Web服务器集群的负载均衡中遗传算子的设计  

DESIGNING GENETIC OPERATORS IN LOAD-BALANCING OF WEB SERVER CLUSTERS SYSTEM

作　　者：张维勇[1] 

机构地区：[1]西南大学计算机与信息科学学院,重庆400715

出　　处：《计算机应用与软件》2010年第4期209-211,共3页Computer Applications and Software

摘　　要：应用遗传算法进行作业调度已被越来越多的学者关注。在Web服务器集群环境中,对于客户端的Web请求分配问题,采用常规的遗传算法进行负载均衡并不总是有效的,好的遗传算子对算法收敛性及收敛到"最优解"非常重要。基于集群环境中Web请求分配的特点,设计了有针对性的遗传算子,即改进的内外结合交叉算子和主动变异算子。模拟实验结果与分析表明这些算子对Web集群的请求分配是有效的。Job scheduling carried out by genetic algorithm-based approaches have been attracting growing attentions from academia.Normal genetic operators are not always effective for load balancing when responding to the Web assignment request from the client in Web server clusters system.Good genetic operators are extremely important for genetic algorithms in its convergence and converging to＂optimal solution＂.Three genetic operators are designed according to the characteristics of Web assignment request in clusters condition：the improved external crossover,the internal crossover and the active mutation.Simulation experiment results and their analysis show that these genetic operators are effective for assigning requests of Web clusters.

关 键 词：负载均衡 遗传算法 交叉 变异 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]