Research On Resource Scheduling Strategy Based On Load Balancing In Cloud Environment

The traditional resource allocation model based on physical machine is coarse-grained,which results in low resource utilization.The virtualization technology can divide physical resources into isolated virtual machines to improve resource utilization with resource sharing.Cloud computing can consolidate a variety of physical resources to build a virtualized resources pool which is used to achieve the allocation of resources on demand and the dynamic extension of resources.The cloud data center contains a large number of physical machines and virtual machines with dynamic workload.The overloaded physical machine will affect the performance of the upper application,and the underloaded physical machine will not make full use of the limited resources.Therefore,it is of great significance to study the resource scheduling strategy in cloud environment for efficient and balanced use of resources.In this paper,the virtual machine placement algorithm and the virtual machine migration strategy are studied to deal with the load balancing of physical machine.For the virtual machine placement problem,the influence of two scenarios about resource utilization on load balancing is analyzed,and the targeted objectives are proposed to measure the load balancing comprehensively as possible.Based on the modeling of the problem,a multi-objective virtual machine placement algorithm based on load balancing is proposed to solve the problem with multi-objective ant colony optimization algorithm.For the problem of dynamic migration of virtual machine,the cause of dynamic change of physical machine load is analyzed,and the framework of virtual machine migration is built.Several key issues for the virtual machine migration problem are analyzed to propose corresponding strategies.There is when to migrate,which physical machines,which virtual machines and where to move.Finally,a virtual machine migration strategy is proposed to balance load and reduce the migration times.The placement algorithm and migration strategy are implemented in open source simulation platform CloudSim.For the multi-objective virtual machine placement algorithm,the parallel implementation of the algorithm is discussed.The experimental results show that the algorithm can effectively optimize multiple objectives to achieve the goal of load balancing in different scales.For the virtual machine migration strategy,the experimental results show that under the same condition,the algorithm can effectively reduce the migration times and achieve better load balancing.