Distributed systems: an algorithmic approach

Most applications in distributed computing center around a set of common subproblems. Distributed Systems: An Algorithmic Approach presents the algorithmic issues and necessary background theory that are needed to properly understand these challenges. Achieving a balance between theory and practice, this book bridges the gap between theoreticians and practitioners. With a set of exercises featured in each chapter, the book begins with background information that contains various interprocess communication techniques and middleware services, followed by foundational topics that cover system models, correctness criteria, and proof techniques. The book also presents numerous important paradigms in distributed systems, including logical clocks, distributed snapshots, deadlock detection, termination detection, election, and several graph algorithms. The author then addresses failures and fault-tolerance techniques in diverse applications, such as consensus, transactions, group communication, replicated data management, and self-stabilization. He concludes with an exploration of real-world issues, including distributed discrete-event simulation and security, sensor networks, and peer-to-peer networks. By covering foundational matters of distributed systems and their relationships to real-world applications, Distributed Systems provides insight into common distributed computing subproblems,