Research On Key Technologies Of Resource Scheduling For Cloud Network Experimental Platform

Demand for cloud computing environment of the experimental platform to face,requested by the user a variety of tasks,types of experimental tasks are not the same,the equipment resources exist larger differences,for the resource management needs to realize standardization management and the sharing of resources through virtualization technology.Therefore,it is necessary to the resources of the virtual experiment scheduling to meet the needs of users.Therefore,this paper presents the cloud network experimental platform of cloud computing.This thesis analyzes the cloud computing technology,and analyzes the combination of cloud computing and experiments,and the specific research contents are as follows:(1)The three layer architecture model of the cloud network experimental platform is studied,and the model structure of the virtual experiment system is proposed.The model consists of three layers,which are resource layer,platform layer,application layer.The cloud computing virtualization technology and experiment based on resource scheduling model is designed,especially the key technologies such as task partitioning,resource partitioning,resource scheduling,etc..(2)The traditional resource scheduling is improved,the task partition and the encoding model and resource division and encoding model are proposed,which can make the experimental task and the experimental resources be efficiently matched and mapped.Then the research on the three layer cooperative resource scheduling mechanism of the cloud network experiment platform is carried out According to the previous task segmentation and coding model and resource classification and coding model,on the platform of detailed design,by the resource scheduler,load balancer and comprehensive scheduler coordinated resource scheduling,design and implementation of the prototype system,through different parts of the design,make the experiment tasks can be segmentation,experimental resources to construct the resource pool,provide scheduling services for the users.Finally,the prototype system is tested.The task execution time of the cloud network experimental platform in two modes is compared with the overall CPU utilization,and the feasibility and effectiveness of the proposed resource scheduling strategy is verified.