A Measurement Methodology to Calibrate Analytical Models including Operating System Overhead

This paper focuses on reducing computing time in large simulation models by replacing some of their elements with analytically determined solutions. In particular, we study realistic analytical server models that include operating system overhead. This overhead will be determined by measuring the behaviour of a representative workload. In this work, we propose a methodology to reduce the number measurements, which is based on information provided by an analytical workload model. We will focus mainly on the first step of the methodology: the development of the analytical workload model. In absence of a real workload a representative benchmark has been used; TPC-C benchmark was chosen due to our interest in transactional systems.