When a Momentary Power Event Causes Your SmartArray to Lose Track of Its Own RAID
The outage was brief — a flicker, a UPS transfer, a breaker trip.
The server rebooted.
Everything looked normal… except your Virtual Drive was gone.
No filesystem.
No logical volume.
Just disks… but no array holding them together.
If you’re using HP SmartArray, this can happen after sudden power loss, when cached metadata never made it from volatile memory to disk — leaving the controller unsure which version of the RAID layout is real.
This moment is frightening, but fixable — as long as you move carefully.
1. What You See
- Virtual Drive missing from SmartArray Configuration Utility
- Physical drives appear “Unassigned” or “OK” but not part of any array
- No logical volume detected by the OS
- “Configuration Incomplete” or “Previous Configuration Not Found” messages
- Array previously online now absent after a forced/ungraceful reboot
- System boots but storage subsystem appears empty
2. Why It Happens (HP SmartArray–Specific Behaviors)
- Cache-to-disk write did not complete during the power event
- Battery-backed cache (BBWC/FBWC) detected an incomplete flush
- Metadata checkpoint mismatch between cached and on-disk configuration
- SmartArray put the array “on hold” to avoid committing potentially corrupt parity
- Drives reinitialized faster/slower than SmartArray’s expected timing, causing identity mismatch
- Write-back policy left pending updates that were never finalized to disk
- Cache module self-test flagged “inconsistent data” and paused volume presentation
SmartArray’s behavior is protective — not destructive.
3. What NOT To Do
- Do not create a new array using the same disks
- Do not initialize or rebuild anything
- Do not run “Re-enable Logical Drive” without validating configuration state
- Do not clear controller metadata
- Do not swap drives between ports (destroys slot identity)
- Do not accept auto-repair or initialization prompts
Every one of these can permanently overwrite your last good metadata.
4. What You CAN Do
- Enter SmartArray BIOS/SSA and document slot → serial number mapping
- Save/export the current controller configuration (if available)
- Check the SmartArray cache module status (OK / Needs Attention / Inconsistent Data)
- Verify BBWC/FBWC battery health — failed batteries cause incomplete commits
- Review “Controller Event Log” for parity-incomplete or forced restart messages
- Clone/image all members before attempting any metadata operations
- Compare detected array parameters with expected geometry (stripe, parity rotation, size)
5. What This Means for Your Data
- The data often remains intact — the controller lost confidence, not the array
- SmartArray hides a VD when metadata versions don’t align
- With proper imaging and layout verification, the original RAID parameters can be reconstructed
- Recovery depends on preserving both on-disk metadata and any cached remnants
- Power-loss events rarely wipe data — they break controller trust
Diagnostic Overview
- Controller: HP SmartArray (P400/P410/P420/P440/P800 families)
- Observed State: Virtual Drive Missing After Power Loss
- Likely Cause: Metadata commit interrupted; cache-to-disk mismatch
- Do NOT: Initialize or recreate the logical drive
- Recommended Action: Export config; verify cache module; image drives; validate metadata and geometry before reassembly
Back to RAID Controller & Systems Triage