A Non-MDS Erasure Code Scheme for Storage Applications

Abstract

This paper investigates the use of redundancy and self repairing against node failures indistributed storage systems using a novel non-MDS erasure code. In replication method, accessto one replication node is adequate to reconstruct a lost node, while in MDS erasure codedsystems which are optimal in terms of redundancy-reliability tradeoff, a single node failure isrepaired after recovering the entire stored data, thereby consuming more repair bandwidth.The current paper aims at investigating a new type of erasure codes with a reduced repairbandwidth as compared to conventional MDS erasure codes. Specifically, we propose a non-MDS using the proposed code a single node failure can be repaired through connecting to only threenodes which gives the ability to reduce the repair bandwidth comparing to MDS codes.(2k k) code that tolerates any three node failures and more importantly, it is shown