摘要
可扩展Web服务器集群是目前高性能网络服务器的主要架构方法,负载均衡技术是集群系统中任务分配的核心环节。提出了一种基于内容识别的负载均衡算法,引入了访问量阈值的概念,并通过动态的修正访问量阈值以适应网络负载的变化;利用动态反馈机制来获取服务器的负载状态,同时通过保证负载的局部性,减少相同内容在多个服务器中的重复缓存,提高服务器Cache的命中率。
At present, the extendible Web server clusters are the important framers for the high performance network servers. The load distribution algorithm is the main role in system. A content-based load balancing algorithm is proposed, and a new concept of requesting threshold is introduced to adapt the network change by modifying it＇ s value. Using a dynamic feedback mechanism to acquire the load conditions of servers, through load locality guarantee reducing the same content to repetitively replicate and improving the cache hit rate.
出处
《计算机工程与设计》
CSCD
北大核心
2006年第18期3410-3412,共3页
Computer Engineering and Design
关键词
服务器集群
负载均衡
内容识别
动态反馈
访问量阈值
server cluster
load balancing
content-based
dynamic feedback
requesting threshold
