Modern Web applications exploit Cloud infrastructures to scale their resources and cope with sudden changes in the workload. While the state of practice is to focus on dynamically adding and removing virtual machines, we advocate that there are strong benefits in containerizing the applications and in scaling the containers. In this paper we present an autoscaling technique that allows containerized applications to scale their resources both at the virtual machine (VM) level and at the container level. Furthermore, applications can combine this infrastructural adaptation with platform-level adaptation. The autoscaling is made possible by our planner, which consists of a grey-box discrete-Time feedback controller. The work has been validated using two application benchmarks deployed to Amazon EC2. Our experiments show that our planner outperforms Amazon's AutoScaling by 78% on average without containers; and that the introduction of containers allows us to improve by yet another 46% on average.

A discrete-Time feedback controller for containerized cloud applications

BARESI, LUCIANO;GUINEA MONTALVO, SAM JESUS;LEVA, ALBERTO;QUATTROCCHI, GIOVANNI
2016-01-01

Abstract

Modern Web applications exploit Cloud infrastructures to scale their resources and cope with sudden changes in the workload. While the state of practice is to focus on dynamically adding and removing virtual machines, we advocate that there are strong benefits in containerizing the applications and in scaling the containers. In this paper we present an autoscaling technique that allows containerized applications to scale their resources both at the virtual machine (VM) level and at the container level. Furthermore, applications can combine this infrastructural adaptation with platform-level adaptation. The autoscaling is made possible by our planner, which consists of a grey-box discrete-Time feedback controller. The work has been validated using two application benchmarks deployed to Amazon EC2. Our experiments show that our planner outperforms Amazon's AutoScaling by 78% on average without containers; and that the introduction of containers allows us to improve by yet another 46% on average.
2016
Proceedings of the ACM SIGSOFT Symposium on the Foundations of Software Engineering
9781450342186
9781450342186
Adaptive Systems; Cloud Computing; Containers; Control Theory; Software Adaptation; Software
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1009052
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 82
  • ???jsp.display-item.citation.isi??? 57
social impact