Towards Certifiable Autonomic Computing Systems Part I: A Consistent and Scalable System Design
Haffiz Shuaib, Richard John Anthony · 2012
Abstract—Relative to currently deployed Information Technology (IT) systems, autonomic computing systems are expected to exhibit superior control/management behaviour and high adaptability, regardless of op-erational context. However, a means for measuring and certifying the self-management capabilities of these systems is lacking and as result, there is no way of assessing the trustworthiness of these systems. Two things are needed to begin to address the above. The first is a consistent structure for the autonomic computing system (ACS) and a consistent architecture for the autonomic computing manager (AM). The second is a set of metrics by which the operational characteristics of these systems are to be measured within the context of the targeted application domain. In this first part of a two-part paper, a biologically inspired architecture is proposed for the autonomic computing manager. The interfaces and messages by which this architecture communicates with objects within and those without are technically defined. Also discussed in this paper