Scheduling Distributed Applications: the SimGrid Simulation Framework

  • Authors:

  • Arnaud Legrand;Loris Marchal;Henri Casanova

  • Affiliations:

  • -;-;-

  • Venue:
  • CCGRID '03 Proceedings of the 3st International Symposium on Cluster Computing and the Grid
  • Year:
  • 2003

Quantified Score

Hi-index 0.01

Visualization

Abstract

Since the advent of distributed computersystems an active field of research has been the investigation of scheduling strategies for parallel applications.The common approach is to employ scheduling heuristics that approximate an optimal schedule. Unfortunately, it is often impossible to obtain analytical resultsto compare the efficacy of these heuristics. One possibility is to conducts large numbers of back-to-back experiments on real platforms. While this is possible ontightly-coupled platforms, it is infeasible on modern distributed platforms (i.e. Grids as it is labor-intensiveand does not enable repeatable results. The solutionis to resort to simulations. Simulations not only enablesrepeatable results but also make it possible to explorewide ranges o plat orm and application scenarios.In this paper we present the SimGrid frameworkwhich enables the simulation o distributed applications in distributed computing environments or the specific purpose of developing and evaluating schedulingalgorithms. This paper focuses on SimGrid v2, whichgreatly improves on the first version of the software withmore realistic network models and topologies. SimGridv2 also enables the simulation of distributed schedulingagents, which has become critical or current scheduling research in large-scale platforms. After describingand validating these features, we present a case studyby which we demonstrate the usefulness of SimGrid forconducting scheduling research.