A Methodology for Evaluating Complex Interactions between Multiple Autonomic Managers

T. Eze, R. Anthony, C. Walshaw & A. Soper

Abstract:

The very success of autonomic systems has inevitably led to situations where multiple autonomic managers need to coexist and/or interact directly or indirectly within the same system. This is evident, for example, in the increasing availability of large datacentres with multiple [heterogeneous] managers which are independently designed. Potentially, problems can arise as a result of conflict-of-interest when these managers (components) coexist. There is a growing concern that the lack of support for interoperability will become a break issue for future systems. We present an architecture-based solution to interoperability. Our approach is based on a Trustworthy Autonomic Architecture (different from traditional autonomic computing architecture) that includes mechanisms and instrumentation to explicitly support interoperability and trustworthiness. We posit that interoperability support should be designed in and integral at the architectural level, and not treated as add-ons as it cannot be reliably retro-fitted to systems. In this work-in-progress paper, we analyse the issue of interoperability and present our approach using a datacentre multi-manager scenario.

Keywords: autonomic computing; interoperability; datacentre; multi-manager




Fri Mar 14 17:41:02 GMT 2014