Research On Virtual Machine Resource Optimization Allocation Model And Algorithm In Enterprise Private Cloud

With the rapid development of the Internet,enterprise data grows explosively.Large-scale and huge amounts of data needs to be processed in a variety of computing scenarios,which results in the demand for enterprise computing power is far beyond the capacity of its own IT architecture.Thus,it is very necessary to continuously increase system hardware investment so as to achieve the scalability of the enterprise information system.In this context cloud computing emerges.Cloud computing has significant advantages such as pay-on-demand,pay-as-you-buy,flexible scalability,scale economy,etc.Using cloud computing can help enterprises reduce a large amount of hardware and software costs,save investment in the early stage and reduce the cost of operation and maintenance in the later stage.Thus cloud computing is gradually applied into kinds of enterprises.While the advantages of cloud computing are recognized by more and more enterprises,its security is also highly concerned by enterprises.Using private clouds can greatly reduce the security risks of migration,so more and more enterprises construct their own cloud computing platforms and provide cloud platform services.At this stage,the key issue in cloud computing system is how to achieve efficient resource allocation,reduce the operation costs of data centers and reduce the default ratio of service level agreements with the guarantee of user satisfaction.Therefore,it is very important to study the optimal resource allocation in the enterprise private cloud to effectively reduce the operation costs of the data center while satisfying user satisfaction.For this purpose,this thesis applied the queuing theory and stochastic model in operations research to realize the goal of reducing the number of active physical machines while meeting user satisfaction.As for the resource allocation problem for virtual machine in the enterprise private cloud environment,this thesis formulated a vacation queuing model for resource scheduling to find the optimal number of active physical machines.Then this thesis also established an optimization model for resource allocation of virtual machine in the enterprise data center,proposed a resource allocation algorithm by applying the convex optimization theory,and gave the implementation steps of the proposed resource allocation algorithm.Finally,the performance of the resource allocation algorithm was discussed and illustrated by some simulation results in several different scenarios.