Identifying implicitly declared self-tuning behavior through dynamic analysis

Autonomic computing programming models explicitly address self management properties by introducing the notion of ldquoAutonomic Element. However, most of currently developed systems do not employ autonomic self-managing programming paradigms. Thus, a current challenge is to find mechanisms to identify the self-tuning behavior and self-tuning parameters which have implicitly been declared using non-autonomic elements, and to expose them for monitoring or to an analysis framework. Static analysis, although it shows a good potential, it results in many false positives. In this paper, we provide a mechanism to identify the tuning parameters more accurately through dynamic analysis.