Performance analysis of GRID middleware using process mining

Abstract

Performance analysis of the GRID middleware used in a production setting can give valuable information to both GRID users and developers. A new approach to this issue is to use the process mining techniques. Analyzing logs of the middleware activities, performed on the SEE-GRID pilot production Grid infrastructure, objective qualitative and quantitative information on what actually happens can be obtained. Using the appropriate tools like ProM to apply the process mining algorithms, many interesting findings and conclusions can be drawn. In this paper we describe our approach and show some of our conclusions.