检索
E-mail
RSS
作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2012, Vol. 38 ›› Issue (11): 186-188,195.doi: 10.3969/j.issn.1000-3428.2012.11.057

• 人工智能及识别技术 • 上一篇    下一篇

基于混沌PSO算法的Hadoop配置优化

柳 香，李俊红，段胜业   

  1. (河北师范大学数学与信息科学学院，石家庄 050024)
  • 收稿日期:2011-07-21 出版日期:2012-06-05 发布日期:2012-06-05
  • 作者简介:柳 香(1984－)，女，硕士研究生，主研方向：智能算法，分布式计算，云计算；李俊红，副教授、CCF会员；段胜业，硕士研究生
  • 基金项目:
    国家自然科学基金资助项目(10926719)；河北师范大学基金资助项目(L2009Y01)

Configuration Optimization of Hadoop Based on Chaos PSO Algorithm

LIU Xiang, LI Jun-hong, DUAN Sheng-ye   

  1. (College of Mathematics and Information Sciences, Hebei Normal University, Shijiazhuang 050024, China)
  • Received:2011-07-21 Online:2012-06-05 Published:2012-06-05

PDF

963

被引次数

8

摘要： 针对Hadoop的组合优化问题，提出一种基于混沌粒子群优化算法的配置优化策略。为保证算法在进行局部搜索的同时避免陷入局部最优，对陷入局部最优的标准值和混沌处理的迭代次数的不同组合进行实验，获得使Hadoop性能达到最优的配置。实验结果表明，算法得出的最优配置方案能够有效提高Hadoop的性能，具有较好的稳定性。

关键词: Hadoop平台, 分布式系统, 置方案, 能优化, 子群优化算法, 沌算法

Abstract: Hadoop configuration is a NP-hard combinatorial optimization problem. Aiming at the configuration problem, a configuration optimization strategy is proposed. To ensure that the algorithm not only can search in the local, but also can avoid failing into local optimum, it selects different criteria values and different iteration times to do experiment, so as to search out the values making Hadoop performance optimal. Experimental results show that optimal configuration derived from the algorithm can effectively improve the performance of Hadoop, and has good stability.

Key words: Hadoop platform, distributed system, configuration scheme, performance optimization, Particle Swarm Optimization(PSO) algorithm, chaos algorithm

中图分类号: