CLAM: Managing cross-layer adaptation in service-based systems

Abstract

Service-based systems (SBS) have a complex layered structure where the service-based application (SBA) is implemented through a composition of services, which run on top of service oriented infrastructures. Taking into account the heterogeneous and dynamic execution context of such complex systems, adaptation is not straightforward. While several state-of-the-art adaptation approaches, unaware of each other, target different problems at specific parts of the system, the isolated enactment of those adaptations results in ignoring the overall impact of the adaptation on the whole SBS. In this work, we propose an approach that introduces a cross-layer adaptation manager (CLAM) to tackle this issue. The approach relies on a cross-layer meta model of the SBS and a set of predefined domain specific rules to integrate and coordinate existing analysis and adaptation tools. It assesses the impact of an initial adaptation trigger at different system levels, and if needed, proposes additional adaptations, consistent with the overall system. The paper introduces the proposed approach and presents preliminary results on its first implementation with concrete analysis and adaptation tools.