摘要： 提出了集群服务器并行网页预取模型,模型采用了马尔科夫链分析访问路径并在Web集群服务器的各节点上并行预取页面,把集群技术的高性能和高可靠性与预取技术的快速响应能力结合起来。实验表明,将此模型应用于集群服务器的分发器上,服务器系统具有更高的请求命中率和更大的吞吐量。

关键词: 集群服务器, 并行网页预取, 马尔科夫链

Abstract: This paper proposes a parallel Web prefetching model to Web cluster server.The model adopts Markov chain to analyze the access paths and prefetch Web pages in parallel on each node so that the characteristics of quick response ability,good performance and high reliability can be combined in the Web cluster server.Experimental results show that the system is provided with higher cache hit rate and larger throughput after using the proposed model.

中图分类号: 

• TP393.092