The network layer
Unit is good, volume is good. Changed IP address. Switch port not functioning. Your SMB version no longer matches. Windows disabled SMBv1 and orphaned a great many older units overnight.
Not a data fault
A NAS is a small PC with some disks, and a failure anywhere from the network to the file system will produce the same experience for you at your desk. Identifying how far along that chain things are still working takes a few minutes and will dramatically alter everything after.
Network, then share, then volume, then array — in that order because each is cheaper and safer to test than the one below it. Jumping straight to your disks is how permissions problems become recovery jobs.
"I can't access my files" describes four rather different situations, three are far less serious than the fourth.
A NAS is a small computer with some disks attached. There are a network, an operating system, a storage array and a file system between your files and your screen — and a failure anywhere along that chain produces the same experience for you at your desk.
Therefore the initial logical step isn't to think that there is a failure on your disks. The first step should be to identify where down the chain things are working. This you can quickly assess for yourself in a matter of minutes and completely alters what occurs after.
These two tests take ten seconds each. They will allow you to identify the problem location.
If it responds, the NAS has booted, loaded its OS, initialised its network interface, and is currently running the network stack. Power supply, motherboard and networking layers are therefore OK; the problem lies below those levels in the storage layer.
If the web interface loads, the unit is fully operational. Therefore, as the problem narrows even further, the only possible faults could be related to either mounting a volume or starting an appropriate sharing service against it.
Most will tell you outright when you're into one. Degraded indicates that there was a loss of one member and the redundancy has been consumed. Get your data copied out first!
Then it's not booting up. This can be either from your power supply, board, or – worth noting – operating system. On most units the OS lives on the disks themselves as opposed to living in the enclosure.
What they look like & how serious they are.
Unit is good, volume is good. Changed IP address. Switch port not functioning. Your SMB version no longer matches. Windows disabled SMBv1 and orphaned a great many older units overnight.
The array assembles but the volume won’t mount — often after a power failure cut off a write process or when a thin-provisioned pool quietly filled to capacity.
One disk failed and no one noticed until another failed. Most units come from here, and this is work on an array rather than the NAS.
Failed power supply, bad board, or firmware update that never finished. Usually the disks are unscathed — the obstacle is the box, not the storage.
Each will be offered as a solution and each can end the job.
Do not start a rebuild before copying your data off. The cheapest form of recovery you have available if your unit reports a degraded array but continues to serve files is the rebuild window. A rebuild takes hours reading each sector of every remaining member at exactly the time another disk with undetected bad sectors also fails.
Do not reinsert a disk the unit ejected. There are many cases where it appears as though there could be more information than what the controller is aware of; in such cases, reinserting the disk can trigger a resync operation in the opposite direction, thus over-writing good data with old data.
Do not remove all the disks at once. For the majority of units running under a Linux based OS, Synology, QNAP, Buffalo, LaCie, the operating system exists within partitions located on the member disks that mirror those partitions among themselves. If all of the disks are removed simultaneously, there is no longer anything left upon which the unit may boot itself and therefore no way to restore good data back to new replacement disks via a rebuild.
At its core, most of these units run conventional forms of Linux-based storage (i.e., using mdadm with an LVM pool and then placing ext4 or Btrfs on top). After imaging the disks used by a dead NAS appliance; you would be able to remove said appliance completely from the process — which is one reason a dead NAS appliance is generally viewed as being less daunting than it initially appears. NAS recovery in Manchester is from £500 +VAT after a free 48-hour diagnostic.
Try pinging it by IP. If it responds, the box has booted, loaded an OS and started network. So the problem is with your storage layer below. The data itself should be good. If you don't get a response from Ping, it didn't boot.
Your box is working. But there's something wrong with your storage layer. Either the array won't assemble. Or the volume won't mount. Or the sharing service can't start against it. Log in to admin interface via browser and see if that helps narrow down what is going on.
It might not be at all. Many older units were orphaned when Windows disabled SMBv1, and they then present exactly like a failure while being completely healthy. Check if you can still access admin interface before assuming worst case.
Yes. Degraded means the redundancy has already been spent. Any further fault means total loss. Get your data off asap. Replacing anything else or starting a rebuild will close out that window for the cheapest recovery possible.
Not all at once. On most Linux systems the OS resides in a portion of the member disk(s), duplicated on each of those disks – pull all the disks out and there's no way for the system to boot or to rebuild when you replace them.
Yes. Usually. And fairly easily. Most run normal Linux RAID software with an LVM layer on top – therefore as soon as the disks are imaged, the entire RAID will be reassembled outside the appliance.