The Research And Improvement Of Fault Monitoring Based On Cloud Platform

With the popularity of internet technology and the improvement of information technology, the demands of information are higher and higher from all the areas of the society, the data that will be processing is also increasing. Cloud computing has been applicated from the concept to the practical application. It’s development has matured, it has been developed for the customization available cloud, scalable cloud, service-oriented cloud of private or public cloud. The quality of service is very important for cloud platform, monitorning is an important part of the cloud computing platform, it is the premise of network analysis, systems management, job scheduling, load balancing, event prediction, fault detection and recovery operations. It can help dynamic resource use, testing services defects, found the patterns of users, resource scheduling module auxiliary decision of cloud platform. It can improve the quality of service on cloud platform.BC-PDM(Big Cloud of Parallel Data Mining) is a system that based on the world’s largest telecom companies in the business intelligence applications, and it is designed for mass data to provide efficient, accurate and conveninent data analysis services. The system is developed based on Hadoop cluster, this paper describes the research and implementation of fault monitorning based on cloud platform.This paper describes the background and the status of the research firstly, then for the demands of the project itself, give the overall design and the design of each module design. This article uses the open source monitorning tools of Ganglia and Nagios. Through the in-depth research of the monitorning tools, I summarizes its working principle and the advantages and disadvantages and so on. The system will combine the advantages of ganglia and Nagios and optimization the fault tolerance mechanisms of ganglia simultaneously. The system achieve fault monitoring and resource monitoring. Ganglia and Nagios both have some problem on the storage of monitoring data. The system makes the monitoring data to mysql database through the persistent storage tool and make the data management and data analysis, optimization the storage problems of monitoring data.The system uses the open source montoring tools of Ganglia and Nagios, then make the requirements analysis of the system, the key research of the system, finally completed the resource monitoring and fault monitoring. The system realized the physical resources, virtual resources, services and resources for comprehensice monitoring and analysis of resource utilization. According to the analysis,it achieve the fault monitoring through email, text messages and other means to monitor the source and fault. To ensure the normal operation of the cloud platform.Finally applying the research all above, the cloud platform monitoring system is implemented, and its running results show that this strategy is feasible and effective.