Скачать презентацию Course Software Engineering II academic year 2007 -2008 Скачать презентацию Course Software Engineering II academic year 2007 -2008

ba93c4d7b7b8379f03a035c456e459d4.ppt

  • Количество слайдов: 14

Course: Software Engineering II academic year: 2007 -2008 Course Web-site: [www. di. univaq. it/cortelle/] Course: Software Engineering II academic year: 2007 -2008 Course Web-site: [www. di. univaq. it/cortelle/] Lecture 22: Performance Evaluation with SHARPE Lecturer: Catia Trubiani Computer Science Department University of L'Aquila - Italy Catia. trubiani@di. univaq. it

Copyright Notice » The material in these slides may be freely reproduced and distributed, Copyright Notice » The material in these slides may be freely reproduced and distributed, partially or totally, as far as an explicit reference or acknowledge to the material author is preserved. SEA Group 2

Roadmap Performance modeling Generalized Stochastic Petri Nets (GSPNs) SEA Group 3 evaluation Symbolic Hierarchical Roadmap Performance modeling Generalized Stochastic Petri Nets (GSPNs) SEA Group 3 evaluation Symbolic Hierarchical Automated Reliability and Performance Evaluator (SHARPE)

Software Performance Software developers’ world (vocabulary) is intrinsically distant from performance analysts’ one. SEA Software Performance Software developers’ world (vocabulary) is intrinsically distant from performance analysts’ one. SEA Group 4 APPLICATION MODEL APPLICATION PERFORMANCE MODEL

Performance Notations Queueing Network s Markov processe s Generalized Stochastic Petri Nets Stochastic Process Performance Notations Queueing Network s Markov processe s Generalized Stochastic Petri Nets Stochastic Process Algebras Simulation SEA Group 5

SHARPE: screenshot Symbolic Hierarchical Automated Reliability and Performance Evaluator (SHARPE) SEA Group 6 SHARPE: screenshot Symbolic Hierarchical Automated Reliability and Performance Evaluator (SHARPE) SEA Group 6

SHARPE evaluation Steps for Performance evaluation with SHARPE: § Design of a generalized stochastic SHARPE evaluation Steps for Performance evaluation with SHARPE: § Design of a generalized stochastic Petri net model § Definition of constants, variables and functions § Analysis of the model § Plotting the results of a model SEA Group 7

1 - Sharpe design § Design of a generalized stochastic Petri net model producer 1 - Sharpe design § Design of a generalized stochastic Petri net model producer SEA Group 8 tasks consumer

2 - Sharpe definitions § Definition of constants, variables and functions producer tasks consumer 2 - Sharpe definitions § Definition of constants, variables and functions producer tasks consumer lambda, buffer. Size, mu SEA Group 9

3 - Sharpe analysis § Analysis of the model § Steady-state average number of 3 - Sharpe analysis § Analysis of the model § Steady-state average number of tokens in the given place § Steady-state THROUGHPUT for a transition § Steady-state UTILIZATION for a transition § Steady-state probability that the given place is empty SEA Group 10

4 - Sharpe plotting § Plotting the results of a model different rates for 4 - Sharpe plotting § Plotting the results of a model different rates for the transition “consumer” SEA Group 11

Exercise Comparison of two simple models: (1) producer tasks consumer 2 (2) producer SEA Exercise Comparison of two simple models: (1) producer tasks consumer 2 (2) producer SEA Group 12 consumer 1 tasks consumer 3

System Performance Model Results § Plotting the results of the models… Throughput of consumers System Performance Model Results § Plotting the results of the models… Throughput of consumers in comparison, what’s the meaning of gaps? SEA Group 13

Basic Readings » [Performance evaluation] http: //www. ee. duke. edu/~kst/ Software packages, Sharpe, Tool Basic Readings » [Performance evaluation] http: //www. ee. duke. edu/~kst/ Software packages, Sharpe, Tool and Interface Manual Robin Sahner, Kishor S. Trivedi, Antonio Puliafito “PERFORMANCE AND RELIABILITY ANALYSIS OF COMPUTER SYSTEMS” SEA Group 14