When Synology Says “Degraded” — Even Though Every Disk Passes SMART
This is one of the most unsettling Synology moments.
You log into DSM expecting everything to be fine — all your drives show Healthy, no SMART alerts, no red icons.
But at the top of Storage Manager, you see the warning:
“RAID Group Degraded.”
Nothing failed.
Nothing was replaced.
Nothing obvious happened.
Yet Synology is treating your perfectly normal-looking array as damaged.
And here’s the truth Synology doesn’t explain clearly:
A RAID group can degrade even when every disk is technically healthy — because the problem lives in metadata, not in the drives.
1. What You See
- RAID Group: Degraded
- All drives show Normal/Healthy
- No SMART warnings, no bad sectors reported
- No recorded drive drops or failures
- DSM logs show warnings about:
- “I/O error during rebuild”
- “System partition mismatch”
- “Write failure”
- “mdadm: resync aborted”
- Storage Manager refuses to repair, or repair goes nowhere
2. Why It Happens (Synology Real-World Behavior)
Synology arrays depend on Linux MD metadata, DSM layers, and dual system partitions.
A RAID group degrades with no failed drives when:
Metadata drift occurs across members
- Slightly out-of-sync MD epoch counters
- Partially written superblocks
- DSM updates interrupted mid-write
- Cache not flushed on shutdown
A drive briefly dropped and came back
Even a 1-second disconnect can:
- Un-sync system partitions
- Flag dirty/uncertain regions
- Cause mdadm to mark the array degraded for safety
Sync failure during background repair
If a drive returns a single UNC read during routine parity check:
- mdadm halts
- DSM surfaces “Degraded”
- Drives remain “Healthy” — misleading
Filesystem layer conflict
Btrfs or ext4 metadata may disagree with RAID layer.
Drives look fine — but the MD layer knows parity is not trustworthy.
3. What NOT To Do
- Don’t remove/re-add drives “because they’re healthy”
- Don’t run DSM “Repair” repeatedly
- Don’t reboot in hopes it will clear the message
- Don’t try to rebuild from Storage Manager without understanding the mismatch
- Don’t assume SMART clean = array safe
These actions risk overwriting metadata that proves which blocks changed.
4. What You CAN Do
- Check DSM logs for: I/O errors, aborted resync, system partition mismatch
- Run dmesg tail for controller, SATA, or timeout warnings
- Export MD configuration (mdadm –detail –scan)
- Clone each drive BEFORE any forced operation
- Validate:
- mdadm UUIDs
- Superblock epochs
- Member order
- System partition agreement
- Check SMART pending sectors (not just reallocated)
- Determine if the array was mid-write during power loss or update
5. What This Means for Your Data
- Drives may be healthy but disagree about the last known write
- Parity mismatch means the RAID can’t rebuild safely
- The array is still recoverable — but only if metadata is preserved
- Forced operations can overwrite the only evidence of parity alignment
Synology’s caution is correct: it’s protecting your data by refusing to guess.
Diagnostic Overview
- Device: Synology NAS (DSM / Linux MD)
- Observed State: RAID Group Degraded — All Drives Healthy
- Likely Cause: Metadata drift, partial writes, or system partition mismatch
- Do NOT: Re-add drives, repeatedly repair, or reboot hoping it will clear
- Recommended Action: Export MD metadata, clone drives, validate epochs and alignment