This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Performance Evaluation Techniques Summer 2005 Workload Modeling and Generation Dr.Ing. Andreas Willig Telecommunication Networks Group (TKN) Technical University Berlin [email protected] June 26, 2005 Andreas Willig Module Summary • In this module we discuss basics of workload (and error) modeling as well as selected examples of specific models • Acknowledgements: some slides have been inspired by Holger Karls lecture “Leistungsbewertung und Simulation” http://wwwcs.unipaderborn.de/cs/agkarl/teaching/ws0405/vllbsim.html Workload Modeling and Generation, slide 1 Andreas Willig Overview Overview • Overview on Workload Modeling and Generation • Tracedriven Simulation • Fundamentals of Synthetic Workloads • IID Sequences of Random Variates from wellknown Distributions • IID Sequences of Random Variates from Empirical Distributions • Selected Traffic Models based on Stochastic Processes Workload Modeling and Generation, slide 2 Andreas Willig Overview on Workload Modeling and Generation Overview on Workload Modeling and Generation • The workload is a key component of any performance study • Requirements: – Specified in terms of services of the system under study: types, interarrival times, and resource demands of service requests – Representative – Relevant: * To find the maximum throughput you have to saturate the system * To test a congestion control algorithm you need congestion * To test convergence speed of an adaptive algorithm you need workload variations – Reproducible Workload Modeling and Generation, slide 3 Andreas Willig Overview on Workload Modeling and Generation Overview on Workload Modeling and Generation II • “Realworld” workloads typically have random components, e.g. random interarrival times • Two fundamental types of workloads: – “Real workloads” / traces – Synthetic workloads: artificially generated stochastic processes Workload Modeling and Generation, slide 4 Andreas Willig Overview Overview • Overview on Workload Modeling and Generation • Tracedriven Simulation • Fundamentals of Synthetic Workloads • IID Sequences of Random Variates from wellknown Distributions • IID Sequences of Random Variates from Empirical Distributions • Selected Traffic Models based on Stochastic Processes Workload Modeling and Generation, slide 5 Andreas Willig Tracedriven Simulation Tracedriven Simulation • Basic idea: – Conduct a measurement at a real system, observe the service requests (type, arrival times, resource demands, further parameters) – Save all this into a logfile / tracefile – Feed the simulation from the tracefile • Example: to obtain an Ethernet workload, you can use tools like tcpdump or ethereal • The real system must be available and accessible to obtain traces Workload Modeling and Generation, slide 6 Andreas Willig Tracedriven Simulation Advantages of Tracedriven Simulation • The workload is “realistic”, which increases credibility of your results...
View
Full Document
 Spring '09
 H.Y
 Normal Distribution, The Land, Probability theory, random variates, workload modeling, Andreas Willig

Click to edit the document details