Design And Implementation Of Monitoring System For Large Scale Data Center

In recent years, the scientific information and social information are driven by the continuous development of our national economy. The amount and scale of domestic data centers are increasing. Traditional infrastructure management and maintenance level are gradually difficult to deal with the ever-expanding network environment and growing application requirements.In the data center of institute of High Energy Physics, servers and network devices are massive increasing, and the sub-center is under construction planning. The accuracy、scalability and flexibility of the exiting monitoring system can’t meet the current demand. So, the administrators need a centralized monitoring system for the large scale of network devices、servers and services. The system must have high real-time, effectiveness and scalability and has the ability of24-hour surveillance and alarm notification.The main work of this thesis is summarized as follows:1) According to the needs analysis, carried on comparative analysis for Ganglia、Cacti and Nagios, the three existing open source monitoring system. Researched the advantage of Nagios’s frame and it’s deficiencies of applications. To supply a monitoring system for the large scale data center, it determined to use the framework of Nagios and to deeply design due to its shortage.2) Achieved overall system design、completed the definitions of monitoring objects、determined the strategy of scheduling management、achieved the custom development of plugins、completed the data collecting、Converted the structure of the data storage format for graphics display development、implemented the design of web-based configuration management、achieved the permission settings and alarm functions、completed the design and implementation of a distributed architecture.3) Based on the characteristics of monitoring needs, completing the design and implementation of high-availability used of open source software.4) Through the planning and deployment, the system is applied to environment of the data center.