With the continuing development of the construction of campusinformatization, campus network application has become more and more mature,campus network video applications are closely linked our daily study life,including our scientific research work, and daily learning and life and otheraspects. However, when video live request when increased, the server willproduce load unbalanced problem, the customer’s request delay will occur. Whena request for reducing, idle equipment problem has become more serious, and theserver not for the other systems, caused the server waste of resource. So how touse the equipment to improve the utilization of the resources, and do it inequipment switch between services, where there is no trembling conversion,become video live broadcast system has the problem of challenge.In order to improve the use of the campus network live video systemperformance, and don’t need to add more equipment. On the basis of the campusnetwork live video system.Through the analys is some indicators of data on servertransmission, combine with the cloud computing infrastructure, for existingbroadcast system reform, put forward a prototype system that the live video ofbased on cloud computing. The system provides handling of the single point error.Key improved load unbalanced problem, improve the utilization rate of the server,reflected the system availability and other advantages. Thus provide the user withthe smooth live video environment.To analyze and compare the feasibility and correctness between the modelsystem and the traditional live system in the paper, and then test on the system.Test environment inc luding two parts that physical video live broadcast systemand virtual video live broadcast system. The former is deployed in physicalservers, the latter through Xen virtualization software deployment in the virtual server. The system fully deployed in the virtual machine of virtual resource pool,and through Eucalyptus monitor the resource pool of to load and operationcondition, LVS according to the flow conditions are responsible for the flow ofload balance. |