改进的分布式系统负载平衡策略  

Improved Strategy of Distributed System's Load Balancing

作　　者：董立岩[1,2] 黄乐[1,2] 王利民[1,2] 毛锐[1,2] 殷涵[1,2] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012 [2]吉林大学符号计算与知识工程教育部重点实验室,长春130012

出　　处：《北京工业大学学报》2011年第12期1903-1906,共4页Journal of Beijing University of Technology

基　　金：'十一五'国家科技支撑计划重大项目(2006BAK01A33)

摘　　要：针对服务器计算能力及带宽等方面的异构性导致的系统负载不均衡问题,提出了一种改进的分布式系统负载平衡策略.将模糊综合评判理论运用到服务器性能评价中,选出性能最佳的服务器,使负载均衡,提高了分布式资源的利用率.实验结果表明,运用此策略可以准确地选出最佳服务器.In a distributed system, its load imbalance is often caused by the heterogeneous of computational capacity and different bandwidths among servers. An improved strategy of a distributed system＇s load balancing is proposed, which uses the fuzzy judgement algorithm, to evaluate the server performance and choose the best server. The load of the system is balanced and the utilization of distributed resources is improved. Experiments show that the best server can be accurately selected through by this strategy.

关 键 词：分布式系统 负载平衡 模糊综合评判 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]