Fast, on-line failure recovery in redundant disk arrays

The authors describe and evaluate two algorithms for performing online failure recovery (data reconstruction) in redundant disk arrays. An implementation of disk-oriented reconstruction, a data recovery algorithm that allows the reconstruction process to absorb essentially all the disk bandwidth not consumed by the user processes, is presented, and this algorithm is compared to a previous-proposed parallel stripe-oriented approach. The disk-oriented approach yields better overall failure-recovery performance. Performance is evaluated via detailed simulation on two different disk array architectures: the RAID level five organization and the declustered parity organization. The benefits of the disk-oriented algorithm can be achieved using controller or host buffer memory no larger than the size of three disk tracks per disk in the array. The tradeoffs involved in selecting the size of the disk accesses used by the failure recovery process are also investigated.