Distributed Systems (2nd Edition) [Hardcover]

Sape Mullender (Editor)

Book Description

Series: Acm Press Frontier Series | Publication Date: July 1993

Technological progress in distributed computing systems continues at a rapid pace, and this second edition has been completely revised to reflect the current state of the art. Examples and case studies of commericial and experimental systems are provided by a distinguished author team, whose work reflects the cutting edge of modern developments in the field.

Editorial Reviews

From the Inside Flap

This book collects the lecture notes used at the Advanced Courseon Distributed Systems which has now been held five times--Artic'88 in Tromso, Norway, Fingerlakes '89 in Ithaca, NY, Bologna '90in Italy, Karuizawa '91 in Japan and Lisboa '92 in Portugal. Itsinauguration will be at the sixth advanced course, Seattle '93 inSeattle, WA.

The first edition was based on the lecture notes used at Artic'88 and Fingerlakes '89. Since then, the course has evolved andsteadily improved: coordination between the lecturers haseliminated duplication of material; it has produced agreement onterminology; and the course now reflects a more balanced view ofboth 'theoretical' and 'practical' aspects of research indistributed systems.

As a consequence, the material in this second edition is almostentirely new. Only one chapter, in fact, has survived unscathedfrom the first edition. This book, I believe, has now become acoherent treatment of an increasingly important research area.

DISTRIBUTED SYSTEMS is intended for people in universities andindustry interested in distributed systems. Teachers can usethis book torefresh their knowledge of distributed systems. Graduate students can use it to get acquainted with the field andas a reference to other work. Distributed systems designers andbuilders can use it as a source book when trying to solveproblems they encounter in their work. It can be (and is, bysome of the authors) used in post-graduate distributed systemscourses. It is aimed at graduate students in computer science: the reader should have a good general knowledge of computerorganization, computer networks, and operating systems.

The book can be roughly partitioned into seven parts--anintroduction (Chapters 1-3), a part on theoretical foundations(Chapters 4-8), a part on Communication (Chapters 9-11), a parton file systems and database systems (Chapters 12-14), a singlechapter on kernel support (Chapter 15), a part on real-timesystems (Chapters 16-19), and a part on security (Chapters20-21).

In Chapter 1, Schroeder explains what distinguishesd distributedsystems from other systems, and he discusses properties ofdistributed systems and techniques for achieving them. Then, inChapter 2, Schneider discusses the importance of making modelsofthe behaviour of distributed systems and the failures that needto be masked by them. Chapter 3, by Weihl, introduces aspecification method for concurrent systems and then derivesspecifications for the alternat-bit protocol, mutexes andcondition variables, remote procedure call, and a name service.

The second group of chapters covers theoretical foundations ofdistributed and fault-tolerant algorithms. Chapter 4, byBabaoglu and Marzullo, describes how one can deduce, through theexchange of messages, what the state is of a distributed system. It introduces consistent cuts and snapshots, causality, logicaland vector clocks, and characterizes stable and unstablepredicates. Chapter 5, by Hadzilacos and Toueg, introducesvarious broadcast mechanisms. Using broadcasts, it is possibleto provide to a set of cooperating processes a consistent view ofa distributed computation. Broadcast protocols play an importantrole in the following three chapters. Chapter 6, by Babaoglu andToueg, gives protocols for non-blocking atomic commitment,Chapter 7, by Schneider, describes one well-known technique forturning a deterministic program running on a single computer intoa distributed one running on a set of computers so that itbecomes fault tolerant, and Chapter 8, by Budhiraja, Marzullo,Schneider and Toueg, shows another technique for transforming aprogram into a fault-tolerant one.

In Chapter 9, Mullender discusses practical aspects of protocoldesign for interprocess communication in distributed systems. Attention is given to achieving at-most-once behaviour ofremote-invocation protocols in unreliable networks. Remoteprocedure call is also described in Chapter 9. Lampson appliesthe specification of methods of Chapter 2 to the derivationof aprotocol for connection establishment in Chapter 10. In Chapter11 Rodeheffer and Schroeder examine a particular distributedapplication--building up the routing tables for a high-speedswitch-based local-area network.

Needham discusses issues associated with naming in Chapter 12 andshows that building naming databases for very large systems is afar from trivial. Weihl describes the protocols used forimplementing atomic transactions in distrubed systems subject toprocessor and network failures in Chapter 13. AndSatyanarayanan, in Chapter 14, shows how empirical data about howfile systems are used can be used to advantage in theimplementation of distributed file systems. One file system,Coda, is also described there in detail.

