Algorithms for Scalable Storage Servers

We survey a set of algorithmic techniques that make it possible to build a high performance storage server from a network of cheap components. Such a storage server offers a very simple programming model. To the clients it looks like a single very large disk that can handle many requests in parallel with minimal interference between the requests. The algorithms use randomization, redundant storage, and sophisticated scheduling strategies to achieve this goal. The focus is on algorithmic techniques and open questions. The paper summarizes several previous papers and presents a new strategy for handling heterogeneous disks