On-Demand Optimal Cloud Service Provisioning Composition across Multi-cloud

The current cloud environment, constituted by many different public cloud providers, is highly independent in terms of pricing models, interfaces and virtualization standard while the ideal cloud environment is homogenous and standardization. In this context, a cloud management system which can provide intermediation and aggregation capabilities to enable users to deploy their service across multiple clouds is necessary. However, most current systems always work in certain clouds and cannot provide advanced service management capabilities to make automatic decisions, such as optimal service component distribution among different cloud. In this paper, we present a modularization service management system (SMS) that can optimally deploy services across multi-cloud automatically in flexible and dynamic infrastructure provisioning environment, in order to achieve a better Quality of Service (QoS) according to service provider´s Service Level Agreements (SLAs). To prove the advantages of the system, we analyze the deployment of a web service in multi-cloud environment using our strategy and compare the result with the traditional works.