Rule engine based lightweight framework for adaptive and autonomic computing

Abstract

The paper describes a framework architecture called the Autonomic Management Toolkit (AMT). This toolkit was implemented to support dynamic deployment and management of adaptation loops. This requires automatic resource discovery, instrumentation and attachment to Autonomic Manager (AM), and furthermore a scalable and easily changed decisionmaking module, which is a major part of the AM. The architecture of a system satisfying these requirements is proposed and described. This system is compared to PMAC (Policy Management Autonomic Computing) - a highly advanced software tool offered by IBM. The central element of AMT is a lightweight AM with Rule Engine as a decisionmaking module. This makes the proposed solution lightweight and flexible. The AM activity is very briefly specified and the process of constructing an execution loop is described. The proposed interfaces are specified. These interfaces are generally sufficient to support a wide range of policies, including standard regulators, well know from control theory. Subsequently, AMT usage is illustrated by a simple example. The paper ends with an overview of related work and conclusions. © 2008 Springer-Verlag Berlin Heidelberg.