SCALAR: Scalable Architecture for Replication of Data

提案编号：CCR-9902023 TITLE：Scalar：Scalable Architecture for the WebPI：Peter Scheuermann项目摘要：万维网的广泛使用和访问Internet的简便性正在极大地改变我们对分布式系统的要求。Web的使用规模呈指数级增长，对Internet的容量造成了压力，导致Web服务的性能和可靠性都很低。复制是解决这些问题的主要技术之一。本项目将致力于开发一个原型系统，它将为Web上的复制提供一个可扩展的体系结构。研究了一种新的客户/服务器体系结构，它将允许用户透明的地理复制。在该系统中，客户端作为签名的小应用程序被下载到市场上可获得的浏览器。支持Servlet的Web服务器可以直接使用Servlet进行扩展，否则可以使用支持Servlet的服务器端代理进行扩展。与其他提出的Web体系结构不同，在该体系结构中，客户端和服务器都是主动的。服务器机器确定要复制/迁移的文档以及文档的最佳位置，而客户端机器根据对服务器先前响应时间的估计来选择提交请求的最佳服务器。将开发基于集群技术的启发式算法，以便确定作为一个单位有利地迁移/复制的文件集。

Proposal Number : CCR-9902023TITLE : SCALAR : Scalable Architecture for Replication on the WebPI: Peter ScheuermannProject Summary:The wide use of the World-Wide Web and the ease of access to the Internetare changing drastically our requirements in distributed systems. The exponentiallygrowing scale of use of the Web is stressing the capacity of the Internet and leads to poor performanceand low reliability of Web service. Replication is one of the major techniques that have beenproposed to overcome these problems.This project will focus on the development of a prototype system that willprovide a scalable architecture for replication on the Web. A new client/server architecturethat will allow for user-transparent geographic replication isinvestigated. In this system, the clients are downloaded as signed applets to commercially available browsers. Web servers that supportservlets can bedirectly extended with servlets, otherwise with server-side proxies thatsupport servlets. Unlike other proposed Web architectures, both clients and servers take activeroles in this architecture.Server machines determine the documents to be replicated/migrated and bestlocations for the documents, while client machines choose the best server to submit theirrequests based on estimates of the servers' prior response times. Heuristicalgorithms based on clustering techniqueswill be developed in order to determine sets of documents that bebeneficially migrated/replicated as a unit.