Research On Resource Distribution Strategy In Cloud Computing

As a new type of service, cloud computing has received great attention from enterprises and institutions abroad and home. By adopting and deploying cloud computing, enterprises can reduce much cost on buying infrastructures, deploying applications and maintaining application systems. It is all because the potential commercial opportunity from cloud computing that many enterprises from home and abroad are deploying their own cloud computing resource pools. Many problems have appeared after deployment of many resource pools. Especially when a certain cloud provider reaches its peak service time, it may lack resources to fulfill its clients’ requests. However, some other cloud providers may have additional resources at the same time. This situation will lead to a waste of resources and affect customer experience.In order to solve the problems, the paper has done some research on the history of cloud computing, key technologies and deploying mode of cloud computing. And then a horizontal dynamic cloud federation architecture has been presented. Based on the proposed architecture, we propose a game-theory resource distribution strategy in cloud computing. When a certain cloud provider reaches its bottleneck on resources, it will send requests to other cloud providers to form a cloud federation. Resources will be provided by the cooperative cloud providers. As a return, they can gain some profit from the primary cloud provider. The price-based strategy can guarantee the fairness among all cloud providers. We also analyzed a non-cooperative resource distribution strategy to compare the performance of proposed strategy. The results show that the proposed strategy has good performance on stability and dynamic pricing. Moreover, it helps cloud providers to increase individual utility and social welfare.