An Implementation Approach for Inter-Cloud Service Combination
Jie Tao, Daniel Franz, Holger Marten, Achim Streit · 2012
Abstract—Increasing cloud platforms have been developed in the recent time and are provisioning different services to customers. The existence of different cloud provides offers the users an opportunity of combining different cloud services to run their scientific workflows with lower cost and higher performance. In this work we developed a framework to allow combining individual services of different clouds to solve complex problems with efficiency. The framework contains a workflow management system that processes users workflow descriptions and starts the related services automatically on the target clouds. A data management component takes care of the data exchange between the underlying clouds. Moreover, a prediction model computes the potential cost and performance of running a workflow on existing clouds, giving users help in selecting the execution target for better performance/cost effect.