Performance pathways

ABSTRACT

We review the status of some recent results in the performance analysis of computer systems which are intrinsically unstable due to the presence of more than one stable operating state. In particular, we consider bistable computer systems which possess two stable states: the typical operating point and an another stable point, concomitant with degraded system performance.

REFERENCES

• 1: {Agarwal et al., 1986} Agarwal, A., Horowitz, M., and Hennessy, J., "An Analytical Cache Model," <i>Tech. Report CSL-TR-86-304, Stanford Univ.</i>, September 1986.
• 2: Subhash C. Agrawal, Metamodeling: a study of approximations in queueing models, Massachusetts Institute of Technology, Cambridge, MA, 1985
• 3: {Cottrell at al., 1983} Cottrell, M., Fort, J-C., and Malgouyres, G., "Large Deviations and Rare Events in the Study of Stochastic Algorithms," <i>IEEE Trans. Automatic Control</i>, <b>AC-28</b>, 907, September 1983.
• 4: {Courtois, 1977} Courtois, P-J., <i>Decomposability</i>, Academic Press, New York, 1977.
• 5: {Denning, 1980} Denning, P. J., "Working Sets Past and Present," <i>IEEE Trans. Soft, Eng.</i>, <b>SE-6</b> no. 1, 64, January 1980.
• 6: {Ellis, 1985} Ellis, R. S., <i>Entropy, Large Deviations, and Statistical Mechanics</i>, Springer-Verlag, 1985.
• 7: {Feynman, 1981} Statistical Mechanics, W. A. Benjamin Inc., New York, 1981.
• 8: {Feynman, 1988} Feynman, R. P., "An Outsider's Inside View of the Challenger Inquiry," <i>Physics Today</i>, February, 1988.
• 9: G. C. Fox , W. Furmanski, Load balancing loosely synchronous problems with a neural network, Proceedings of the third conference on Hypercube concurrent computers and applications: Architecture, software, computer systems, and general issues, p.241-278, January 19-20, 1988, Pasadena, California, United States [doi>10.1145/62297.62327]
• 10: {Gibbens et al., 1988} Gibbens, R. J., Hunt, P. J., and Kelly, F. P., "Bistability in Communication Networks," <i>Report to the Stochastic Networks Group</i>, December, 1988.
• 11: {Greenberg &amp; Weiss, 1986} Greenberg, A., and Weiss, A., "An Analysis of ALOHA Systems via Large Deviations," AT&amp;T Bell Labs., preprint, July 1986.
• 12: N. Gunther, Path integral methods for computer performance analysis, Information Processing Letters, v.32 n.1, p.7-13, July 3, 1989 [doi>10.1016/0020-0190(89)90061-6]
• 13: {Gunther, 1989b} Gunther, N. J., "Performance Models for Teletraffic Networks with Dynamic Call Routing," in preparation.
• 14: N. J. Gunther , G. Shaw, Path integral evaluation of ALOHA network transients, Information Processing Letters, v.33 n.6, p.289-295, February 1990 [doi>10.1016/0020-0190(90)90211-F]
• 15: {Gunther &amp; Shaw, 1989b} Gunther, N. J. and Shaw, J. G., "Instantons and Tunneling Instabilities in Computational Systems," <i>Physical Review Letters</i>, submitted May 1989.
• 16: {Kleeman &amp; Cantoni, 1987} Kleeman, L., and Cantoni, A., "Metastable Behavior in Digital Systems," <i>IEEE Design and Test of Computers</i>, December 1987.
• 17: Randolph Nelson, Stochastic catastrophe theory in computer performance modeling, Journal of the ACM (JACM), v.34 n.3, p.661-685, July 1987 [doi>10.1145/28869.28878]
• 18: {Simon, 1983} Simon, B., "Instantons, Double Wells, and Large Deviations," <i>Bull. Amer. Math. Soc.</i>, <b>8</b>, 323, 1983.
• 19: William D. Strecker, Transient behavior of cache memories, ACM Transactions on Computer Systems (TOCS), v.1 n.4, p.281-293, Nov. 1983 [doi>10.1145/357377.357379]