Research On Cloud Storage Resource Allocation Based On Marginal Analysis And Restricted By Credit

With the rapid development of cloud computing, the cloud storage was been a main subject conveniently. Similar to cloud computing, cloud storage is the use of cluster application features, grid technology and distributed file system, set up network in a large number of various types of storage devices to work together by software, and then provide functions of a complete set of system of data storage and business access. In the face of cloud storage service, effective scheduling of resources was an urgent problem.The rapid growth number of resources is gradually challenge the memory bandwidth, high-speed communication and the power consumption of the equipments in cloud storage space, how to make the storage resources more reasonable and efficient is become the one of the hot issues in the research of cloud storage.Combining with the characteristics of cloud computing, this paper focuses on how to improve the performance and get maximum utility of the cloud storage system. By the analysis of the cloud computing economics model credit as the constraint conditions, using new classical economics principle and the marginal analysis method to resolve the resource allocation problem of the cloud computing environment.Firstly, for the trust between resources and users in the process of resource scheduling, this paper proposed to the reputation model by resources evaluating similarity of user and then rating the reputation classes.Secondly, consideration of resources function, performance, reputation as constraint of the reference objects. In economic models, when the function and performance is consistent, this paper presents a resource scheduling management by reputation which based on super marginal analysis method, it converted the resource allocation into division issues on specialization, adopting the optimal scheme for resource utilization and makes better the performance of the system.Finally, simulation is put forward to the resource scheduling model of reputation which based on super marginal analysis method by Matlab. The experiments show that the scheduling algorithm is efficient on resource utilization ratio and performance of the cloud storage system, comprehensive compare and confirm the super marginal scheme that makes the system total utility is the largest.