S2-RAID: A new RAID architecture for fast data recovery

Abstract

As disk volume grows rapidly with terabyte disk becoming a norm, RAID reconstruction time in case of a failure takes prohibitively long time. This paper presents a new RAID architecture, S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -RAID, allowing the disk array to reconstruct very quickly in case of a disk failure. The idea is to form skewed sub RAIDs (S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -RAID) in the RAID structure so that reconstruction can be done in parallel dramatically speeding up data reconstruction time and hence minimizing the chance of data loss. To make such parallel reconstruction conflict-free, each sub-RAID is formed by selecting one logic partition from each disk group with size being a prime number. We have implemented a prototype S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -RAID system in Linux operating system for the purpose of evaluating its performance potential. SPC IO traces and standard benchmarks have been used to measure the performance of S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -RAID as compared to existing baseline software RAID, MD. Experimental results show that our new S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -RAID speeds up data reconstruction time by a factor of 3 to 6 compared to the traditional RAID. At the same time, S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> -RAID shows similar or better production performance than baseline RAID while online RAID reconstruction is in progress.