Component-distinguishable Co-location and Resource Reclamation for High-throughput Computing

doi:10.1145/3630006

	Component-distinguishable Co-location and Resource Reclamation for High-throughput Computing
	Zhao, Laiping 1; Cui, Yushuai 1; Yang, Yanan 1; Zhou, Xiaobo 1; Qiu, Tie 1; Li, Keqiu 1; Bao, Yungang 2
	2024-05-01
发表期刊	ACM TRANSACTIONS ON COMPUTER SYSTEMS
ISSN	0734-2071
卷号	42 期号:1-2 页码:37
摘要	Cloud service providers improve resource utilization by co-locating latency-critical (LC) workloads with best-effort batch (BE) jobs in datacenters. However, they usually treat multi-component LCs as monolithic applications and treat BEs as "second-class citizens" when allocating resources to them. Neglecting the inconsistent interference tolerance abilities of LC components and the inconsistent preemption loss of BE workloads can result in missed co-location opportunities for higher throughput. We present Rhythm, a co-location controller that deploys workloads and reclaims resources rhythmically for maximizing the system throughput while guaranteeing LC service's tail latency requirement. The key idea is to differentiate the BE throughput launched with each LC component, that is, components with higher interference tolerance can be deployed together with more BE jobs. It also assigns different reclamation priority values to BEs by evaluating their preemption losses into a multi-level reclamation queue. We implement and evaluate Rhythm using workloads in the form of containerized processes and microservices. Experimental results show that it can improve the system throughput by 47.3%, CPU utilization by 38.6%, and memory bandwidth utilization by 45.4% while guaranteeing the tail latency requirement.
关键词	Datacenters resource utilization tail latency co-locating
DOI	10.1145/3630006
收录类别	SCI
语种	英语
资助项目	National Key Research and Development Program of China[2022YFB 4500702] ; Shandong Provincial Natural Science Foundation[ZR2022LZH018] ; National Natural Science Foundation of China[62141218] ; National Natural Science Foundation of China[62372322] ; Zhejiang Lab[2021DA0AM01/003]
WOS研究方向	Computer Science
WOS类目	Computer Science, Theory & Methods
WOS记录号	WOS:001229267600002
出版者	ASSOC COMPUTING MACHINERY
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/40049
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Yang, Yanan; Li, Keqiu
作者单位	1.Tianjin Univ, Coll Intelligence & Comp, Tianjin Key Lab Adv Networking, Tianjin, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Zhao, Laiping,Cui, Yushuai,Yang, Yanan,et al. Component-distinguishable Co-location and Resource Reclamation for High-throughput Computing[J]. ACM TRANSACTIONS ON COMPUTER SYSTEMS,2024,42(1-2):37.
APA	Zhao, Laiping.,Cui, Yushuai.,Yang, Yanan.,Zhou, Xiaobo.,Qiu, Tie.,...&Bao, Yungang.(2024).Component-distinguishable Co-location and Resource Reclamation for High-throughput Computing.ACM TRANSACTIONS ON COMPUTER SYSTEMS,42(1-2),37.
MLA	Zhao, Laiping,et al."Component-distinguishable Co-location and Resource Reclamation for High-throughput Computing".ACM TRANSACTIONS ON COMPUTER SYSTEMS 42.1-2(2024):37.