Terms and Conditions:
Edward D. Lazowska wrote:The publisher has returned the copyright to the authors, who make the material available here for viewing or downloading, in Adobe Acrobat PDF format (free Acrobat reader available here).
This book is written for computer system performance analysts. Its goal is to teach them to apply queueing network models in their work, as tools to assist in answering the questions of cost and performance that arise throughout the life of a computer system.
The approach to the subject arises from the authors' collective experience in contributing to the theory of queueing network modelling
, in embodying this theory in performance analysis tools, in applying these tools in the field, and in teaching computer system analysis using queueing network models in academic and industrial settings.
Case studies appear throughout the book. They are included to illustrate various aspects of computer system analysis using queueing network models. They should not be misconstrued as making general statements about the relative performance of various systems; the results have significance only for the specific configurations and workloads under con sideration.
The authors have summarized a number of important modelling techniques in the form of "Algorithms". The intention is to provide enough informa tion that the reader can understand fully the essential aspects of each technique. This book omits details of significance to the implementation of a technique when these details might obscure the more fundamental concepts.
Practicing computer system analysts are relatively skilled in techniques such as workload characterization, system measurement, interpretation of performance data, and system tuning, and are at least acquainted with basic statistical methods and with simulation. Each of these subjects is well represented in the existing literature, and is given short shrift in the present book. Much interesting and important research work concerning queueing network modelling also is given short shrift; this book discusses the one approach to each problem that is best suited for application. For readers who desire to pursue a topic in greater detail, each chapter concludes with a brief discussion of the relevant literature.