Research And Implementation Of Cloud Management Platform Resource Scheduling Strategy Based On Xen

Cloud computing organizes the geographically dispersed computing resources intoa virtual supercomputer by using the Internet. It manages and schedules the resourcesunifiedly, and constructs them into a computing resources pool to provide on-demandservice. The right network that provides the resources is called "cloud". At present, thetheory of cloud computing is not mature enough. Due to the characteristics of networkresources include distribution, heterogeneity, dynamic, autonomous and so on, the wayto improve the flexibility of resource scheduling and to improve resource utilizationbecome an important and difficult issue of the cloud computing technology.This paper researches and analyzes the resource scheduling strategy of cloudmanagement platform based on Xen in an enterprise level. It takes Platform Company’scloud management product named ISF (Infrastructure Sharing Facility) as the projectbasis, and the cloud computing resource scheduling technology as background whilerelying on virtualization technology Xen. It deeply investigates resource schedulingmodule EGO (Enterprise Gride Orchestrator) of ISF, schedules two kinds of networkresource such as CPU and Memory by using the principle of resource mapping. Andthen it studies and implements resource scheduling strategy: Packing, Striping, CPULoad Balance and Memory Balancing, as well as load balancing in cloud cluster andhigh availability of virtual machine. Thus, network resources can be saved, conflictscaused by cluster effect can be resolved, and stability of virtual computing systems canbe guaranteed. By the help of balance computing of resources and high availability ofthe virtual machine, the goal to load balancing, equitable and reasonable utilization ofresources, reducing IT costs, increasing flexibility; reducing downtime through disasterrecovery, and maintaining business continuity and stability is achieved.