YML-PC: A Reference Architecture Based on Workflow for Building Scientific Private Clouds

Abstract

Cloud computing platforms such as Amazon EC2 provide customers flexible, on-demand resources at low cost. However, while the existing offerings are useful for providing basic computation and storage resources, they fail to consider factors such as security, custom, and policy. So, many enterprises and research institutes would not like to utilize those public Clouds. According to investigations on real requirements from scientific computing users in China, the project YML-PC has been started to build private Clouds and hybrid Clouds for them. In this paper, we will focus on the first step of YML-PC to present a reference architecture based on the workflow framework YML for building scientific private Clouds. Then, some key technologies such as trust model, data persistence, and schedule mechanisms in YML-PC are discussed. Finally, some experiments are carried out to testify that the solution presented in this paper is more efficient.