Research For The Data Transmission Model In Cloud Resource Monitoring

Cloud Computing is a new model of business computing, the computing tasks are distributed to the virtualized resource pool which is composed of a large number of computers. Cloud resource monitoring is one of the important points in service availability protections, it is responsible for collecting the resource load information, and it is the premise of job scheduling, load balance, event prediction, fault detection and recovery, etc. The existing data transmission model based on resource monitoring increase the data consistence between monitoring components and monitored resources through frequent exchange of data which will cause dramatic communication cost and network congestion, and lead to a decrease in the quality of cloud service. Therefore, the research on data transmission model is of great importance to the cloud computing system.This paper describes the development of the cloud computing, the research background, the research status and the significance. Then it analyses the basic concept of cloud computing, the system structure and the service mode. Besides, it expounds the classification of cloud computing, the architecture of monitoring technology and the typical application, as well as Libvirt. Focusing on the problem that frequent data transmission in cloud resource monitoring system will cause dramatic communication cost, a Push-Pull Hybrid Model (PPHM) is proposed. The proposed model takes advantages of the push and pull model. In the monitor terminal, the proposed model predicts changes for resource load by time sequence model to adjust data transferring frequency dynamically. At the same time, it classifies the changes of resource loading and then transfers those critical changes under push model to reduce communication cost and insure the higher consistency. In the monitor terminal, if the monitoring components haven’t receive the information of resource loading from the monitored node within a predetermined time, the model will use the pull model to inquire about the load of monitored resource timely to keep uninterrupted communication between monitoring components and monitored resources.The PPHM model is applied in the virtualization teaching management system Vmmaster to implement the data communication of monitoring module. The experimental results show that the model can effectively improve data consistency and reduce communication overhead compared with the push and pull model, it can effectively improve the monitoring performance of the teaching virtualization management system.