Cloud Resource Orchestration Optimization Bsased On ARIMA

With the increasing scale of the current network commercial operation environment,the requirement of computer is also rising.Cloud computing uses virtualization and other technologies to abstract hardware into virtual services,reducing operating costs and increasing organizational flexibility.Because cloud computing stores the data in a large number of distributed computers,corporate data center has removed before the operation mode of the complex,companies can more easily use the resources,achieve on-demand access.Due to above,cloud computing model is popular among enterprises.But cloud environment resources become more and more complex,it is a new problem to arrange these resources rationally.Many resources in the cloud environment,how to reasonably for resource scheduling,reasonably for the system running in the instance of the allocation of resources,realize the global optimization allocation,integrate resources effectively,reduce the operation cost,to raise the utilization rate of resources under the premise of guarantee the quality of service are the current hot spots.However,most of the current resource management mechanisms are focused on IaaS,there are few considerations for the application.On the other hand,most of the simple "threshold" resource management in the form of alarm,when the system load reach the "threshold",the whole system has risk of overload.In order to solve the problem of resource management in the cloud environment,especially on the PaaS layer,this paper studies the application automation deployment and resource dynamic management.In order to meet user requirements dynamically,the load of high concurrency server is alleviated.In this paper,we have studied the cloud computing management platform,and through the analysis and comparison,we have adopted the widely used OpenStack to provide IaaS service.On the PaaS layer,this paper focuses on the representative resource arrangement of Cloudify,and analyzes the basic components of Cloudify,application deployment process and resource scheduling.Then,the current resource prediction algorithm based on time sequence is studied,and ARIMA prediction model is emphatically studied.The application,evolution process,mathematical expression and modeling process of ARIMA model are explained.This paper proposes a dynamic layout optimization architecture based on ARIMA model for the current PaaS layer.The architecture of cloud resource orchestration based on ARIMA was designed and implemented,which was based on the OpenStack infrastructure as a service,and combined with the Cloudify resource scheduling platform on the PaaS layer.The ARIMA algorithm is used to predict the resource consumption in the next time period,and the resource scheduling algorithm is designed.Finally this paper designed a series of experiments,using the real cloud data sets,to validate the prediction accuracy of ARIMA,the test results show that ARIMA algorithm can obtain better prediction effect.The platform and the validity of resource scheduling,can improve the utilization rate of resources and so on has carried on the experiment,pass through the experimental results show that this architecture can effectively shorten the concurrent response time and improve the utilization efficiency of memory.