On state-dependent workload characterization by software resources
Günter Haring · 1982
A method for the characterization of computer workload at the task level is presented. After having divided the workload into different classes using a cluster technique, each cluster is further analysed by state dependent transition matrices. Thus it is possible to derive the most probable task sequences in each cluster. This information can be used to construct synthetic scripts at the task level rather than the usual description at the hardware resource level.