Research on RAID Coding Schemes for Tolerating Multiple Disk Failures

As the Redundant Arrays of Inexpensive Disks(RAID) scale up,multiple disk failures are likely to take place at the same time.The single erasure code is hard to meet the requirements of high reliable storage.We reviews the coding schemes and features of various MDS array codes for tolerating up to double disk failures and compares their redundancy performances.A class of MDS Abelian group array codes for tolerating up to three disk failures is presented based on circular permutation matrices.Since the encoding and decoding are very efficient,it becomes the trend of application in large-scale RAID storage systems.