一种新的集群Web服务器体系结构及其数据复制算法

A New Architecture of Clustered Web Server and its Data Replication Algorithm

数据副本管理是集群Web服务器(CWS)的一个重要研究内容,它与CWS的体系结构和具体的应用密不可分。文章提出了一种架构于Linux虚拟服务器(LVS)基础之上、应用广泛、可扩展性强的集群Web服务器结构,给出了相应特定的数据复制算法,包括数据定位和数据同步。通过实验对此复制算法进行了测试与分析,并和其它复制方案进行了比较,证明了本算法具有更高的可用性和更好的性能。

Data replication is an important issue in the research of Clustered Web Server (CWS) and it is closely associated with the architecture and application of CWS. This article proposed a new architecture of CWS, which is based on the Virtual Linux Server(LVS) and features wide-range application and strong expansibility. The correspondent specified data replication algorithm was also given, including data location and data synchronization. The algorithm was proved to have better usability and performance than other means through analysis and test.