关于Hadoop集群作业调度算法的探讨  

Research on Algorithm of Hadoop Cluster Job-scheduling

作　　者：林晓婧[1] 吴俊平[1] 

机构地区：[1]华南理工大学自动化科学与工程学院,广州510640

出　　处：《价值工程》2013年第7期187-188,共2页Value Engineering

摘　　要：随着信息和数据爆炸式增长,Hadoop等海量数据处理平台日益重要,作业调度算法是Hadoop集群的核心,本文对Hadoop集群默认的三种作业调度算法以及异构集群环境下的作业调度算法进行了系统的研究和对比,并提出了今后相关科研工作的重点。With the explosive growth of information and data,mass data processing platform such as Hadoop is more and more important. Job-scheduling algorithm is the core of Hadoop cluster.Three acquiescent job-scheduling algorithms of Hadoop cluster and the job-scheduling algorithm of heterogeneous cluster have been researched and compared.The key point of future research is proposed.

关 键 词：HADOOP集群 海量数据 作业调度 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]