Passer à la navigation principale Passer à la recherche Passer au contenu principal

EnLoB: Energy and load balancing-driven container placement strategy for data centers

  • École de technologie supérieure
  • Tomsk Polytechnic University
  • School of Postgraduate Studies and Research

Résultats de recherche: Chapitre dans un livre, rapport, actes de conférence › Participation à un ouvrage collectif lié à un colloque ou une conférence › Revue par des pairs

15 Citations (Scopus)

Résumé

Cloud data centers (DCs) can be aptly regarded as the epicenter of today's business and economy; which support seamless data processing, analysis, and storage. However, various studies advocate that the existing DCs are often underutilized. To be precise, almost 30\% of the installed DCs in the United States are comatose. In addition to this, the existing DC architecture leads to extensive energy utilization, which severely hampers the environment and places a severe risk on the power sector. Thus, it is highly essential to reduce DC's energy utilization through efficient resource consolidation approaches. In this work, we investigate the joint impact of resource consolidation and load balancing on cutting down the energy utilization indices of the cloud DCs. In this vein, we formulate a multi-objective optimization problem (MOOP) for container placement across heterogeneous infrastructure, primarily with the intent to minimize the overall energy consumption and balance the load amongst the operating hosts. However, due to the hardness of the underlying problem and its infeasibility to furnish optimal solutions in polynomial time, we designed an online solution based on the incremental exploration of the solution space to map containers on the available array of hosts such that the objectives mentioned above can be attained. Finally, we evaluated the performance of the proposed algorithm in contrast to an existing algorithm on real-time workload traces obtained from PlanetLab. The obtained results confirm the superior performance of the proposed algorithm relative state-of-the-art.

Langue originaleAnglais
titre2019 IEEE Globecom Workshops, GC Wkshps 2019 - Proceedings
EditeurInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronique)9781728109602
Les DOIs
ÉtatPublié - déc. 2019
Evénement2019 IEEE Globecom Workshops, GC Wkshps 2019 - Waikoloa, Etats-Unis
Durée: 9 déc. 2019 → 13 déc. 2019

Série de publications

Nom2019 IEEE Globecom Workshops, GC Wkshps 2019 - Proceedings

Conférence

Conférence2019 IEEE Globecom Workshops, GC Wkshps 2019
Pays/TerritoireEtats-Unis
La villeWaikoloa
Période9/12/19 → 13/12/19

SDG des Nations Unies

Ce résultat contribue à ou aux Objectifs de développement durable suivants

  1. SDG 7 – Energie propre et d'un coût abordable
    SDG 7 – Energie propre et d'un coût abordable

Classification disciplinaire ASJC Scopus

  • Réseaux informatiques et communications
  • Matériel et architecture
  • Logiciel
  • Contrôle et optimisation

Empreinte digitale

Voici les principaux termes ou expressions associés à « EnLoB: Energy and load balancing-driven container placement strategy for data centers ». Ces libellés thématiques sont générés à partir du titre et du résumé de la publication. Ensemble, ils forment une empreinte digitale unique.

Citer cette ressource