In Chapter 15, Mullender uses the design of the Amoebadistributed system to discuss issues of distributed system kerneldesign and discuss techniques for memory management and processmanagement.

In Chapters 16-19, Kopetz and Verissimo cover the subject ofreal-time systems. Chapter 16 introduces real time and theconcept of dependability. Chapter 17 then covers communicationin real-time systems, while Chapter 18 discusses scheduling ofreal-time processes. Two case studies, MARS and DELTA-4, are thesubject of Chapter 19.

The final part of the book is about security and cryptography. Needham explains how cryptography is used to build secureauthenticated communication channels in Chapter 20. In Chapter21, Lampson unfolds a security architecture for distributedsystems, in which he covers security aspects ranging from securebootstrapping to authenticating a party on the other side of theplanet. 0201624273P04062001

From the Back Cover

Technological progress in distributed computing systems continuesat a rapid pace, and this second edition has been completelyrevised to reflect the current state of the art. Examples andcase studies of commericial and experimental systems are providedby a distinguished author team, whose work reflects the cuttingedge of modern developments in the field.

Highlights of the new edition include:

• A revised structure, providing an easy learning curve from fundamental concepts to advanced applications
• Many new case studies, including studies of real-timedistributed systems
• A strong focus on giving solutions to problems commonlyfaced by system designers

It is designed to be equally valuable for practitionersdeveloping new generations of communication networks, and forstudents taking advanced courses in distributed computing.

Sape Mullender, Professor of Computer Science at the Universityof Twente in the Netherlands, has assembled a world-class teamfrom the world's major centres of excellence in distributedsystems.

0201624273B04062001

See all Editorial Reviews

Product Details

• Hardcover: 595 pages
• Publisher: Longman Group United Kingdom; 2 Sub edition (July 1993)
• Language: English
• ISBN-10: 0201624273
• ISBN-13: 978-0201624274
• Product Dimensions: 9.1 x 6.1 x 1.4 inches
• Shipping Weight: 2.2 pounds
• Average Customer Review: 4.0 out of 5 stars  See all reviews (2 customer reviews)
• Amazon Best Sellers Rank: #517,840 in Books (See Top 100 in Books)

Customer Reviews

Most Helpful Customer Reviews

18 of 18 people found the following review helpful:

5.0 out of 5 stars Best book on Disributed Systems Principles and Architecture, January 24, 2004

By A Customer

This review is from: Distributed Systems (2nd Edition) (Hardcover)

Distributed Systems (2nd Edition) by Sape Mullender is easily the best book available on Distributed Systems. Many authors, yet so coherent and authoritive and well written. Makes reasonably easy yet rigorous reading. It is about 10 years old, but still very up-to-date in terms of principles. For more up-to-date other (also very good) alternatives see either or both of:

1. Distributed Systems - Couloris

OR

2. Distributed Systems - A. Tanenbaum

These are not nearly as deep as Sape Mullender, but quite thorough. Mullender covers distributed real time systems better (given short shrift by the others) and is more rigorous. But due to its age Mullender has hardly any coverage of distributed object middleware (e.g. DCOM, CORBA, RMI) and multi-media systems.

To get or good handle or distrbuted systems design or research however you also need a distrbuted algorithms text. You may want to search on the following authors:

Vijay Garg, Gerard Tel, Nancy Lynch.

Gerard Tel's is the most thorough in terms of problem/topic coverage and mots intuitive. But if you want a very easy to read book, non too mathematical with algorithms in a real programming language (JAVA) to adapt to build your own Distrbuted Software choose: Vijay Garg's Concurrent and Distributed Computing in Java which covers both Concurrent and Distributed Programming algorithms in Java.

Another important book which looks at distributed systems from a thorough, modern distributed objects point of view is 'Engineering Distributed Objects' by Wollgang Emmerich. It covers the distributed objects middleware raison d'etre, principles, and the Java RMI, CORBA and DCOM/COM+ approaches and architectures in an excellent, readable fashion.

2 of 5 people found the following review helpful:

3.0 out of 5 stars its ok, November 26, 2009

By 

Ali Ezad Abbas (Atlanta, GA) - See all my reviews
(REAL NAME)   

Amazon Verified Purchase

This review is from: Distributed Systems (2nd Edition) (Hardcover)

I used this for one of my grad courses. Not very well written but got me an A in the course so I'm not complaining :)