分布式网络信息跟踪  

Distributed Web Information Tracking

作　　者：陈康[1] 沈美明[1] 郑纬民[1] 

机构地区：[1]清华大学计算机系高性能计算研究所,北京100084

出　　处：《小型微型计算机系统》2005年第9期1441-1444,共4页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(60131160743)资助

摘　　要：网络是一个非常动态的世界.每时每刻都有大量的网页加入,删除,更新.对于网络信息跟踪系统来说,需要着眼于跟踪一些固定的网页集合,同时及时地把更新的信息反应给用户.为了跟踪大量在地理上分布的网页集合,采用分布式的技术进行网络信息跟踪是非常有效率的办法.本文研究了采用分布式网络信息跟踪的各项参数特征,包括运行时的数据传输特征,分布式站点的数据负担特征,以及网络占用时间情况特征等.Obviously, the web is an extremely dynamic world i.e. information updates even every second. A lot of pages are added, deleted and modified continuously on the Internet. The web information monitoring system fetches the information from the web continuously, finds some changes by comparing two versions of the same page and then notifies the users about what happened to their concerned pages. The distributed computing technologies are the appropriate way to track those geographically distributed pages. This paper studies the characteristics of such system by considering the parameters of data transferring characteristics during runtime, the load of nodes types in the system and the network time occupied.

关 键 词：网络信息跟踪 分布式系统 带宽占用 泊松过程 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]