Synology SHR, Btrfs & DSM Data Recovery

Every Synology NAS, from a 2010 DS410 to a 2025 DS925+, stores your data in the same basic way: Linux software RAID, often wrapped in Synology Hybrid RAID (SHR), with a Btrfs or ext4 file system on top. When DSM reports that a volume has crashed, the data is usually still on the drives – what has broken is one or more of these layers.

Understanding the layers is what makes Synology recovery possible. This page explains how they fit together, what goes wrong, and how we rebuild them.

How Synology Stores Your Data

• System partitions – every drive holds a small copy of DSM itself, mirrored across all disks. This is why a DiskStation can boot even when one drive fails.
• Data partitions and RAID groups – the rest of each drive is joined into Linux software RAID (mdadm) arrays: RAID 0, 1, 5, 6, 10, or SHR.
• SHR and SHR-2 – with drives of different sizes, SHR slices each drive into partitions and builds several RAID groups from them, then joins those groups together using LVM. A four-drive SHR volume with mixed sizes typically contains two or more separate RAID arrays.
• The file system – ext4 on older and entry-level models, or Btrfs (since DSM 6.0) with checksums, snapshots and compression.
• Encryption – optional encrypted shared folders (eCryptfs) or, since DSM 7.2, full-volume encryption. These need your password or key file.

DSM Versions, 2010 to Today

What Goes Wrong

• Volume crashed – too many RAID members have failed or dropped out, or the RAID metadata is inconsistent
• Storage pool degraded, then crashed during repair – a second drive fails while DSM rebuilds
• Btrfs corruption – the volume becomes read-only or won’t mount, often after a power cut or a failing drive returns bad data
• Drives showing as “Initialized” or “Not initialized” after being moved between units
• Encrypted shared folder won’t mount, or the volume encryption key has been lost
• Deleted shared folders, snapshots or Hyper Backup versions

How We Rebuild a Synology Volume

We image every drive first, repairing any failed disks in our cleanroom. From the images, we read the RAID metadata on each partition to reconstruct every RAID group – including the multiple groups inside an SHR volume – and then rebuild the LVM layer that joins them. Finally, we repair and read the Btrfs or ext4 file system. For Btrfs, we can often recover files from earlier snapshots or tree generations, even after deletion or corruption.

Backups and Sync

We can also recover from Synology’s own backup formats: Hyper Backup repositories (.hbk folders), USB Copy, Snapshot Replication targets, and Synology Drive and Cloud Station sync folders – including restoring a Hyper Backup set without the NAS that created it.

Frequently asked questions