Resource Utilization Analysis of Alibaba Cloud

被引:8
作者
Deng, Li [1 ,2 ]
Ren, Yu-Lin [1 ,2 ]
Xu, Fei [1 ,2 ]
He, Heng [1 ,2 ]
Li, Chao [3 ]
机构
[1] Wuhan Univ Sci & Technol, Coll Comp Sci & Technol, Wuhan 430065, Hubei, Peoples R China
[2] Hubei Prov Key Lab Intelligent Informat Proc & Re, Wuhan 430065, Hubei, Peoples R China
[3] Hubei Univ, Dept Informat Dev & Management, Wuhan 430062, Hubei, Peoples R China
来源
INTELLIGENT COMPUTING THEORIES AND APPLICATION, PT I | 2018年 / 10954卷
基金
中国国家自然科学基金;
关键词
Cloud platform; Online services; Batch jobs; Resource utilization ratio; EFFICIENT;
D O I
10.1007/978-3-319-95930-6_18
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Currently, low resource utilization and high costs of cloud platform are becoming big challenges to cloud provider. However, due to confidentiality, few cloud platform providers are willing to publish resource utilization data of their cloud platform. This poses great difficulties in designing an effective cloud resource scheduler. Fortunately, Alibaba released its cloud resource usage data in September 2017. This paper analyzes Alibaba cloud trace data deeply from different aspects and reveals some important features of resource utilization. These features will help to design effective resource management approaches for cloud platform: (1) The maximum resource utilization of online services is closely related to their average utilization. (2) The longer a batch instance runs, the longer it may last. (3) The type of job that runs in a container can be estimated according to the amount of consumed resources and life time of this container. (4) Actual resources used by different batch jobs vary with time greatly and static resource allocation would make resource wasted seriously.
引用
收藏
页码:183 / 194
页数:12
相关论文
共 13 条
[1]  
Abadi M, 2016, PROCEEDINGS OF OSDI'16: 12TH USENIX SYMPOSIUM ON OPERATING SYSTEMS DESIGN AND IMPLEMENTATION, P265
[2]  
[Anonymous], 2011, P USENIX NSDI
[3]  
[Anonymous], P 10 EUROPEAN C COMP
[4]  
[Anonymous], 2012, Proceedings of the 3rd ACM Symposium on Cloud Computing (SOCC), DOI 10.1145/2391229.2391236
[5]   Resource Central: Understanding and Predicting Workloads for Improved Resource Management in Large Cloud Platforms [J].
Cortez, Eli ;
Bonde, Anand ;
Muzio, Alexandre ;
Russinovich, Mark ;
Fontoura, Marcus ;
Bianchini, Ricardo .
PROCEEDINGS OF THE TWENTY-SIXTH ACM SYMPOSIUM ON OPERATING SYSTEMS PRINCIPLES (SOSP '17), 2017, :153-167
[6]   HCloud: Resource-Efficient Provisioning in Shared Cloud Systems [J].
Delimitrou, Christina ;
Kozyrakis, Christos .
ACM SIGPLAN NOTICES, 2016, 51 (04) :473-488
[7]   Quasar: Resource-Efficient and QoS-Aware Cluster Management [J].
Delimitrou, Christina ;
Kozyrakis, Christos .
ACM SIGPLAN NOTICES, 2014, 49 (04) :127-143
[8]   Synthesis of Au-Ni Bimetallic Nanoparticles with Tunable Aspect Ratio [J].
Han Tianhao ;
Liang Fuxin ;
Yang Zhenzhong .
CHEMICAL JOURNAL OF CHINESE UNIVERSITIES-CHINESE, 2017, 38 (11) :1921-1926
[9]  
[李超 Li Chao], 2017, [小型微型计算机系统, Journal of Chinese Computer Systems], V38, P1945
[10]  
Lu C., 2017, P IEEE INT C BIG DAT