Locality of sampling and diversity in parallel system workloads

Dror G. Feitelson*

*Corresponding author for this work

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

21 Scopus citations

Abstract

Observing the workload on a computer system during a short (but not too short) time interval may lead to distributions that are significantly different from those that would be observed over much longer intervals. Rather than describing such phenomena using involved non-stationary models, we propose a simple global distribution coupled with a localized sampling process. We quantify the effect by the maximal deviation between the global distribution and the distribution as observed over a limited slice of time, and find that in real workload data from parallel supercomputers this deviation is significantly larger than would be observed at random. Likewise, we find that the workloads at different sites also differ from each other. These findings motivate the development of adaptive systems, which adjust their parameters as they learn about their workloads, and also the development of parametrized workload models that exhibit such locality of sampling, which are required in order to evaluate adaptive systems.

Original languageEnglish
Title of host publicationProceedings of ICS07
Subtitle of host publication21st ACM International Conference on Supercomputing
Pages53-63
Number of pages11
DOIs
StatePublished - 2007
Event21st ACM International Conference on Supercomputing, ICS07 - Seattle, WA, United States
Duration: 17 Jun 200721 Jun 2007

Publication series

NameProceedings of the International Conference on Supercomputing

Conference

Conference21st ACM International Conference on Supercomputing, ICS07
Country/TerritoryUnited States
CitySeattle, WA
Period17/06/0721/06/07

Bibliographical note

Funding Information:
The author would like to thank individually all the participants, from private and public organizations located in France, Switzerland and the UK, that have taken part in the research and interviews. Such support and open discussions highlighting project experiences, have substantively contributed to both match research aim and goals, and provide a handout for practical reflection and learning. The author would like also to acknowledge the help and resources provided by the University of Southampton and the allocated supervisor's contribution.

Keywords

  • Locality
  • Workload characterization
  • Workload modeling

Fingerprint

Dive into the research topics of 'Locality of sampling and diversity in parallel system workloads'. Together they form a unique fingerprint.

Cite this