MESCAL - 2011 - Annual activity report

MESCAL

MESCAL - 2011

Project Team Mescal

Members

Overall Objectives

Presentation
Objectives
Highlights

Scientific Foundations

Large System Modeling and Analysis
Management of Large Architectures
Migration and resilience
Large scale data management

Application Domains

On-demand Geographical Maps
Nano simulations
Seismic simulations
Electromagnetic Fields simulations
Embedded Systems
Wireless Networks

Software

Tools for cluster management and software development
OAR: Batch scheduler for clusters and grids
FTA: Failure Trace Archive
SimGrid: simulation of distributed applications
TRIVA: interactive trace visualization
$ψ$ and $ψ^{2}$ : perfect simulation of Markov Chain stationary distributions

New Results

Perfect simulation
Economic models for clouds
Game theory and networks
Mean field analysis for networks
Idleness and failure prediction in large infrastructures
Scheduling and Game Theory
Validity study of flow-based network models.
Vizualisation
Experimental methodology
Multi-core platforms
High performance computing
Input-Output

Contracts and Grants with Industry

Contracts with Industry
Grants with Industry

Partnerships and Cooperations

Regional Initiatives
National Initiatives
European Initiatives
International Initiatives

Dissemination

Animation of the scientific community
Teaching

Bibliography

Major publications
Publications of the year

Home | Next next

2011 Team Activity Report

MESCAL

Middleware Efficiently SCALable

Research centre: Grenoble - Rhône-Alpes

In partnership with: CNRS, Institut polytechnique de Grenoble, Université Joseph Fourier (Grenoble 1)

In collaboration with: Laboratoire d'Informatique de Grenoble (LIG)

Field: Networks, Systems and Services, Distributed Computing

Theme: Distributed and High Performance Computing

Keywords: High Performance Computing, Game Theory, Grid'5000, Scheduling, Stochastic Modeling