基于URL索引划分的Web内容自适应算法

Web集群服务的请求分配算法大多使用Hash方法对请求URL进行散列,并按一定规则对请求内容进行负载均衡调度.提出了一种基于URL词典排序及全部URL按其词典序列分为k*n个集合的URLALLOC算法.该算法通过对URL进行词典序排序并将全部URL按其词典序列分为k*n个集合,依访问流量排序及分段互补等一系列方法将Web负载尽可能均匀地分布到多个后端服务器中.仿真实验结果表明:URLALLOC算法比现有的URL散列方法具有更好的负载均衡能力.

A WEB CONTENT ADAPTIVE ALGORITHM BASED ON URL INDEX DISPATCH

The scheduling algorithms of Web cluster service mostly use Hash method to hash URL's request,and the load balancing can benefit from content-based web application which schedule the applications according to their contents after firstly hashing the URLs to different servers. In the paper it proposes a URLALLOC algorithm based on URL Index Dispatch and kn sets. The proposed algorithm firstly sorts URLs according to their lexicographic order and divides them into kn sets in sorted order,then accomplishes a m...