Publication Details

Performance-Cost Optimization of Moldable Scientific Workflows

JAROŠ, M.; JAROŠ, J. Performance-Cost Optimization of Moldable Scientific Workflows. In Job Scheduling Strategies for Parallel Processing. Lecture Notes in Computer Science. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Portland, Oregon USA: Springer International Publishing, 2021. p. 149-167. ISBN: 978-3-030-88223-5. ISSN: 0302-9743.
Czech title
Výkonostně cenová optimalizace tvarovatelných sestav úloh pro vědecké účely
Type
conference paper
Language
English
Authors
Jaroš Marta, Ing., Ph.D. (DCSY)
Jaroš Jiří, doc. Ing., Ph.D. (DCSY)
URL
Keywords

task graph scheduling, workflow, genetic algorithm, moldable tasks, makespan estimation

Abstract

Moldable scientific workflows represent a special class of scientific workflows where the tasks are written as distributed programs being able to exploit various amounts of computer resources. However, current cluster job schedulers require the user to specify the amount of resources per task manually. This often leads to suboptimal execution time and related cost of the whole workflow execution since many users have only limited experience and knowledge of the parallel efficiency and scaling. This paper proposes several mechanisms to automatically optimize the execution parameters of moldable workflows using genetic algorithms. The paper introduces a local optimization of workflow tasks, a global optimization of the workflow on systems with on-demand resource allocation, and a global optimization for systems with static resource allocation. Several objectives including the workflow makespan, computational cost and the percentage of idling nodes are investigated together with a trade-off parameter putting stress on one objective or another. The paper also discusses the structure and quality of several evolved workflow schedules and the possible reduction in makespan or cost. Finally, the computational requirements of evolutionary process together with the recommended genetic algorithm settings are investigated. The most complex workflows may be evolved in less than two minutes using the global optimization while in only 14s using the local optimization.

Published
2021
Pages
149–167
Journal
Lecture Notes in Computer Science, no. 12985, ISSN 0302-9743
Proceedings
Job Scheduling Strategies for Parallel Processing
Series
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISBN
978-3-030-88223-5
Publisher
Springer International Publishing
Place
Portland, Oregon USA
DOI
10.1007/978-3-030-88224-2_8
UT WoS
000869960400008
EID Scopus
2-s2.0-85117449650
BibTeX 
@inproceedings{BUT175770,
  author="Marta {Jaroš} and Jiří {Jaroš}",
  title="Performance-Cost Optimization of Moldable Scientific Workflows",
  booktitle="Job Scheduling Strategies for Parallel Processing",
  year="2021",
  series="Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
  journal="Lecture Notes in Computer Science",
  number="12985",
  pages="149--167",
  publisher="Springer International Publishing",
  address="Portland, Oregon USA",
  doi="10.1007/978-3-030-88224-2\{_}8",
  isbn="978-3-030-88223-5",
  issn="0302-9743",
  url="https://link.springer.com/book/10.1007%2F978-3-030-88224-2"
}
Back to top