Invention:
This patented invention provides a significant improvement in speed, storage space and array failure rate required for encoding and decoding double-parity systems (e.g. RAID-6) to provide fault tolerance. Applications include data storage such as disk drives as well as data transmission systems. Matlab code based on the use of quintuple parity has been created and demonstrated to be effective at protecting against data loss.
Background:
Fault tolerant data storage systems, such as those used in data centers, rely on redundant array of independent disks (RAID) technology to significantly reduce the impact of errors. Data loss can occur due to reasons other than an apparent disk failure, including silent data corruption due to undetected discs errors (UDE), failed disks operating in degraded mode, and/or inability to detect or correct errors operating in a degraded mode. Given the ever increasing storage density of disk drives and potential increase in errors rates as computing needs expand, an improved approach to RAID is critical.
Applications:
- Data storage manufacturing and providers
- Semiconductor and chip manufacturing
- Operating system development
- RAID Hardware manufacturing
- Correcting for errors in signal transmission such as telecommunications
Advantages:
- Cost efficient
- Greater read/write speeds
- Reduced storage overhead
- Orders of magnitude improvement in failure rate
Patents:
Granted US Patents 10,997,024 and 11,531,591
