Measurement and Evaluation for Docker Container Networking

被引:27
作者
Zeng, Hao [1 ]
Wang, Baosheng [1 ]
Deng, Wenping [1 ]
Zhang, Weiqi [1 ]
机构
[1] Natl Univ Def Technol, Coll Comp, Changsha, Hunan, Peoples R China
来源
2017 INTERNATIONAL CONFERENCE ON CYBER-ENABLED DISTRIBUTED COMPUTING AND KNOWLEDGE DISCOVERY (CYBERC) | 2017年
关键词
Cloud Computing; Data Center; Container; Container Network; Model; Performance;
D O I
10.1109/CyberC.2017.78
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Container technology is a virtualization technology at the operating system level. It is more in line with future data center needs for its lightweight resource management, second-level start-up features. As the bridge between containers, container network plays a very important role in container life cycle system. This paper investigates and analyzes the current status of the development of container network, mainly container network model and network solutions. Besides we design experiments to measure and evaluate the performance of three mainstream network solutions (Flannel, Swarm Overlay, Calico). Our experiments show that Calico is of the highest performance and its TCP throughput is almost the same to host, but the configuration is the most complicated; Flannel configuration is simple, but it is designed for Kubernetes and the performance is not high; Swarm overlay configuration is simple with low performance. For it is originally native to Docker, its potential applications are very considerable.
引用
收藏
页码:105 / 108
页数:4
相关论文
共 10 条
[1]  
Baier Jonathan., 2015, Getting Started with Kubernetes
[2]  
Bernstein D., 2014, IEEE CLOUD COMPUTING
[3]  
Boettiger C., 2015, ACM SIGOPS OPERATING
[4]  
Brewer E., 2015, ACM S
[5]  
Bui T., 2015, COMPUTER SCI
[6]  
Haines R., 2016, J PHARM SCI
[7]  
Mangano N., 2010, IEEE ACM INT C AUT S
[8]  
Merkel D., 2014, LINUX J, DOI DOI 10.5555/2600239.2600241
[9]  
Nesmachnow S., 2016, SCHEDULING ALGORITHM
[10]  
Vohra D., 2016, Kubernetes microservices with docker