Autonomic Adaptation solution based on Service-Context Adequacy Determination

Abstract

Autonomic adaptation is a very ambitious emerging domain aiming to build self-adaptable systems. The most important advantages of these systems are: easier complexity management, autonomous service evolution and proactive behaviour. In the ‘classical’ service adaptive systems, the developer must solve ‘manually’ two problems. The first one is to determine a priori the service adequacy to each possible context. The second one depends on the first one and it is to specify a strategy, a reconfiguration suite, which will transform an inadequate service into an adequate one. In order to replace the developer-based reasoning with a machine-based one, we propose a meta-model describing the service and its context into a common graph representation. A set of general rules and operators applied on this meta-model enables the machine to check the service adequacy to its context. The same meta-model is used for searching the adaptation strategy.