Using quantitative analysis to implement autonomic IT systems

Abstract

The software underpinning today’s IT systems needs to adapt dynamically and predictably to rapid changes in system workload, environment and objectives. We describe a software framework that achieves such adaptiveness for IT systems whose components can be modelled as Markov chains. The framework comprises (i) an autonomic architecture that uses Markov-chain quantitative analysis to dynamically adjust the parameters of an IT system in line with its state, environment and objectives; and (ii) a method for developing instances of this architecture for real-world systems. Two case studies are presented that use the framework successfully for the dynamic power management of disk drives, and for the adaptive management of cluster availability within data centres, respectively.