Large scale distributed systems

Responsable / In charge of : Fabrice Huet (Fabrice.HUET@univ-cotedazur.fr)

Résumé / Abstract :
Large scale distributed infrastructures leverage high performance networks to federate computing resources from multiple institutions. Distributed computing has become a critical tool in many scientific disciplines. This lecture describes the foundation of distributed computing infrastructures, including the main computing models exploited in Grids and Clouds to evolve from cluster computing towards virtualized resources and cross-institutional user communities.

Prérequis / Prerequisite :
The course is given in english.
Although the course is self-contained, notions of distributed systems, parallel computing and probabilities will help.

Objectifs / Objectives :

  • Become familiar with large-scale distributed computing infrastructures
  • Learn distributed computing principles and underlying technologies
  • Identify distributed computing capabilities and limitations
  • Design performing distributed applications
  • Be alert to emerging technologies and research trends


Contenu / Contents :

  • Distributed computing and models
  • Remote services
  • Grid infrastructures
  • Workload and performance modeling
  • Workflows for distributed computing infrastructures
  • Cross-institutional security
  • Distributed data management


Références / References :

• I. Foster, K. Kesselman, “The GRID 2: blueprint for a new computing infrastructure”. Elsevier, 2004. • D. Lingrand, J. Montagnat, J. Martyniak, D. Colling. "Optimization of jobs submission on the EGEE production grid: modeling faults using workload" in Journal of Grid Computing, 8(2):305-321, Springer, March 2010.
• E. Caron, F. Desprez. “DIET: A Scalable Toolbox to Build Network Enabled Servers on the Grid”. International Journal of High Performance Computing Applications, 20(3):335-352, 206.
• T. Truong Huu, G. Koslovski, F. Anhalt, J. Montagnat, P. Vicat-Blanc Primet. "Joint Elastic Cloud and Virtual Network Framework for Application Performance-cost Optimization" in Journal of Grid Computing (JOGC), 9 (1), pages 27-47, Springer, March 2011.

Acquis / Knowledge :

Evaluation / Assessment :
Two written exams with a coefficient of 50% each : one focused on the lectures and one focused on the labs.