Course on Distributed Systems

Additionaly, see also the course home-page (warning: many parts are in Danish).

Distributed Systems

This course introduces basic concepts and techniques for distributed systems, i.e. systems in which a set of work stations and other computers are connected in a network - allowing them to share common resources and facilities, such as files, printers, and internet access. The computing equipment at DAIMI is an excellent example of a distributed system.

Contents

The course describes different kinds of system architectures, e.g. the client/server model. There is an introduction to protocols and networks, including local area networks. It is shown how different processes can interact by means of remote procedure calls. It is discussed how to obtain exclusive access to critical sections, how to elect coordinators, and how to use atomic transactions. As a concrete example of a part of a distributed system, we discuss concepts and techniques to implement file systems, including files that can be used by several users simultaneously. Finally, we look at different examples of distributed operating systems.

It is much more difficult to design, validate, and test distributed systems than traditional sequential programs. It is easy to introduce deadlocks and to end up in situations, where the result of a program execution (in an undesirable way) depends on the order in which the different operations are executed. High-level Petri nets are introduced as a modelling language, adequate to describe and investigate communication, synchronisation and resource sharing.