Web请求的时延刻画和降低时延的协作分配模型

1 引言随着互联网的发展,基于Web的协作系统越来越多,如基于Web的分布式设计系统,基于Web的网格计算,基于Web的电子商务,以及基于Web的工作流办公系统等。这些服务都是以Web为平台,任何客户只要连接上互联网,就可以享受这些系统提供的服务,且客户请求和响应服务都以Web页面存取。互联网的尽力而为的服务模式,可能导致系统不能满足客户的要求。客户的服务质量(QOS)除安全性外最重要的是Web请求的响应时延。另外,客户请求的媒体除文本外,音频和视频也越来越多,这些都会增加时延,仅用足够的网络带宽不能完全解决问题。目前,许多测量表明Web

The Analysis of the Response Delay Times of Web Requests

With evolution of Internet, many cooperative information systems are developed and deployed. These systems may not meet the requirements of clients because of best.efforts in Internet sometimes,so a lot of techniques are provided to satisfy the Web.QOS(Quality Of Services)now. It is important to understand Web delay and to decrease the delay. In this paper,the Web delay is characterized from both network protocols and network architecture,and some basic techniques are discussed to shorten the delay, furthermore assessments of these techniques are given. Finally,a cooperative assigning model is presented to balance workloads and to decrease the delay on the basis of measuring performances of both networks and servers.