Research On Virtual Machine Resource Dynamic Integration Algorithm In Cloud Computing Environment

Cloud computing is a model of centralized management of hardware computing resources based on virtualization technology.It provides computing services for users by constructing a virtualized resource pool with mass ive computing resources.Virtualization technology can dynamically manage computing resources,customize resource supply for cloud computing tasks,improve resource utilization and servi ce capacity,enhance cloud server reliability and security,and signif icantly reduce the cost of server construction for users.The static resource allocation strategy is difficult to deal with the ever-changing resource demand and energy consumption.Impr oving resource utilization,reducing energy consumption,and ensuring service of quality(Qo S)are hot research directions in recent years.Based on virtualization technology,this paper has carried out extensive theoretical analysis and technical research around cloud computing resource scheduling.This paper analyzes the application of virtualization technology in data center resource scheduling,and conducts research and demonstration on load identification migration and receiving host search.According to the load change,the load prediction model and the selection strategy of migrated VM for load prediction are proposed by analyzing the characteristics of load data change.According to the characteristics of data center construction structure,this paper studies the physical host search strategy under multi-objective with multi-condition and puts forward a target host selection strategy with multi-objective optimization.The main research contents are as follows.Firstly,to improve resource utilization,a predictive load classification model for virtual machines is proposed according to the characteristics of data center load data changes.This model is based on the classification of load coupling complementary resources between cloud computing tasks,th is paper proposes a step weight load prediction algorithm(SWLPA),realize the cloud computing data center more efficient energy efficiency.Experiments show that the SWLPA can effectively reduce energy consumption and frequency of migration.This strategy can reduce energy consumption by 50%,reduce SLA default rate by 60%,and effectively reduce unnecessary resource overhead and Qo S degradation caused by virtual machine migration.Secondly,for the physical host equipment structure of the data center,this paper proposes a multi-objective host search algorithm(MOHSA).The policy is based on network topology and resource usage data centers to en sure Qo S,reduce the execution time of migration,and improve resource utilization without re ducing the Qo S.Experiments show that using MOHSA can effectively improve migration execution efficiency and resource utilization,reduce energy consumption by 55%,SLA default rate by 80%,and downtime by 60%,achieving goals such as reducing operating expenses.Based on the research method,this paper desi gns and implements the dynamic integration and allocation strategy of computing resources in cloud computing environment.This strategy has the advantages of computing resource load perception and prediction,load type recognition and division,accurate an d efficient scheduling ability.This strategy can achieve the goals of reducing energy consum ption,improving performance,and reducing the operating costs of cloud computing providers.The validity and stability of the proposed strategy were verified by simulation experiments.