基于CHC遗传算法的Hadoop作业调度研究  

Research on Hadoop Job Scheduling Based on CHC Genetic Algorithm

作　　者：薛涛[1] 燕明磊 

机构地区：[1]西安工程大学计算机科学学院,西安710048

出　　处：《计算机工程》2016年第3期61-68,共8页Computer Engineering

基　　金：国家发改委高科技产业化基金资助项目(陕发改高技[2009]1365号);西安工程大学博士科研启动基金资助项目(BS0725)

摘　　要：作业调度是影响Hadoop平台性能的重要因素。基于基本遗传算法(SGA)的作业调度进化代数大、收敛速度慢,且其编码与解码、遗传操作过程中存在过多冗余计算,增加了作业总完成时间。为此,提出一种基于CHC遗传算法的作业调度算法。采用双目标函数的控制方式和最优解保留策略,优化作业总完成时间与平均完成时间,从而加快最优解的收敛速度。实验结果表明,与SGA算法相比,CHC算法在作业调度效率、资源利用率等方面有较大的性能提升。Job scheduling is an important factor influencing the performance of the Hadoop platform. The job scheduling based on the Simple Genetic Algorithm（ SGA） has disadvantages that the evolution generation is long and the rate of convergence is slow. In addition,there is redundant computation in encoding and decoding as well as the process of genetic operation,which increases the total scheduling time. To solve the problems,this paper proposes a job scheduling algorithm based on CHC genetic algorithm. It uses the double objective function and the method of keeping the optimal solution to decrease the total scheduling time and the average execution time of the job,so it speeds up the covergence rate of the optimal solution. Experimental results shows that CHC has great improvementthat in the scheduling efficiency and the utilization rate compared with SGA.

关 键 词：CHC遗传算法 基本遗传算法 最优解 双目标函数 作业调度 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]