are usually untouched — the obstacle is the box, not the storage.
Each is offered as a fix and each can end the job.
Do not start a rebuild before copying your data off. If the unit reports a degraded array and still serves files, that window is the cheapest recovery available. A rebuild reads every sector of every surviving member for hours — precisely when a second disk with latent bad sectors gives up.
Do not reinsert a disk the unit ejected. It may hold more than the controller thinks, and reinserting can trigger a resync in the wrong direction, overwriting good data with stale data.
Do not remove all the disks at once. On most Linux-based units — Synology, QNAP, Buffalo, LaCie — the operating system lives in a partition on the member disks, mirrored across them. Take them all out and the unit has nothing left to boot with, and cannot rebuild onto replacements.
Underneath the interface, most of these run conventional Linux storage: mdadm with an LVM pool and ext4 or Btrfs on top. Once the disks are imaged, the appliance can be removed from the equation entirely — which is why a dead NAS is usually a better prospect than it looks. NAS recovery in Manchester is from £500 +VAT after a free 48-hour diagnostic.
Ping it at its IP address. If it replies, the unit has booted, loaded its operating system and brought networking up — so the fault is in the storage layer beneath, and the data is usually intact. If it does not reply, it is not booting.
The unit is working and the storage layer is not — either the array will not assemble, the volume will not mount, or the sharing service cannot start against it. Try the admin interface in a browser to narrow it further.
Possibly not at all. Windows disabling SMBv1 orphaned many older units, which then present exactly like a failure while being completely healthy. Check whether the admin interface still loads before assuming the worst.
Yes. Degraded means the redundancy is already spent, so any further fault means total loss. Copy your data off now, before replacing anything or starting a rebuild — that window is the cheapest recovery available.
Not all at once. On most Linux-based units the operating system lives in a partition on the member disks, mirrored across them — remove them all and the unit has nothing to boot with and cannot rebuild onto replacements.
Usually, and fairly easily. Most run standard Linux software RAID with an LVM pool on top, so once the disks are imaged the array is reassembled entirely outside the appliance. The box failing is often the best of the possible faults.