Challenges and requirements for an effective trace exploration tool

Abstract

Building efficient tools for the analysis and exploration of large execution traces can be a very challenging task. Our experience with building a tool called SEAT (software exploration and analysis tool) shows that there is a need to address several key research questions in order to overcome these challenges. SEAT is intended to integrate several filtering techniques to tackle the size explosion problem that make traces hard to understand. However, the incorporation of these techniques into one efficient tool raises many issues. This paper focuses on these issues and underlies future research directions to advance the area of dynamic analysis of large software systems.