Dynamic scheduling of scientific workflow applications on the grid: a case study

  • Authors:
  • Radu Prodan;Thomas Fahringer

  • Affiliations:
  • University of Innsbruck, Technikerstraße, Innsbruck, Austria;University of Innsbruck, Technikerstraße, Innsbruck, Austria

  • Venue:
  • Proceedings of the 2005 ACM symposium on Applied computing
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

The existing Grid workflow scheduling projects do not handle recursive loops which are characteristic to many scientific problems. We propose a hybrid approach for scheduling Directed Graph (DG)-based workflows in a Grid environment with dynamically changing computational and network resources. Our dynamic scheduling algorithm is based on the iterative invocation of classical static Directed Acyclic Graphs (DAGs) scheduling heuristics generated using well-defined cycle elimination and task migration techniques. We approach the static scheduling problem as an application of a modular optimisation tool using genetic algorithms. We report successful implementation and experimental results on a pilot real-world material science workflow application.