Design And Implementation Of Multidimensional Resource Control Subsystem In Cloud Computing Platform

With the rapid development of information technologies and the popularization of Internet applications,more and more companies and developers pay close attention to the cloud computing technologies.As the cloud computing platforms grow fast,especially the rapid innovation in IaaS cloud computing platforms,problems in traditional data centers like low resource utilization and high operating cost could be effectively resolved.Application development based on IaaS cloud computing platform allows developers pay more attention to the core business,and greatly reduce the operational cost and difficulty.How to achieve effective control on multidimensional resources and efficient resource scheduling strategies while providing SLA guarantee and increasing resource utilization are the key issues in cloud computing platform.In this paper,a multidimensional resources control system in IaaS is designed and implemented after a set of investigation and analysis in mainstream cloud computing platforms.This system could support the effective performance isolation of multidimensional resources and make the cloud computing platform provide SLA guarantee.This system also provides many resource scheduling strategies,and further improves the resource utilization through the static and dynamic resource scheduling.This paper begins with the introduction on the background of cloud computing technologies and the investigation on the resource management solutions of mainstream cloud computing platforms.The requirements of resource control subsystem in cloud computing platform is then put forward based on the analysis of the research.Based on these requirements,the design scheme is proposed,which includes the creation,performance isolation,performance evaluation,performance monitoring and resource scheduling of multidimensional resources.Next,the detail introduction about the implementation of each module is put forward.In the end,the function and performance tests are involved to verify the effectiveness of the multidimensional resources control subsystem raised in this paper.