Storage codes: Managing big data with small overheads

Abstract

Erasure coding provides a mechanism to store data redundantly for fault-tolerance in a cost-effective manner. Recently, there has been a renewed interest in designing new erasure coding techniques with different desirable properties, including good repairability and degraded read performance, or efficient redundancy generation processes. Very often, these novel techniques exploit the computational resources available `in the network', i.e., leverage on storage units which are not passive entities supporting only read/write of data, but also can carry out some computations. This article accompanies an identically titled tutorial at the IEEE International Symposium on Network Coding (NetCod 2013), and portrays a big picture of some of the important processes within distributed storage systems, where erasure codes designed by explicitly taking into account the nuances of distributed storage systems can provide significant performance boosts.