Using Distributed Data Structures for Constructing Cluster-Based Services

The emergence of the Internet as the global, ubiquitous networking infrastructure, and the accompanying Internet computing model, where much of the computing takes place on servers rather than local machines, is driving a new class of Internet service applications. While these services promise unprecedented computing power, connectivity, convenience, and low cost, their construction is a challenging exercise—examples of costly failures abound (e.g., [11, 19, 22]).