We continue our study of algorithms for distributed systems by looking at how processes can coordinate their actions in a distributed system. We look at how to implement algorithms for ensuring mutual exclusion and how to perform elections (choosing one process among a set of peer processes to perform a specific task). An important expect of this lecture is how we evaluate the different algorithms with respect to performance and fault tolerance.
Literature[DS] Chapter 12.1-12.3
LinksA nice java applet showing the behaviour of the bully algorithm.
Exercises
DS 12.4, 12.5, 12.6, 12.7, 12.8