Research And Implementation Of Digital Experimental Platform Based On Cloud Computing

Under the development of information technology and the upgrade of traditional experiment system, the improvement of resource utilization and service capability for experiment platform has become the research hot spot. Cloud computing provides a strong support for open digital experimental platform, but how to use cloud computing technology effectively, building a standardized open digital experiment platform which assures the interactive ability and high resource utilization rate, still faces technical challenges.Focusing on the tasks of experimental environments which the needs of users and the resources of open digital experimental platform, this work carries out researches as follows:Firstly, this investigates the current situation and open requirements of the digital experimental platform, analysises the necessity of the combination of digital experimental platform and cloud computing, investigates the principle, common architecture, implementation mechanism, resource scheduling of cloud computing, and VMware cloud computing, and analysises the feasibility of the combination of digital experimental platform and cloud computing.Secondly, according to the specific requirements of digital experimental platform, the three-layer architecture of digital experimental platform based on cloud computing (DEP2C in short) is given, including resources layer, platform layer and application service layer. In this architecture, platform layer is the core to realize the cohesion of user and resources, providing services which the users need. Through investigating task division, resource division and resource scheduling strategy deeply, the task division model and resource division model are proposed respectively, then priority coding after task division, function level coding after resource division, and the resource scheduling strategy is proposed to achieve suitable match of tasks and resources.Thirdly, this paper designs DEP2C in detail, bring in task division, resource division and resource scheduling strategy to achieve standardized management, give a three-layer collaboration resource scheduling model, through resources scheduler, load balancer and comprehensive scheduler to achieve the scheduling of resources. Using the cloud computing technology of VMware, the prototype system of DEP2C is implemented. In the prototype system, VMware ESX4.1is used to implement the virtualization of servers, OSN is used to realize the virtualization of FC storage and IP storage. Finally, resource pool is formed, via VMware vCenter to realize unified management. The prototype system can divide the virtualization resource into several virtual machines which can carry out different experiment tasks.Finally, in this work, DEP2C was implemented. In the prototype system of DEP2C, two experimental environments which the users need are implemented; for example of the render application, render task is divided and resources is allocated, then through the performance test of resource scheduling, the validity of resource scheduling strategy is verified. This work has conducted the initial attempt of cloud computing applications, and established the foundation for DEP2C application to promote.