Transient Probability Models of Computing System Dynamic Behavior.
Abstract
The analysis of computing systems using mathematical models is most often concerned with the determination of long-term statistical trends or the prediction of steady-state probabilistic behaviour. Valuable insight into the dynamic operation of computing systems, however, is not always directly obtainable by such approaches. For this purpose, the application of certain transient probability models may be beneficial in some instances. A two-part investigation of transient probability models of computing systems is described. In the first investigation, the dynamic operation of a single peripheral processor as it performs an assigned task in a CDC 6600 computing system is studied. The transient analysis of queuing networks frequently used to model computing systems was studied in the second investigation. A probability flow technique for numerically analyzing networks of single queues is described. Models formulated by the method presented provide an enhanced level of detail not available in conventional queuing models. Three case studies are included to demonstrate the capabilities of this modeling technique.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1975
- Accession Number
- ADA013806
Entities
People
- Baxter F. Womack
- Wilfred S. Litzler
Organizations
- University of Texas at Austin