Research on the Quality of Service of Docker Container based on the Linux Control Group

DOI

10.2991/iceeecs-16.2016.107How to use a DOI?

Keywords

container; Quality of Service; Docker; Virtualization

Abstract

Container, such as Docker, is becoming more and more popular in recent years in cloud computing area. More and more people started using container as their choice to run services on the cloud. While containers are just like processes on the Linux, in fact, so it is quite difficult to do resource management like traditional virtual machine does. To meet different demands of cloud users in using cloud computing resources under the cloud computing environment, especially disk I/O resource, an effective disk I/O QoS (Quality of Service) for Docker or some other container-based cloud computing technique should be used. Thus, we designed dDiskQoS, a solution to do disk I/O QoS for container-based cloud computing technique, which can effectively manage and control disk I/O traffic. The key idea of dDiskQoS is to leverage the advantage of Linux cgroups technique and implement a multi-weighted algorithm to satisfy the disk I/O requirements from both time-sensitive containers and big data throughput containers. The evaluations show that the solution can meet the demands of different scenarios while accepting different parameters, to guarantee system efficiency. Additionally, there is very small overhead in loading our QoS solution. With dDiskQoS, operators can keep more containers running in one host, which will certainly save the cost and energy.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Weicheng Sun
PY  - 2016/12
DA  - 2016/12
TI  - Research on the Quality of Service of Docker Container based on the Linux Control Group
BT  - Proceedings of the 2016 4th International Conference on Electrical & Electronics Engineering and Computer Science (ICEEECS 2016)
PB  - Atlantis Press
SP  - 529
EP  - 533
SN  - 2352-538X
UR  - https://doi.org/10.2991/iceeecs-16.2016.107
DO  - 10.2991/iceeecs-16.2016.107
ID  - Sun2016/12
ER  -