Research And Implementation Of Performance Optimizationon On A Monitoring System TANC Based On Cloud For Large-Scale Measurement

Network applications continue to rapidly enhance all aspects of personal, business and social activities in recent years. Technology is changing, and so are the demands and the expectations of global Internet users. Good Internet performance is the foundation of satisfying Internet service. The measurement of network performance not only can find out potential problems and help maintenance workers take actions, but also provide an opportunity for researchers to learn about the situation of the Internet and allocate resources reasonably in the future. This is what makes the large-scale measurement with cloud monitoring platform meaningful.TANC is a distributed system based on cloud monitoring, which consists of probes and management system. It can measure the performance of links with the programs running on probes. Its performance and the user experience directly influence the efficiency and effectiveness of the measurements. The system has been focusing on measurement between LAN nodes and some WAN nodes. So the topology maps do not have many nodes or levels. When deploying TANC into Internet in the real world, the system faces many problems, such as Internet attack, multi-nodes deployment and user interaction etc. TANC needs functional enhancement and performance optimization to support future study and large-scale measurement.PlanetLab is a shareable, large-scale and open global platform for the development and testing of the next generation Internet and its applications. This paper designs and implements automatically calculation of the topology map of PlanetLab nodes to accord with the demands of the system and the scale of this map. This topology map supports the deployment of several nodes at the same time. Also, users can measure links based on geographical position with this topology map. In the meanwhile, users can relate the nodes to other existing maps to implement large-scale measurement in the real world. With RSA encoding the password and Struts2, the system can become more reliable and safer. By using HTML5data cache function and uniting pages and interfaces, users will find the system more flexible and friendly. With the help of improving the system’s functions and performance, the system will help a lot for large-scale measurement and further analysis.This paper firstly introduces related technical background about large-scale measurement with cloud monitoring platform, including PlanetLab nodes’topology map, multi-nodes deployment and user interaction for measurement. Subsequently this paper improves the security of the system based on the common problems of web systems and the demands of TANC. Finally, based on the research before, we design and implement the security enhancement and user-experience improvement. According to the test results, we draw the conclusion that the new functions and relative improvements make the system more stable and be ready for large-scale measurement.