An optimization approach for service deployment in service oriented clouds

In service oriented cloud, in order to meet the different users' service requirements, the cloud venders need to manage the service applications' deployment effectively first. During the deployment process, the deployment costs and the influence of service deployment to the cloud are very important issues to be paid attention to, which lacks in current research. In this paper, we design a novel optimization approach for service deployment to address above issues, which works in a service deployment framework. The approach can improve the deployment efficiency and reduce the deployment costs. Atom-services, the basic units of service application, are partitioned to different service families with its compatibility and installation policy. We present three algorithms to automatically standardize the installation expressions, simplify and optimize the installation expression series during the service deployment process respectively. An analysis about the service deployment result is given in the paper and the experiment data demonstrates the approach's efficiency.