数字图书馆中基于机群的Web Server数据分布方法

随着数字图书馆中信息量的急剧膨胀,一个WebServer已经很难满足众多用户并发查询的要求。文章用机群作WebServer,它由多个处理节点构成,多个节点协同工作,既增加了系统吞吐量,又减轻了每个节点的负担。因此,如何在多个处理节点上分布数据就成为数字图书馆研究的关键问题。该文提出了一种新的数据分布方法,既考虑了图书馆中数据的相似性,又考虑了多个处理节点的工作负载。该方法使一个查询在尽量少的处理节点上完成,减少了网络传输时间。在多个并发用户查询时,查询的效率远远高于传统的数据分布方法。该方法已经用于笔者自行研制的支持数字图书馆的并行文本数据管理系统PDoc中。

Cluster-based Data Allocation Method of Web Servers in Digital Library

With the exploding of data in Digital library,one Web Server can not satisfy the need of a large amount of concurrent users.This paper uses a cluster as a Web Server composed of a lot of processing nodes.How to allocate data on multiple processing nodes of the Web Server is a key issue in digital library.This paper proposes a new data allo-cation method,which considers both the similarity of the data and the workload of the processing nodes.It tries to use as few nodes as possible to execute one query,which reduces the time needed to gather the query results.The method has been used in the digital library PDoc.Theoretic analysis and practice show that the more the concurrent users are,the high the method's performance is.