The objective is to obtain a synthetic workload based on a real workload generated from a few client requests to a local web server, having already identified the work points of the web server itself. An overview is then given of the different phases implemented.
This phase consists of evaluating throughput and response time under varying loads, and identifying points of usable and knee capacity.
This phase consists of system workload characterisation, in terms of low-level and high-level parameters, and also verifying that the synthetic workload is representative of the real workload.
This phase consists of studying the influence of various factors on response time, and assessing their significance and importance.