| The project,"High-performance Remote Technology Supporting System", is supported by Electronic Development fund of Information Industry Ministry. With the rapid development of Internet, the number of the Internet users and the hosts are increasing greatly, the Internet behavior is getting more and more complicated. It is urgent for Internet Service Provider (ISP) and network software designers to understand the performance of network in order to make the best of the existing resources. Measurement and analysis of the network behavior is an effective way to get the first-hand parameters of the network.The high-performance Remote Technology Supporting System consists of control servers, relay servers, communication nodes. It transmits its data between clients and servers by relay servers. How to deploy the Relay Server (RS) is an imperative problem while implementing the"High Performance Remote Technology Supporting System". This thesis aims at how to optimize the performance of the system, considering this system has a high demand of real-time ability, it analyses and discusses the network latency between end-hosts and server.We choose a server in our laboratory and deploy 12 clients that will send HTTP requests in different cities throughout the country. Because the communication between clients and server will generate large numbers of data, we collect them for two months or so, then analyze them and transform the data into graphs so that they will express the rule clearly.As can be demonstrated from the graphs, the longer the distance from the clients and the servers is, the larger the network latency is. So, when the high-performance Remote Technology Supporting System deploy its RS throughout the country, we can plot out some RS in several regions firstly, and then choose a RS in the small region based on rule that we found in our experiment. Not only can it deploy the RS rationally but also enhance the performance of the high-performance Remote Technology Supporting System. |
PDF Full Text Request