OpenMOLE

OpenMOLE (Open MOdeL Experiment) is a workflow management system, making it simple to run applications on distributed computing environments. The only requirement is that this application can be run on Linux.

The typical usages of OpenMOLE are high performance model calibration, model exploration, machine learning, optimization, data processing, …

OpenMOLE is:

  • Programming language agnostic: it supports any JVM-enabaled language (Java, Scala, NetLogo, …) natively, compiled languages through Linux binaries (C, C++, …), and scripts whose interpreter is available on Linux (R, Python, SciLab, …)
  • Designed for Distributed computing - Works on multi-core machines, clusters, grids, desktop grid.
  • Expressive – with a workflow language to describe a naturally parallel processes.
  • Scalable - Handles millions of tasks, years of computation, and GBs of data.
  • Mature - Developed since 2008 and widely used.
  • Free and Open Source – under the AGPLv3 free software license.

Related publications:

  • Passerat-Palmbach et al., Frontiers Neuroinformatics 2017
  • Reuillon et al., HPCS 2015
  • Passerat-Palmbach et al., HPC-MICCAI 2014
  • Reuillon et al., FGCS 2013
  • Reuillon et al., HPCS 2010

Components of the OpenMOLE platform

This work is done in collaboration with Romain Reuillon and Mathieu Leclaire.