Predicting the Performance of Component-based Applications: A Quantitative Approach

Queueing Theory has many applications in reality. Models based on Queueing Theory become important tools in the design and analysis of computer systems. Queueing network models in particular have been often adopted as models for the evaluation of software performance.

In this talk, we first introduce some background knowledge of Queueing Theory, such as Qeueing Nework (QN) models and the analysis techniques. Then we briefly discuss the state of art work in software performance engineering towards predicting software performance at the design level. QN models have been applied to our PP-CB approach, which has been devised to predict the performance of component-based applications. The development of performance models using QN for a proposed architecture design of J2EE-based applications will be presented in details.

Jenny Liu
Researcher
NICTA Empirical Software Engineering Program

Date:Thu Nov 11 2004
Time: 1pm to 2pm
Location: Level 1 Seminar Room, CSE Bldg (K17), UNSW Kensington Campus

Return to index


Last updated by rhuuck at Fri Nov 12 14:48:42 2004 GMT+1100