Speaker: Mark S. Squillante, IBM T.J. Watson
Date: December 2, 1999
Web Traffic Modeling and Server Performance Analysis
The control and optimization of various performance measures in high-volume Web sites requires a fundamental understanding of the user request patterns and the performance impact of such traffic patterns. In this talk we will present a study of the request patterns for a dynamic and heavily-accessed Web server environment, and their impact on Web server performance. Our study is based on the data from the official Web site for the 1998 Winter Olympic Games in Nagano, Japan. Our analysis of the user request data illustrates traffic patterns that exhibit both light-tailed and heavy-tailed behaviors. We then analyze the waiting-time behavior of the geographically-distributed Web server system under such user request patterns in a completely general setting, thus gaining a better understanding of the latency encountered by user requests, a key measure of quality of service (QoS).
This is joint work with David Yao and Li Zhang.
Mark Squillante received the Ph.D. degree in computer science from the University of Washington, Seattle, WA, in 1990. He has been a Research Staff Member at the IBM Thomas J. Watson Research Center, Yorktown Heights, NY, since 1991, and an adjunct faculty member of the Department of Computer Science at Columbia University, New York, NY, from 1991 through 1998. He currently manages the Systems Design, Analysis and Theory department. From 1982 to 1985 he was a Member of the Technical Staff at Bell Telephone Laboratories, Murray Hill, NJ. His research interests concern the fundamental theory, analysis, design and implementation of computer systems, including mathematical analysis, modeling and optimization; scheduling theory, optimization and practice; algorithms; and distributed, parallel and meta-computing systems. He is a member of ACM, IEEE, IFIP W.G. 7.3, INFORMS and SIAM. He is an editor of Performance Evaluation.