Research On Virtual Machine Resource Scheduling Method Based On Openstack Cloud Platform

Cloud computing changes the traditional IT service pattern.Through virtualization technology,cloud computing takes a large amount of physical resources into a pool of virtualized resources.Users can access the computing,storage,network and other resources at any time.Because of its flexibility,ease of expansion,and elastic computing,more and more enterprises are moving their services to cloud computing platforms,which not only improves the utilization of hardware resources,but also reduces the maintenance costs of the enterprise.With the development of cloud computing and the popularization of cloud services,the heterogeneity of cloud data center resources and the diversity of resource requirements of cloud users have brought great challenges to resource management of cloud platform.How to allocate resources reasonably to users in a complex cloud environment is an important topic of cloud computing research.The virtual resources of cloud computing center are organized and distributed in the form of virtual machines,so the scheduling of virtual machines becomes the core of research.Virtual machine scheduling has an impact on resource utilization,system power consumption,load balancing,and user quality of service.Therefore,the virtual machine scheduling strategy needs to consider multiple optimization goals.Virtual machine scheduling can be divided into virtual machine allocation and virtual machine migration.The virtual machine allocation can be described as a virtual machine and physical machine associated mapping.How to find the best mapping between multiple virtual machines and multiple physical machines under the premise of ensuring multi-optimization goals,which is the problem that virtual machine allocation strategy needs to solve.In this paper,a physical resource balance model and a system load balancing model are established for the optimization of physical resource balance and system load balancing.A virtual machine allocation strategy based on particle swarm optimization is designed.The algorithm is implemented by Cloud Sim simulation software,and compared with First-Fit and Best-Fit algorithm,the validity and algorithm performance of the algorithm are verified.Due to the dynamic nature of the cloud platform,the system load will change with the operation of the platform.Therefore,it is necessary to dynamically adjust the cloud platform through the virtual machine migration technology.For the design of virtual machine migration strategy,this paper designs the migration triggering strategy,the virtual machine selection strategy and the target host selection strategy respectively from the three processes of migration,and regards the user service quality and system load balancing as the optimization target.Finally,this paper analyzes the principle of virtual machine scheduling strategy by using OpenStack as the platform environment,and extends the virtual machine migration strategy based on double threshold to OpenStack cloud platform,and implements OpenStack Cloud platform for virtual machine automated migration management,optimization of the OpenStack virtual machine scheduling strategy.