Research And Implementation Of Cloud Resource Allocation Method On Isomerism Autonomous Platform

In recent years,cloud computing technology had developed rapidly,and resource allocation is an important research content in the field of cloud computing.Under the heterogeneous cloud computing architecture,due to the different CPU,memory,hard disk and other resources among the autonomous servers of different architectures,there are inevitable performance differences in the processing of different types of services.The traditional static resource allocation method can not keep the efficient utilization of computing resources.In order to meet the requirements of heterogeneous resource allocation and improve resource utilization,this paper studies and implements the resource allocation system under the heterogeneous autonomous cloud platform.According to the process of resource allocation,this paper improves from three aspects:resource pool division,computing resource allocation and service organization.Firstly,a scheme of business cloud partition based on security domain is designed,which is composed of application classification,multi-domain tenant resource isolation and cross domain agent access.The resource pool is divided while the security of cloud platform is guaranteed.On this basis,a resource allocation algorithm based on deep learning is designed,which quantifies the data characteristics and adds work load forecasting model to increase the consideration of business flow and make the given resource allocation scheme and demand more match;based on this algorithm,the resource allocation module of service group is designed,and the application system service is organized.Resource planning and scheduling have been realized by group,and the solution of service organization for traditional application have been provided by setting up department plan to meet different user needs and semantic service organization specification.Through the experimental analysis on Taiji cloud platform,it is proved that the resource allocation system implemented in this paper can adapt to the heterogeneous autonomous cloud platform,and make up the difference between heterogeneous processors on the software level.The research results of this paper lay a part of the foundation for the efficient and stable operation of the independent cloud platform and make a certain contribution to the development of cloud services in China.