Auctioning Cluster Resources

Lee Or Alon, Orna Agmon Ben-Yehuda, Sigal Oren

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Organizational clusters are often shared among users who compete over resources, but the organization's goal is to increase the overall utility produced by the cluster. That is, to increase the aggregate benefit drawn from the cluster. To overcome this problem, we enhanced SLURM with an auctioning system. We evaluated our work on several real cluster traces. Our auctioning scheduler increases the responsiveness to highly valued jobs. It reduces the weight of the queued jobs (the sum of multiplication of jobs by their wait time by their number of required nodes and by their bid) by 3X - 33X compared with backfilling.

Original languageEnglish
Title of host publicationHPDC 2022 - Proceedings of the 31st International Symposium on High-Performance Parallel and Distributed Computing
PublisherAssociation for Computing Machinery, Inc
Pages285-286
Number of pages2
ISBN (Electronic)9781450391993
DOIs
StatePublished - 27 Jun 2022
Event31st International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2022 - Virtual, Online, United States
Duration: 27 Jun 202230 Jun 2022

Publication series

NameHPDC 2022 - Proceedings of the 31st International Symposium on High-Performance Parallel and Distributed Computing

Conference

Conference31st International Symposium on High-Performance Parallel and Distributed Computing, HPDC 2022
Country/TerritoryUnited States
CityVirtual, Online
Period27/06/2230/06/22

Keywords

  • auction
  • cluster resources
  • slurm

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Computer Science Applications
  • Software

Fingerprint

Dive into the research topics of 'Auctioning Cluster Resources'. Together they form a unique fingerprint.

Cite this