Accidentally Deleted NAS Volume or Pool: Synology and QNAP Recovery

Quick answer: If a volume or storage pool on your NAS was deleted, formatted or recreated by accident, power the device off immediately. In most cases the real data is still on the disks, what got deleted is only the top layer table and the RAID definition. As long as the device stays on, the new structure starts to overwrite the old data. Do nothing to the disks and call us. Emergency line: +90 536 662 38 09.

Why is "init" or "format" not the end of your data?

Creating a volume on a NAS looks like wiping the disks, but it actually does far less. Synology and QNAP build a layered structure on the disks: at the bottom a RAID array (usually Linux md-raid), a logical volume (LVM) on top of it, and the file system at the very top (mostly btrfs or ext4 on Synology, ext4 on QNAP). When you create or format a volume, the system usually only writes the initial tables and metadata of these layers. Terabytes of actual data blocks stay where they are.

So when you say "I accidentally created a volume" or "I deleted the pool", what is lost is the map, not the treasure itself. If the correct parameters are recalculated, the old file system can be reassembled and the data read again.

The first and unbreakable rule: power off

As long as a NAS stays on, it keeps working in the background. A newly created empty volume can start spreading btrfs or ext4 metadata across the disks, scrubbing, or resyncing the RAID. Each of these risks overwriting the old data. RAID resync in particular means recalculating parity from scratch, and the longer it runs the lower the recovery odds.

So the first move is clear: power the NAS off from the button, do not copy any file to the new volume, do not run "repair" or "recreate" wizards. Pulling the disks out and trying them one by one in another computer is also risky, the order and pairing can break.

Why is the structure different on Synology SHR and QNAP?

Unlike classic RAID, Synology Hybrid RAID (SHR) combines multiple md-raid layers with LVM in the background to use disks of different sizes efficiently. This adds flexibility but means rebuilding a deleted SHR pool requires knowing more parameters than standard RAID 5: which disks belong to which md layer, the chunk size, the disk order.

On the QNAP side there is usually a flatter md-raid with LVM and ext4, but QNAP's own thin provisioning and qtier layers can complicate things. With both vendors the issue is the same: find the correct RAID parameters (order, chunk, parity layout) again and reassemble the upper layer.

Layers in the recovery process

Layer Synology QNAP Role in recovery
RAID array md-raid (SHR multi layer) md-raid Disk order and parity recovered
Logical volume LVM LVM Volume boundaries resolved
File system btrfs / ext4 ext4 File tree rebuilt
Upper data btrfs subvolume thin provision Snapshots and shares return

In the DSET lab the process works like this: each disk is first imaged bit by bit, and all work is done only on these copies. Then the RAID parameters (disk order, chunk size, parity cycle) are recalculated in software and the array is virtually reassembled. The btrfs or ext4 file system map is extracted on top and files are delivered with names and folder structure. In a synchronized multi disk array this is delicate, a single wrong parameter produces corrupt files.

Why does multi disk synchronization matter?

In a NAS array, data is not on a single disk, it is striped across all disks. Pieces of one file sit in sequence across different disks. So for recovery the disks must be brought together in the correct order and alignment. Misplacing a single disk, or one disk lagging at an old sync point, makes the whole file structure unreadable. This is exactly the part that demands expertise.

Connection to other NAS scenarios

If your NAS did not lose a volume but one or more disks failed, the process is a little different. For Synology see our Synology NAS crash SHR data recovery article, for QNAP see our QNAP NAS data recovery article. If you suffered a general RAID level crash, the RAID 5 crash recovery process article is a good start for cost and time expectations.

DSET's accidentally deleted NAS approach

We have been recovering NAS arrays in our Ankara Hacettepe Teknokent Beytepe lab since 2003. In accidentally deleted or recreated volume cases the most critical factor is how long the device stayed on after the deletion. Arrays powered off immediately recover almost entirely. Our success rate is 99.4 percent. First diagnosis is free, no data no fee.

Frequently Asked Questions (FAQ)

I deleted the volume and created a new one, is my data gone?

If you powered off immediately, most likely not. The new volume only writes top layer tables, the actual data blocks usually stay untouched. But if you copied files to the new volume or left it on for a long time, some may have been overwritten.

Will the disks read if I pull them and connect them to a computer?

Connecting them one by one is risky. The order can break, and some operating systems suggest auto "initializing" and write to the disk. The safest path is to bring the disks to the lab without touching them.

Is Synology SHR recovery the same as normal RAID 5?

No. SHR combines multiple md-raid layers with LVM to use different sized disks. Rebuilding it requires more parameters and is more complex than standard RAID.

I had thin provisioning on QNAP, is that a problem?

It can be. Layers like thin provisioning and qtier keep extra metadata tracking which blocks were actually written. Recovery has to interpret these layers correctly, so expertise matters in QNAP cases.

The device is still on, what do I do?

Power it off from the button right away. Do not run any repair wizard, do not copy files to the new volume. Then call us for a free diagnosis.

Sources