自适应负载指标权值的负载均衡算法

摘要

为解决在Web集群负载均衡算法中预先指定权值来评估服务器节点综合负载不能体现各负载指标动态变化情况的问题, 提出一种自适应负载指标权值的负载均衡算法.根据服务器节点各负载指标的实际观测值动态调整各负载指标的权值, 更加准确地评估服务器节点的综合负载情况;采用朴素贝叶斯算法对服务器节点进行分类, 根据分类结果, 综合考虑服务器节点性能以及当前综合负载来指导任务的分配.实验结果表明, 自适应负载指标权值的负载均衡算法能有效减少平均响应时间, 提高系统吞吐量, 有较好的负载均衡效果.%To solve the problem that pre-assigning weight in Web cluster load balancing algorithm can not reflect the dynamic change of each load index when evaluating the comprehensive load of server node, a load balancing algorithm with adaptive load index weight was presented.The weight of each load index was dynamically adjusted according to the actual observed value of each load index to evaluate the true load status of the server nodes more accurately.The server nodes were classified according to the load information using naive Bayesian algorithm, and both the node performance and current load were considered to guide the allocation of assignments.Experimental results show that the proposed load balancing algorithm can effectively reduce average response time and increase throughput, which obtains better load balancing results.