A Network Attached Storage (NAS) device is essentially a small computer that has multiple hard drives connected to it. Therefore, “I can't get to my files” could be describing one of four different types of problems and only one is related to data. Most NAS devices use standard Linux RAID software under their OEM interface; therefore, an electronic failure in a unit means it's very rare for the entire volume to be lost. It's what occurs within the first hour after failure that determines the final outcome. Do not attempt to rebuild, do not attempt to re-seat any components, and do not initialise any storage. Send all of the drives, with identifying labels on each individual drive showing the order.
People will attempt to solve their issue with one of four instinctual actions that make sense at the time. All four of these instinctual actions could turn a recoverable volume into a lost one.
Rebuilding reads all sectors of all remaining disks and writes across the replacement; when you are rebuilding an array that has been quietly degraded for several weeks, this constant activity is exactly what causes another disk to fail during the process.
Removing a drive to test it will lose the original order in which the drives were arranged, and order is part of what keeps data from being unrecognisable once the volumes are rebuilt. Before anything comes out, label each bay.
The “create volume”, “initialise”, “repair” and “factory reset” options all create new metadata that will overwrite existing metadata. We would call this the most dangerous button on your entire user interface, and it is rarely labelled as such.
Any degraded array that continues to run is using up the margin that protects your data. When an array degrades, shut it down correctly and identify each disk with the correct bay number. After you have done both, then remove the whole set from service.
You should go through all of these steps before assuming the worst. Three of them don't require you to spend money.
The single factor that allows most NAS recovery projects to succeed, and the primary reason a failed NAS box is usually recoverable.
Almost all consumer and small business NAS run standard Linux software RAID under their own vendor interface: mdadm-based arrays, usually with an LVM layer on top, and either Btrfs or ext4 for the actual filesystem. The hard drives do not use a proprietary format specific to that device.
This matters greatly in practical terms. When the hardware dies, the hard drives can be read by completely different devices, and the array can be reconstructed using the superblocks. You don't need a working NAS of the same make/model. You also won't need to send the drives back to the maker.
Things get complicated at higher levels. Synology Hybrid RAID puts partitions across different size disks. QNAP uses LVM thin provisioning which creates a mapping layer that will have to be recreated before your filesystem becomes accessible again. These are both very well understood problems, but each one will just create additional steps.
A 4-bay unit will get purchased by a company when they are small; it does just what is needed, and the purchase is never revisited. Three years later it contains all of the company's working document files, the accounts and the design history, and it is the only copy of items that people haven't thought about in a while. No one owns this unit, and no one monitors the email notifications. The first anyone finds out about a failed disk is when a second failure occurs and the volume drops.
This is the sequence behind most NAS (Network Attached Storage) jobs coming from the Cambridge science parks and the Milton Keynes/Huntingdon business estates. It is recoverable, provided that the first hour of the process goes as outlined at the beginning of this page.
Send every disk, labelled by bay. Not just the one that failed. All members of the entire set will be needed for rebuilding the array, and if a member is missing on a RAID 5 with one disk already gone, it can’t be rebuilt because there isn’t anything to build from.
Multi-bay NAS recovery begins at £500 +VAT, regardless of what badge is on the front or how many bays you have. Single-bay units begin at £300 +VAT as regular single-drive work.
The price will be fixed in writing after your complimentary 48-hour diagnostic period. If additional mechanical work needs to be done on any individual disks beyond the array rebuild, that will be identified and quoted in addition to your array rebuild price during your diagnostic, not afterwards.
If the data is more important than convenience, don’t rebuild. Rebuilding causes all of your other disks to be constantly under full surface use; so on an array that has been degraded for a while, you are exactly in the right place to have a 2nd drive fail at the time of rebuilding. Copy the data off the array first and rebuild later.
Almost always. Synology, QNAP and most others use standard Linux mdadm arrays, so the disks can be read on another machine and put back together into an array. You don't have to replace the unit with one that matches the original model.
The answer is yes. Just send us the disks, labelled by bay. The way the array was created is still there in the full set of disks, and the position each disk was placed in is part of that structure. Therefore simply label the disks as you pull them out of your unit (bay 1, bay 2, bay 3 etc.). We cannot accept either the enclosure or the controller.
In many cases where the volume is unmounted, it's not due to a failed disk. Often it's because of a thin-provisioning metadata problem. The mapping layer that connects the logical volume to the physical blocks has become corrupted. This issue is recoverable, but do not let the unit attempt a repair first.
No. This is the misconception behind most of the jobs we do. A RAID system will protect against a hard drive failure. However, it will provide no protection against deletion, ransomware, fire, theft or having the RAID controller write garbage to each disk simultaneously.
Typically 5-10 working days. Each individual disk is imaged prior to any attempt to reconstruct. On a four-bay unit using large disks, imaging alone can consume much of that timeframe.
From £500 +VAT after a free diagnostic, regardless of what the label says on the front. Please send each disk, labelled by bay if possible, and please do not allow another rebuild, or a scheduled scrub, to occur while you decide.