A real-time adaptive control of autonomic computing environments

Abstract

Autonomic computing has received a great deal of attention from the research community in recent years. Many techniques have been proposed to monitor, analyze, and change the system under observation, but less attention has been paid to adapting the autonomic computing loop itself. Based on previous results on the design and implementation of a reference real-time architecture for autonomic computing, a self-adapting loop based on system-specific adaptation knowledge, which includes the types and properties of autonomic computing components, behavioural constraints, and strategies for adaptation is proposed in this paper. The proposed system is an integral part of the real-time system that controls the behaviour of the computing environment, evaluating its global behaviour using criteria that take into account the mathematical description of the time variation of the number of users in the system. Based on this evaluation, the adaptive system changes the control structure of the autonomic computing environment by replacing its controller with one matching the user time variation law. Elements of the self-adapting loop and the tradeoff of introducing additional overhead to the autonomic computing processes are discussed. A case study illustrates the self-adapting technique proposed.