分布式数据流处理系统管理中负载均衡问题建模与求解  

Modeling and Solution for Load Balancing Optimization in Distributed Stream Processing System Management

作　　者：唐颖峰 陈世平[2] TANG Ying-feng;CHEN Shi-ping(Academic Affairs Section, Shanghai University of International Business and Economics, Shanghai 201620, China;Management School, University of Shanghai for Science and Technology, Shanghai 200093, China)

机构地区：[1]上海对外贸经贸大学教务处,上海201620 [2]上海理工大学管理学院,上海200093

出　　处：《运筹与管理》2021年第4期155-162,共8页Operations Research and Management Science

基　　金：国家自然科学基金资助项目(61170277,61472256);上海市教委科研创新重点项目(12zz137);上海市一流学科建设项目(S1201YLXK)。

摘　　要：对分布式数据流处理系统管理中,处理节点负载均衡问题进行了研究。阐述了分布式数据流处理系统的运行机理以及节点负载不均衡的成因,并提出了对系统负载均衡调整的优化方案;对提出的优化方案建立模型,并对模型的适用条件进行理论分析;然后采用蚁群算法对模型进行求解,并针对分布式数据流处理系统实时性的需求对算法进行改进;最后用实验证明本文所建立的模型及其求解方法对于解决分布式数据流处理系统管理中节点负载均衡问题的有效性。The paper concentrates on the load balancing problem in the management of distributed stream processing system. The operation mechanism of distributed stream processing system and the causes of unbalanced load of nodes is expounded, and an optimization scheme for load balancing adjustment is proposed in the paper. The model of the proposed optimization scheme is then established, and the applicable conditions of the model are theoretically analyzed. Then the model is solved with ant colony optimization, and the algorithm is then improved to meet the real-time requirement of distributed stream processing system.Finally, the validity of the model and its solution in solving the node load balancing problem in the management of distributed stream processing system is proved by experiments.

关 键 词：系统管理 分布式数据流处理系统 负载均衡 组合优化问题 蚁群算法 

分 类 号：C931[经济管理—管理学]