Research On Virtual Machine Consolidation Method Based On Load Forecasting And Virtual Machine Selection

In recent years,with the popularity of the Internet and the rapid expansion of cloud computing market,a series of cloud computing products emerge as the times require,cloud data center is one of them.With the development of data centers,resource utilization is declining,energy consumption is too large,and service performance is declining.Virtual machine integration technology is an important means to deal with the above problems.The essence of virtual machine migration and consolidation is to use virtual machine migration technology to manage and schedule host resources,and the services provided by virtual machines before and after migration are not affected.At the same time,the number of active hosts in data center has decreased significantly.Compared with dormant hosts,active hosts generate higher energy consumption,so it can be reduced by reducing the number of active hosts.In order to further improve the comprehensive performance of the data center,this paper optimizes the related algorithms.The virtual machine consolidation process of cloud data center is divided into three main parts: the judgment of overload host,the selection of vm to be migrated,and the matching of virtual machine and host.In the overload host judgment part,the dynamic exponential smoothing algorithm is applied to the host load prediction module to adjust the key parameters,and the host load prediction algorithm FDES based on dynamic exponential smoothing is proposed.The algorithm is used to predict the future load of the host and reduce the probability of host overload.In the selection part of the virtual machine to be migrated,the migration time and time of the virtual machine are considered Based on the correlation between virtual machines,a new virtual machine selection algorithm MTC is proposed.The results show that the algorithm can not only improve the service quality of the data center,but also reduce the energy consumption to a certain extent.Finally,in view of the different emphasis of the two algorithms,the two algorithms are combined,and a virtual machine consolidation scheduling strategy FDES?MTC based on load forecasting and virtual machine selection is proposed,the proposed strategy effectively solves the balance between reducing energy consumption and quality of service,and improves the comprehensive performance of the data center.In this paper,the virtual machine consolidation scheduling process is simulated with Cloudsim platform,and the data used in the experiment are real datasets of Planet Lab platform.To prove the reliability of the strategy proposed in this article,this paper evaluates the proposed virtual consolidation strategy.Compared with the other six strategies,The indicators of cloud data center based on FDES?MTC are the lowest.In the 10 day test set of Planet Lab workload,compared with the strategy of getting larger value,the energy consumption index EC is reduced by 23.23%,the service level agreement violation rate index Slav is reduced by 45.71%,and the comprehensive performance index ESV is reduced by 53.31%.These data show that the proposed FDEs?MTC strategy can effectively reduce the energy consumption and service level agreement violation rate of the data center,and has a good balance between reducing energy consumption and ensuring the service quality level,and the comprehensive performance of the data center has been improved.