Skip to main content

TS

Four-fold auto-scaling on a contemporary deployment platform using docker containers

Authors

Philipp Hoenisch, Ingo Weber, Stefan Schulte, Liming Zhu and Alan Fekete

Vienna University of Technology
Austria

NICTA

UNSW

University of Sydney

Abstract

With the advent of Docker, it became popular to bundle Web applications (apps) and their libraries into lightweight linux containers and offer them to a wide public by deploying them in the cloud. Compared to older approaches, which deploy apps in a cloud-provided virtual machine (VM), the use of containers allows faster start-up and less overhead. However, having containers inside VMs makes the decision about elastically scaling up or down more flexible but also more complex. In this contemporary approach to service provisioning, four aspects of scaling have to be considered: each of the VM and the containers can be adjusted horizontally (changes in the number of instances) and vertically (changes in the power of the instances). In this paper, we address this four-fold auto-scaling by forming the scaling decision as a multi-objective optimization problem. We evaluated our approach with realistic apps, and found that using our approach, we can reduce the average cost per request by about 20-28%.

BibTeX Entry

  @inproceedings{Hoenisch_WSZF_15,
    author           = {Hoenisch, Philipp and Weber, Ingo and Schulte, Stefan and Zhu, Liming and Fekete, Alan},
    year             = {2015},
    month            = nov,
    keywords         = {auto-scaling, optimization, docker, light-weight containers, cloud},
    title            = {Four-fold Auto-scaling on a Contemporary Deployment Platform using Docker Containers},
    booktitle        = {International Conference on Service Oriented Computing},
    pages            = {316-323},
    address          = {Goa, India}
  }

Download

Served by Apache on Linux on seL4.