Computer Sciences Dept.

CS 547: Computer System Modeling Fundamentals

Spring 2009
Instructor: Prof. Mary Vernon

Tues. & Thurs. 11:00a-12:15p in 1289 Computer Sciences


Texts and Readings

Course Text:

  • [LZGS] E. D. Lazowska, J. Zahorjan, G. S. Graham, K. C. Sevcik, Quantitative System Performance, Prentice Hall, 1984.

    This text is available online here.

    We will cover chapters 1-6 in the above text. Note that the copyright date is 1984. Hence the example systems modeled in the text are hugely out of date(!). On the other hand, the modeling techniques covered in the text are timeless. We will cover modern applications in class, illustrating the interesting point that the same techniques can be applied to current system performance analysis and design.


Other Course Texts:

  • [AA] A. O. Allen, Probability, Statistics, and Queueing Theory with Computer Science Applications, Second Edition, Academic Press, 1990.

  • [SR] Sheldon Ross, A First Course in Probability, Fifth Edition, Prentice Hall, 2002.

  • [LK] L. Kleinrock, Queueing Systems Volume I: Theory, Wiley Interscience, 1975.

Copies of these other course texts will be on reserve at the Wendt Library and/or will be available to check out from the CS 547 TA. Relevant sections to read will be provided in the Schedule of Topics.


Tutorials

  • Matlab Tutorial

    You can use Matlab or excel or any other software to plot your results in this course. Note that there are many other Matlab tutorials available on the web.

  • Queueing Theory Tutor

    This on-line tutorial contains a graphical simulation of the M/M/1 queue. In this queue, customers arrive to a server at random points in time and service times are moderately bursty. The simulation illustrates the burstiness of random arrivals as well as the "busy period" behavior. Note the time it takes for the queue to empty when a backlog builds up.


Additional Resources


Application Papers

Below are papers in the literature that describe analytic performance models. A few of these models will be covered in the course.

  • Applications of probability and random variables:

    J. Padhye, V. Firoiu, D. Towsley, J. Kurose, "Modeling TCP Throughput: A Simple Model and its Empirical Validation", Proc. ACM SIGCOMM '98, August 1998.

    C. Samios and M. K. Vernon, "Modeling the Throughput of TCP Vegas", Proc. ACM SIGMETRICS 2003, San Diego, June 2003.

  • Customized queueing models:

    Bier, G.E., and M. K. Vernon, "Measurement and Prediction of Contention in Multiprocessor Operating Systems with Scientific Application Workloads", Proc. 1988 Int'l. Conf. on Supercomputing, St. Malo, France, July 4-8, 1988, pp. 9-15.

    R. Berry and K. M. Chandy, "Performance Models of Token Ring Local Area Networks", Proc. ACM SIGMETRICS '83, Minneapolis, August 1983, p. 266-274.

    J. Zahorjan, "Workload Representations in Queueing Models of Computer Systems", Proc. ACM SIGMETRICS '83, Minneapolis, August 1983.

    Tan, H., D. L. Eager, M. K. Vernon, and H. Guo, "Quality of Service Evaluations of Multicast Streaming Protocols", Proc. ACM SIGMETRICS 2002, Marina del Rey, CA, June 2002.

  •  
    Computer Sciences | UW Home