Configuration-Space Performance Anomaly Depiction Christopher Stewart Kai Shen Department of Computer Science, University of Rochester Arun Iyengar Jian Yin IBM T. J. Watson Research Center Complex distributed systems (like those based on J2EE platforms) are designed to perform well for a variety of application workloads and configuration settings. In practice, however, the system performance may not meet the expectation at all execution conditions. This short paper describes our research on depicting performance anomaly manifestations over a large execution condition space for distributed systems. Specifically, we provide a case examination of our research target and describe our work in three areas---performance anomaly identification, configuration-space anomaly depiction, and the utilizations of depiction results.