Elements of Distributed Computing
May 2002, Wiley-IEEE Press
As distributed systems become increasingly available, the need for a fundamental discussion of the subject has grown. Designed for first-year graduate students and advanced undergraduates as well as practicing computer engineers seeking a solid grounding in the subject, this well-organized text covers the fundamental concepts in distributed computing systems such as time, state, simultaneity, order, knowledge, failure, and agreement in distributed systems.
Departing from the focus on shared memory and synchronous systems commonly taken by other texts, this is the first useful reference based on an asynchronous model of distributed computing, the most widely used in academia and industry. The emphasis of the book is on developing general mechanisms that can be applied to a variety of problems. Its examples-clocks, locks, cameras, sensors, controllers, slicers, and synchronizers-have been carefully chosen so that they are fundamental and yet useful in practical contexts.
The text's advantages include:
- Emphasizes general mechanisms that can be applied to a variety of problems
- Uses a simple induction-based technique to prove correctness of all algorithms
- Includes a variety of exercises at the end of each chapter
- Contains material that has been extensively class tested
- Gives instructor flexibility in choosing appropriate balance between practice and theory of distributed computing
Model of a Computation.
Verifying Clock Algorithms.
Clocks of Different Dimensions.
Mutual Exclusion: Using Timestamps.
Mutual Exclusion: Tokens and Quorums.
Drinking Philosophers Problem.
Observing Global Predicates.
Observing Conjuctive Predicates.
Control of a Distributed Computation.
Causal Message Ordering.
Synchronous and Total Message Ordering.
Computation of a Global Functon.
Repeated Global Computation of a Global Function.
Distributed Shared Memory.
Knowledge and Common Knowledge.
Consensus Under Asynchrony.
Consensus Under Synchrony.
Easy Problems in Asychronous Systems.
Checkpointing for Recovery.
Message Logging for Recovery.
Appendix: Partial Orders.