Distributed Database Systems Slides

Please register - see below.

The following are the slides that accompany Principles of Distributed Database Systems by M.T. Özsu and P. Valduriez, Prentice-Hall, 1991. Both the regular slides and student handouts (two slides-per-page) are provided. Academics can make copies of these slides for use in higher education courses provided that they register. Registration will generate an email as well as an official letter to you (to be sent by regular post) giving you permission to make copies. This letter is necessary in some countries to duplicate copyrighted material. It will also allow us to keep track of the users of these notes so that we can inform you of changes, etc. Use for all other purposes requires written permission from the authors. Please send email to ozsu@cs.ualberta.ca.

There are some errors in the book. If you have the first printing, please get a copy of the corrections. These errors are fixed in later printings, but, we have discovered other errors in later printings as well. There are links under appropriate chapters for these. All of these errors will be corrected in the second edition which is now in the works. If you discover any new ones, please send message to ozsu@cs.ualberta.ca.

Beginning Slides

These are the beginning slides which provide an overview of the entire contnets and a comprehensive list of additional references on distributed databases. Handouts

Last change: January 16, 1995

Introductory Slides

These slides correspond to Chapter 1 of the book. They define what a distributed DBMS is, places it in the general context of distributed computing systems, presents the fundamental assumptions that are made about the distributed DBMS technology and discusses the various problems that need to be addressed to make these systems a reality. The slides also provide a quick overview of the promises of this technology and the current commercial state-of-the-art. Handouts

Last change: January 16, 1995

Background Material

Not yet ready.

Handouts

Distributed DBMS Architecture

These slides correspond to Chapter 4 of the book. The discussion starts with an overview of the types of transparencies that should be provided by a distributed DBMS and how they can be provided. Then the ANSI/SPARC architecture is introduced and its extension to a distributed environment is discussed. This is a datalogical view of the architecture which is followed by a component view. The architectural discussion is motivated by a taxonomy of different ways of developing a DBMS which identifies the differences between "distributed DBMSs" and what are commonly called "distributed multidatabase systems". This section ends with a brief overview of the directory management issues. Handouts

Last change: January 20, 1995

Distribution Design

These slides correspond to Chapter 5 of the book. The first part provides an overview of the distribution design problem, placing it in its proper context. The second part of the notes discuss fragmentation. The algorithms for horizontal, derived horizontal and vertical fragmentation are given and their correctness proven. The final part of the notes discuss data allocation. Handouts

Corrections to Chapter 5 are here.
Last change: January 20, 1995

Semantic Data Control

These slides correspond to Chapter 6 of the book. There are three issues that are covered: view management, data protection and integrity enforcement. The part on view management is important for some of the discussion later on (query modification concept), so it may be wise to include it under some other topic (perhaps when the ANSI sparc architecture is introduced during the Architecture discussion) even if this part is skipped in a course. Handouts

Last change: January 21, 1995

Distributed Query Processing

These notes correspond to Chapters 7-9 of the book. Following an introduction of the problem and a discussion on the importance of query optimization, a query processing methodology is introduced consisting of four steps. The rest of the notes cover these steps in detail. Handouts

Last change: January 21, 1995

Distributed Transaction Processing

These notes correspond to Chapters 10-12 of the book. The first part of the notes introduce the transaction concept, what a schedule (history) is, and what different transaction models are. The second part concentrates on concurrency control issues and the last part on reliability issues. There is an error in Section 12.6.2 of the book in the subsections which describe termination and recovery protocols. Handouts

Corrections to Chapters 10-12 are here in postscript format.
Last change: February 23, 1995

Distributed Database Operating Systems

These notes correspond directly to Chapter 13 of the book. They provide an overview of the problems in interfacing distributed DBMSs with distributed OSs and then discusses some of the issues. It ends with a look at the various architectural options. Handouts

Last change: February 20, 1995.

Interoperability and Multidatabases

These notes roughly correspond to Chapter 14 of the book. This chapter will undergo significant change for the second edition and this chapter contains some material (such as the standards activities, work on Pegasus, etc) that does not exist in the current form of Chapter 14. In a sense, the notes are currently a mixture of what is in the book and what is not. This set of notes is likely to undergo significant change, so watch out for the last change date.Handouts

Last change: February 10, 1995

Parallel Database Systems

These notes approximately correspond to Sections 15.1 and 15.2 of the book. The correspondence is only approximate since this part of the book is undergoing heavy revision for the second edition which is now underway. In the second edition, these section, together with Section 15.3 will become a separate chapter with more depth. The notes will need to be revised at that time.Handouts

Last change: February 14, 1995

Distributed Object Management

Not yet ready.

Handouts


February, 1995