This effort
can be expressed in terms of performance measures (e.g., utilisations for resources,
response and processing times for behaviour elements) and/or costs; it starts in the
infrastructure and propagates back to the higher layers. In summary, our approach
consists of the following two phases (see Fig. 8.4): a top-down calculation of the
workloads imposed by the users; this provides input for a bottom-up calculation of
performance measures.