| 不同MapReduce运行系统的性能测试与分析 |
| |
| 引用本文： | 易修文,李天瑞,张钧波,滕 飞.不同MapReduce运行系统的性能测试与分析[J].计算机科学,2015,42(5):24-27, 50. |
| |
| 作者姓名： | 易修文 李天瑞 张钧波 滕 飞 |
| |
| 作者单位： | 西南交通大学信息科学与技术学院 成都610031 |
| |
| 基金项目： | 本文受国家自然科学基金(61175047,3),国家自然科学基金联合基金(U1230117),四川省科技支撑计划项目(2012RZ0009),中央高校基本科研业务费专项资金(SWJTU11ZT08)资助 |
| |
| 摘 要： | 随着云计算技术的发展,许多MapReduce运行系统被开发出来,如Hadoop、Phoenix和Twister等.直观上,Hadoop具有很强的可扩展性、稳定性,适合处理大规模离线应用;Phoenix具有运行速度快等优点,适合处理数据密集型任务;Twister是轻量级的迭代系统,非常适合迭代式的应用.不同的应用在不同的MapReduce运行系统中有着不同的性能.通过测试不同应用在这些运行系统上的性能,给出了实验比较和性能分析,从而为大数据处理时选择合适的并行编程模型提供依据.
|
| 关 键 词： | 云计算 MapReduce Hadoop Phoenix Twister |
Performance Testing and Analysis among Different MapReduce Runtime Systems |
| |
| YI Xiu-wen,LI Tian-rui,ZHANG Jun-bo and TENG Fei.Performance Testing and Analysis among Different MapReduce Runtime Systems[J].Computer Science,2015,42(5):24-27, 50. |
| |
| Authors: | YI Xiu-wen LI Tian-rui ZHANG Jun-bo and TENG Fei |
| |
| Affiliation: | School of Information Science and Technology,Southwest Jiaotong University,Chengdu 610031,China,School of Information Science and Technology,Southwest Jiaotong University,Chengdu 610031,China,School of Information Science and Technology,Southwest Jiaotong University,Chengdu 610031,China and School of Information Science and Technology,Southwest Jiaotong University,Chengdu 610031,China |
| |
| Abstract: | With the development of cloud computing technology,several implementations which adopt MapReduce mo-del,e.g.,Hadoop,Phoenix and twister,have been developed.Hadoop has high scalability and stability,thus is suitable for handling large-scale off-line applications.The primary advantage of Phoenix,which is especially appropriate for data-intensive tasks,is its processing speed.Twister,a lightweight iterative runtime system,is designed for iterative applications.Different applications produce different levels of performance on different MapReduce runtime systems.By testing various applications using the aforementioned runtime systems,the experimental comparison and performance analysis were presented,providing the basis for the selection of parallel programming models for big data processing. |
| |
| Keywords: | Cloud computing MapReduce Hadoop Phoenix Twister |
| 本文献已被 万方数据 等数据库收录！ |
|
点击此处可从《计算机科学》下载全文 |
|
