The Art of Computer Systems Performance Analysis

As a key step in the selection of processors, operating systems, database, systems networks, and languages, performance analysis plays an integral role in the design and procurement of new computer systems. Presenting a unique blend of measurement, simulation, and modeling methods in a straightforward, problem-oriented fashion, The Art of Computer Systems Performance Analysis is a computer professional's treasure trove for practical performance analysis techniques. Integrating essential queueing theory with simulation, measurement techniques, data analysis, and experimental design, the book is divided into the following comprehensive parts: Overview of Performance Evaluation-addresses common mistakes in the evaluation process, criteria for selecting evaluation techniques and performance metrics, and the establishment of performance requirements - Measurement Techniques and Tools-examines workload selection and characterization, monitors, capacity planning and benchmarking, data presentation, and ratio games - Probability and Statistics-discusses the summarization of measured data, the comparison of systems using sample data, and regression models - Experimental Design and Analysis-provides the maximum amount of information with the minimum number of experiments - Simulation-clarifies the analysis of results, model verification, validation, and random-number generation - Queueing Models-models and techniques for the solution of a wide variety of common performance problems. Analysis techniques are supported by more than 150 examples and case studies-all from the computer science field-and many are summarized with results in handy reference boxes. In addition, common mistakes in the application of various techniques are descussed, and a host of special tricks to demonstrate system superiority are revealed. The Art of Computer Systems Performance Analysis is an essential tool for performance analysts, designers, purchasers, and users of computer systems.

Autorentext

RAJ JAIN is a Senior Consulting Engineer in the Distributed Systems Architecture and Performance Group at Digital Equipment Corporation. With over sixteen years of experience in the field of computer systems performance, he is currently responsible for analyzing various design alternatives for DEC's networking architecture. He received the Ph.D. degree from Harvard and has taught courses on performance at Massachusetts Institute of Technology. "Dr. Jain is a Fellow of the IEEE and is listed in Who's Who in the Computer Industry, 1989."

Klappentext

The Art of Computer Systems Performance Analysis "At last, a welcome and needed text for computer professionals who require practical, ready-to-apply techniques for performance analysis. Highly recommended!" -Dr. Leonard Kleinrock University of California, Los Angeles "An entirely refreshing text which has just the right mixture of theory and real world practice. The book is ideal for both classroom instruction and self-study." -Dr. Raymond L. Pickholtz President, IEEE Communications Society "An extraordinarily comprehensive treatment of both theoretical and practical issues." -Dr. Jeffrey P. Buzen Internationally recognized performance analysis expert ". it is the most thorough book available to date" -Dr. Erol Gelenbe Université René Descartes, Paris ". an extraordinary book.. A worthy addition to the bookshelf of any practicing computer or communications engineer" -Dr. Vinton G. Cer??? Chairman, ACM SIGCOMM "This is an unusual object, a textbook that one wants to sit down and peruse. The prose is clear and fluent, but more important, it is witty." -Allison Mankin The Mitre Washington Networking Center Newsletter

Inhalt

Partial table of contents: AN OVERVIEW OF PERFORMANCE EVALUATION. Common Mistakes and How to Avoid Them. Selection of Techniques and Metrics. MEASUREMENT TECHNIQUES AND TOOLS. Types of Workloads. Workload Characterization Techniques. Monitors. Ratio Games. PROBABILITY THEORY AND STATISTICS. Summarizing Measured Data. Simple Linear Regression Models. Other Regression Models. EXPERIMENTAL DESIGN AND ANALYSIS. One-Factor Experiments. Two-Factor Full Factorial Design without Replications. Two-Factor Full Factorial Design with Replications. SIMULATION. Analysis of Simulation Results. Testing Random-Number Generators. Commonly Used Distributions. QUEUEING MODELS. Analysis of a Single Queue. Operational Laws. Convolution Algorithm. Appendix. Solutions to Selected Exercises. References. Author Index. Subject Index.