Cooperative Regenerating Codes for Distributed Storage Systems
Distributed storage system provides a scalable solution to the ever-increasing demand of reliable storage. The storage nodes are distributed in different geographical locations, and in case some disastrous event happened to one of them, the source data would remain intact. There are two common strategies for preventing data loss against storage node failures. The first one, employed by the current Google file system, is data replication. Although replication-based scheme is easy to manage, it has the drawback of low storage efficiency.