2D-RAID-based array self-adaptation strategy

This paper proposes a strategy which is able to improve the reliability of 2D-RAID array by reorganizing the effective disks left when some disks in the array fail, so as to reduce the potential impact from extra disk failures and effectually increase the reliability class of effective disks left in the array. In addition, this paper also shows how to use the strategy in 3×3 2D-RAID array and the transformation of 2D array to 1D array. The analysis on reliability shows that such strategy is able to increase the MTTDL (Mean time to data loss) of the array by 2 to 3 order of magnitudes.