This is one of the scariest Synology moments.

DSM boots normally.
Storage Manager shows Healthy.
SMART tests pass.
No red icons.

But when you browse your shared folders:
No files.
Sometimes whole shares are missing.
Sometimes the NAS mounts, but the directories appear empty.

It doesn’t make sense — the hardware is fine, the RAID group looks normal, and DSM isn’t warning about corruption. Yet something is clearly wrong.

Here’s the truth Synology doesn’t tell you:
A “Healthy” status doesn’t mean your filesystem is intact — it only means the RAID layer isn’t reporting a failure.


  • RAID Group: Healthy
  • Storage Pool: Healthy
  • Shared folders: Missing, empty, or unmounted
  • DSM won’t mount specific volumes or says “Folder path does not exist”
  • File Services still run — but expose nothing
  • Btrfs volumes may show:
    • No snapshots
    • No shared-folder metadata
    • No visible subvolumes
  • Logs show:
    • “Failed to mount volume”
    • “btrfs: open_ctree failed”
    • “Superblock mismatch”

Filesystem corruption at Btrfs or EXT4 layer

  • RAID can be perfectly healthy
  • But the filesystem tree/root/superblocks can still be damaged
  • DSM reports “Healthy” because RAID ≠ FS

System partitions out of sync

A brief drive drop may:

  • Cause metadata mismatch
  • Prevent DSM from mounting system-level structures even though RAID is clean

Unexpected shutdown during snapshot/metadata commit

  • Btrfs is vulnerable to mid-transaction inconsistency
  • Leads to a “healthy RAID, broken tree root” condition

DSM update partially applied

If DSM updates system partitions but fails before writing volume metadata:

  • Volumes appear empty
  • Shares disappear
  • RAID still shows “Healthy”

Volume mounted read-only

To prevent damage, DSM may mount Btrfs RO.
Users see “empty folders” because the writable layer never attaches.


  • Don’t format the volume just because DSM recommends it
  • Don’t delete/recreate shared folders (destroys metadata that proves layout)
  • Don’t run fsck.ext4 or btrfs check –repair without cloning
  • Don’t reset DSM
  • Don’t initialize storage pool unless your goal is permanent data loss

DSM’s repair prompts are dangerous when the filesystem root is unstable.


  • Check /var/log/messages and kernel logs for btrfs/EXT4 mount errors
  • Confirm whether DSM mounted the volume read-only
  • Export MD metadata and RAID layout (even though RAID looks fine)
  • Clone drives BEFORE any filesystem-level repair
  • Validate superblocks, tree roots, checksums
  • Use btrfs-progs offline against cloned images — never on live NAS media
  • Recover snapshots directly from binary if metadata tree partially exists

  • Your data is likely still present on disk
  • RAID being healthy is a good sign — the blocks aren’t missing
  • The filesystem index (the map of where files live) is damaged
  • Recovery is high-probability if no write operations overwrite metadata
  • Forced DSM repair often permanently destroys the remaining index structures

This problem looks catastrophic, but it’s usually recoverable with correct triage.

Diagnostic Overview

  • Device: Synology NAS (DSM / Btrfs or EXT4)
  • Observed State: RAID Healthy — Shared Folders Empty
  • Likely Cause: Filesystem metadata corruption, system partition mismatch, or read-only mount
  • Do NOT: Format, recreate shares, reset DSM, or run repair tools on live disks
  • Recommended Action: Clone drives, check logs, validate superblocks and FS tree roots offline