Dynamic Cloud Environment Virtual Machine Optimization

Cloud computing is a new business and service model, allowing the user toaccess to the application and related data at any place. Its goal is to simplify thecomputing and storage resources like as the public water and electricity easily touse. Network can be used conveniently, as long as users paid by the amount.From the perspective of supplier and users, economic benefit and service qualityare the most concerned issues. In dynamic cloud computing environment, in theallocation of resources management process to avoid excessive use of resourcesand underutilized. In order to ensure that the service level agreement, andimprove resource utilization, virtual machine configuration for online dynamicoptimization is necessary.We combines the characteristics of cloud systems analysis the service formof cloud computing, and illustrates the importance of virtual resource dynamicoptimization, in view of the cloud applications for problems encountered.Virtualization is the key technologies for IT supporting system evolved intocloud platform. This paper describes three different virtualization technology,focusing on the Xen virtualization technology, and describes the dynamicallocation scheme of Xen virtual machine briefly. Dynamic optimization of thevirtual machine configuration needs to solve the problem of virtual machine livemigration, we propose an A-HRL algorithm to configure virtual machinesreal-timely and automatically. Ants leave pheromone during the traversal, andpheromones change rate is introduced into the concept of roughness. And createsthe hierarchical sub-goals according to the threshold of roughness, thus able toexplore more effective. A-HRL algorithm is applied to adaptive applications inthe virtual machine configuration, this method creates a task group and virtualmachine availability assessment, supervising each task progress and quality.Living migration of workloads to optimize the dynamic configuration of virtualmachines. The algorithm is used to analyse the virtual machine configuration inCloudSim simulation, and the results show that this algorithm has betterperformance than the conventional algorithm, and improve resource imbalance problem and utilization rate of resources effectively.