PEGASEF: A provenance-based big data service framework for efficient simulation execution on shared computing clusters

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

1 Scopus citations

Abstract

Over the past years high-performance computing (HPC) simulation programs have been aggressively employed to solve complex problems in a variety of computational science and engineering disciplines. As those programs are shared in an online platform, many users can easily run their simulations on the platform as long as they are connected on the web. However, repetitive simulations from users have charged a significant burden on the platform’s limited computing and storage resources. To address the concern of inefficiency in simulation execution, we propose a big data service framework based on past simulation records. Such records are called provenances, which capture various properties in simulation. By utilizing the provenances, the platform can perform more efficient simulations via duplicate elimination and assist users with enhanced simulation service such as result prediction, execution-time estimation, and input-parameter clustering.

Original languageEnglish
Title of host publicationBig Data Applications and Services 2017 - The 4th International Conference on Big Data Applications and Services
EditorsCarson K. Leung, Wookey Lee
PublisherSpringer Verlag
Pages175-182
Number of pages8
ISBN (Print)9789811306945
DOIs
StatePublished - 2019
Event4th International Conference on Big Data Applications and Services, BigDAS 2017 - Tashkent, Uzbekistan
Duration: 15 Aug 201718 Aug 2017

Publication series

NameAdvances in Intelligent Systems and Computing
Volume770
ISSN (Print)2194-5357

Conference

Conference4th International Conference on Big Data Applications and Services, BigDAS 2017
Country/TerritoryUzbekistan
CityTashkent
Period15/08/1718/08/17

Keywords

  • Big data
  • Duplicate elimination
  • Efficient simulation
  • HPC
  • Provenance
  • Service framework

Fingerprint

Dive into the research topics of 'PEGASEF: A provenance-based big data service framework for efficient simulation execution on shared computing clusters'. Together they form a unique fingerprint.

Cite this