RAID 60 Triage – ADR Data Recovery ~ Advanced RAID & Server Recovery Services

When RAID 60 Goes Offline, Looks Healthy But Won’t Mount — This Is Your Roadmap.

RAID 60 is marketed as the “bulletproof” enterprise RAID level — dual parity on each group (RAID 6) and striping across both (RAID 0).

In real incidents, it also fails in predictable, diagnosable, and recoverable patterns.

This hub page is your symptom-to-cause directory — every major failure mode RAID 60 can produce, each linked to a dedicated deep-dive page with full technical details and citations from TN-R60-001, ADR’s authoritative reference on RAID 60 behavior.

1. The Most Common RAID 60 Failure Symptoms

1. RAID 60 Failed — Group-Level Inconsistency

Why the array dies even though all disks look good.
Group A and Group B no longer agree on write history.

→ https://www.adrdatarecovery.com/raid-triage-center/raid-60-group-level-inconsistency/

2. RAID 60 Showing All Drives Healthy — But Won’t Mount

Array is intact, but the RAID-0 layer cannot assemble because parity epochs diverged.

→ https://www.adrdatarecovery.com/raid-triage-center/raid-60-healthy-no-mount/

3. RAID 60 Rebuild Won’t Start (0% or Immediate Abort)

Controller refuses because the two RAID-6 groups disagree on geometry or sequence.

→ https://www.adrdatarecovery.com/raid-triage-center/raid-60-rebuild-wont-start/

4. RAID 60 Corruption After Drive Swap — Now Unreadable

A replacement drive triggers metadata comparison, exposing mismatched groups.

→ https://www.adrdatarecovery.com/raid-triage-center/raid-60-swap-corruption/

5. RAID 60 Virtual Disk Missing — Or Wrong Size After Restart

Controller sees a foreign config conflict or cannot decide which group is authoritative.

→ https://www.adrdatarecovery.com/raid-triage-center/raid-60-virtual-disk-missing/

6. RAID 60 Online But Directories Empty or Wrong Size

RAID-0 rebuilt over one degraded RAID-6 group, creating ghost sectors.

→ https://www.adrdatarecovery.com/raid-triage-center/raid-60-online-empty/

3. The Real Reason RAID 60 Fails (What No OEM Explains)

RAID 60 is really two arrays pretending to be one.

If Group A and Group B disagree on:

parity epoch
sequence number
stripe width
commit order
geometry
slot identity

…the RAID-0 layer cannot assemble the array safely.

The controller hides the array as a protective action, not a destructive one.

This is why RAID 60 arrays often show:

all disks GOOD
no volume
no mount
rebuild won’t start
foreign configs detected
wrong LUN size

This is normal behavior when groups diverge.

TN-R60-001 documents the exact internal mechanics.

4. When To Stop Immediately (Critical Warning)

Do not perform any of the following on a degraded RAID 60:

Import foreign config
Replace multiple drives
Force rebuild
Initialize or “fix” metadata
Run filesystem repair
Run check/verify on live drives
Let the controller attempt auto-rebuild

Every one of these can permanently overwrite the only remaining correct parity domain.

5. How ADR Recovers RAID 60 Safely

Step 1 — Clone all disks (including “good” ones)

Silent sector errors on survivors destroy RAID 60 math.

Step 2 — Extract both groups’ metadata

We pull parity epochs, commit sequences, and group-level mapping.

Step 3 — Virtualize both RAID-6 groups independently

Treat them as separate entities.

Step 4 — Reconcile parity domain mismatch

Identify which group drifted.

Step 5 — Rebuild RAID 60 virtually

Only after both groups’ math is corrected.

Step 6 — Mount safely once parity is fully validated

Diagnostic Overview

Device: RAID 60 (two RAID-6 groups striped as RAID-0)
Observed State: Offline, missing VD, or mounts but data missing
Likely Cause: Cross-group parity divergence, staggered stripe commits, rebuild residue, or foreign-config asymmetry
Do NOT: Import foreign config, start rebuild, replace more drives, or reinitialize
Recommended Action: Clone all disks, extract group metadata, virtualize both groups, determine true parity epoch, then rebuild RAID 60 in a virtual model before committing changes

RAID 60 doesn’t degrade gracefully —
it fails catastrophically when even one subgroup falls out of line.