Abstract
A typical performance evaluation consists of the identification of resources, the definition of system boundaries, the measurement of external and internal performance variables, and finally the interpretation of data and projection of system performance to hypothetical environments. These projections may be used to estimate the cost savings to be expected when changes are made to the system.
The fundamental external performance measures such as response time and thruput are intimately related, but may be defined differently depending on how the system is defined. They can be analyzed with respect to the internal performance measures (such as activities, queue lengths and busy times) by applying one or more interpretations such as: absolute utilizations, normalized busy times, system profiles, analysis of response, workload relaxation, and resource consumption hyperplanes. These models, which are generally free of assumptions regarding interarrival and service time distributions, can be adjusted to represent potential changes to the system. Then the interpretations may be used to evaluate the predicted external performance measures.
- 1 Bell, T.E., Boehm, B.W., and Watson, R.A., "Framework and Initial Phases for Computer Performance Improvement", Proceedings, FJCC, 1972. pp. 1141-1154.Google ScholarDigital Library
- 2 Cockrum, J.S. and Crockett, E.D., "Interpreting the Results of a Hardware Systems Monitor", Proceedings SJCC, 1971, pp. 23-38.Google ScholarDigital Library
- 3 IBM Data Processing Techniques Manual - Analysis of Some Queuing Models in Real-Time Systems, Manual No. F20-0007-1.Google Scholar
- 4 Kimbleton, Stephen, R., "Performance Evaluation - A Structured Approach", Proceedings SJCC, 1972, pp. 411-416.Google Scholar
- 5 Kobayashi, Hisashi, "Some Recent Progress in Analytic Studies of System Performance", IBM Technical Report RC 3990 (#17999), 1972.Google Scholar
- 6 Lassettre, L.R. and Scherr, A.T., "Modelling the Performance of the OS/360 Time Sharing Option (TSO)", Statistical Computer Performance Evaluation, Academic Press (1972), pp. 57-72.Google Scholar
- 7 Mendenhall, W., Introduction to Linear Models and the Design and Analysis of Experiments, Duxbury Press, 1968.Google Scholar
- 8 "Predicting the Effects of Hardware Changes", EDP Performance Review, Vol. 2, No. 2, February 1974, pp. 1-5.Google Scholar
- 9 Racite, M.P., "The Use of Pure and Modified Regression Techniques for Developing Systems Performance Algorithms," Statistical Computer Performance Evaluation, Academic Press (1972), pp. 347-369.Google Scholar
- 10 Stimler, S. and Brons, K.A., "A Methodology for Calculating and Optimizing Real-time System Performance", Communications of the ACM, Vol. 11, No. 7, July 1968, pp. 509-516. Google ScholarDigital Library
- 11 Traub, Arthur C. and Zachman, William F, "A GPSS Model of a Complex On-Line Computer System", Proceedings, Symposium on the Simulation of Computer Systems, 1973, pp. 17-20. Google ScholarDigital Library
- 12 Yeh, A.C., "An Application of Statistical Methodology in the Study of Computer System Performance", Statistical Computer Performance Evaluation, Academic Press (1972), pp. 287-327.Google Scholar
Index Terms
- Interpretive analysis of computer system performance
Recommendations
Interpretive analysis of computer system performance
SIGMETRICS '74: Proceedings of the 1974 ACM SIGMETRICS conference on Measurement and evaluationA typical performance evaluation consists of the identification of resources, the definition of system boundaries, the measurement of external and internal performance variables, and finally the interpretation of data and projection of system ...
Performance analysis of a telephone system with both patient and impatient customers
In this paper, the performance of a telephone system in which there are both patient and impatient customers is studied. Depending on the waiting time in the buffer, a customer may become impatient and therefore the call is incomplete. An impatient ...
Comments