Energy-aware service composition in multi-Cloud

Abstract

Service composition is widely used in multiple scenarios to meet users’ various demands. In a multi-Cloud environment (MCE), a composite request (service request) needs atomic services (service candidates) located in multiple clouds with various functions. Service composition composes atomic services from multiple clouds together as a new service. Prior work focused on how to compose services and ignored energy consumption caused by the execution of atomic services. In this paper, we examine an energy-aware heuristic for service composition (EASC) under a multi-Cloud environment to reduce energy consumption from executing atomic services. To meet our requirements, we try to compose services in one cloud to reduce energy consumption for transferring files between atomic services. Beyond that, we also consider the influence of the split-point positions to energy consumption and other metrics. Simulation results show that our proposed method has shown good performance in reducing execution time and energy consumption.