Research On Virtual Machine Migration Algorithm In Cloud Computing And Implementation Of Cloud Simulation System

Cloud computing is a new type of computing model.Many servers and network devices are connected together based on the Internet and virtualization technology.All resources are configured and managed uniformly by cloud data.Moreover,it can provide computing,storage,bandwidth and other resources for users and enterprises.However,due to the continuous expansion of cloud computing data centers,how to schedule virtual machines to reduce energy consumption reasonably is a issue worthy of attention.For this,this paper proposes a feasible virtual machine migration strategy for the rational allocation and utilization of resources to reducing the overall energy consumption of the data center.Fitstly,propose a virtual machine migration algorithm KP-VMC based on load prediction.The algorithm predicts the CPU and memory usage in the next period based on the KNN model,and comprehensively considers the current load and the load usage in the next period during the migration process.The host is divided by three states:high load,normal load and low load.For high-load hosts,the minimum migration time algorithm is adopted to reduce the influence of the service.Aiming at the low-load host set,an improved ant colony algorithm is adopted to comprehensively consider the three factors of migration times,energy consumption and SLA violation rate to find the optimal migration plan.By comparing our algorithm with existing algorithms,experiments show that compared with other algorithms,this algorithm has a certain reduction in energy consumption,migration times and SLA violation rate.Secondly,starting from the demand analysis,the existing simulation engine has fewer simulation configuration functions,complicated operations,poor availability,and inability to schedule dynamic task flows.We design and implement a cloud simulation system based on the CloudSim sumulation engine.The system is extended based on the CloudSim engine and adopts a layered architecture model to divide the system into a presentation layer,a simulation layer and a data layer.