The Mechanism
All hard drives have a manufacturer-rated limit for Unrecoverable Read Errors (UREs)—microscopic sectors that physically degrade over time. In a perfectly healthy RAID 6 array, hitting a bad sector is a non-issue; the dual parity instantly reconstructs the missing piece of data.
The Impact on RAID 6
The math changes when the array is already severely degraded:
- If one drive has failed and the controller hits a URE during a rebuild, the second layer of parity easily fixes it. The rebuild continues safely.
- If two drives have failed, the array has zero remaining redundancy. If the controller encounters even a single URE or "silent data corruption" block on any of the surviving drives while trying to rebuild, it has no remaining parity to calculate the missing data. The rebuild aborts, and the array crashes.
