A Case for Adaptive Resource Management in Alibaba Datacenter Using Neural Networks

doi:10.1007/s11390-020-9732-x

	A Case for Adaptive Resource Management in Alibaba Datacenter Using Neural Networks
	Wang, Sa 1,2,3; Zhu, Yan-Hai 4; Chen, Shan-Pei 4; Wu, Tian-Ze 1,2; Li, Wen-Jie 1,2; Zhan, Xu-Sheng 1,2; Ding, Hai-Yang 4; Shi, Wei-Song 5; Bao, Yun-Gang 1,2,3
	2020
发表期刊	JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY
ISSN	1000-9000
卷号	35 期号:1 页码:209-220
摘要	Both resource efficiency and application QoS have been big concerns of datacenter operators for a long time, but remain to be irreconcilable. High resource utilization increases the risk of resource contention between co-located workload, which makes latency-critical (LC) applications suffer unpredictable, and even unacceptable performance. Plenty of prior work devotes the effort on exploiting effective mechanisms to protect the QoS of LC applications while improving resource efficiency. In this paper, we propose MAGI, a resource management runtime that leverages neural networks to monitor and further pinpoint the root cause of performance interference, and adjusts resource shares of corresponding applications to ensure the QoS of LC applications. MAGI is a practice in Alibaba datacenter to provide on-demand resource adjustment for applications using neural networks. The experimental results show that MAGI could reduce up to 87.3% performance degradation of LC application when co-located with other antagonist applications.
关键词	resource management neural network resource efficiency tail latency
DOI	10.1007/s11390-020-9732-x
收录类别	SCI
语种	英语
资助项目	National Key Research and Development Program of China[2016YFB1000201] ; National Natural Science Foundation of China[61420106013] ; National Natural Science Foundation of China[61702480] ; Youth Innovation Promotion Association of Chinese Academy of Sciences and Alibaba Innovative Research (AIR) Program
WOS研究方向	Computer Science
WOS类目	Computer Science, Hardware & Architecture ; Computer Science, Software Engineering
WOS记录号	WOS:000512098800013
出版者	SCIENCE PRESS
引用统计	被引频次：4[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/14690
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Zhu, Yan-Hai
作者单位	1.Chinese Acad Sci, Inst Comp Technol, State Key Lab Comp Architecture, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China 3.Peng Cheng Lab, Shenzhen 518055, Peoples R China 4.Alibaba Inc, Hangzhou 311121, Peoples R China 5.Wayne State Univ, Dept Comp Sci, Detroit, MI 48202 USA
推荐引用方式 GB/T 7714	Wang, Sa,Zhu, Yan-Hai,Chen, Shan-Pei,et al. A Case for Adaptive Resource Management in Alibaba Datacenter Using Neural Networks[J]. JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,2020,35(1):209-220.
APA	Wang, Sa.,Zhu, Yan-Hai.,Chen, Shan-Pei.,Wu, Tian-Ze.,Li, Wen-Jie.,...&Bao, Yun-Gang.(2020).A Case for Adaptive Resource Management in Alibaba Datacenter Using Neural Networks.JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,35(1),209-220.
MLA	Wang, Sa,et al."A Case for Adaptive Resource Management in Alibaba Datacenter Using Neural Networks".JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY 35.1(2020):209-220.