Programming Parallel and Distributed Systems for Large Scale Numerical Simulation Application Christian Perez INRIA researcher IRISA Rennes, France.

Slides:



Advertisements
Similar presentations
XtreemOS IP project is funded by the European Commission under contract IST-FP XtreemOS: Building and Promoting a Linux-based Operating System.
Advertisements

Elton Mathias and Jean Michael Legait 1 Elton Mathias, Jean Michael Legait, Denis Caromel, et al. OASIS Team INRIA -- CNRS - I3S -- Univ. of Nice Sophia-Antipolis,
A Workflow Engine with Multi-Level Parallelism Supports Qifeng Huang and Yan Huang School of Computer Science Cardiff University
All-in-one graphical tool for grid middleware management Eddy Caron, Abdelkader Amar, Frédéric Desprez, David Loureiro LIP ENS Lyon, INRIA Rhône-Alpes,
Christian Delbe1 Christian Delbé OASIS Team INRIA -- CNRS - I3S -- Univ. of Nice Sophia-Antipolis November Automatic Fault Tolerance in ProActive.
GridRPC Sources / Credits: IRISA/IFSIC IRISA/INRIA Thierry Priol et. al papers.
A Nation Wide Experimental Grid The Grid’5000 project: architecture and objectives Building a nation wide experimental platform for Grid researchers –
1 Software & Grid Middleware for Tier 2 Centers Rob Gardner Indiana University DOE/NSF Review of U.S. ATLAS and CMS Computing Projects Brookhaven National.
Session 2: task 3.2 GCM, Kracow, June l Current status of GCM Denis Caromel (10 mn each talk) l Wrapping CCA Components as GCM Components Maciej.
Applying the collective entity concept to POP-C++ Pierre Kuonen & Christian Pérez EIA-Fr & INRIA WP3 – Task 3.1 Krakow Meeting, June 27 th 2006.
Dynamic adaptation of parallel codes Toward self-adaptable components for the Grid Françoise André, Jérémy Buisson & Jean-Louis Pazat IRISA / INSA de Rennes.
Denis Caromel1 Denis Caromel, et al. OASIS Team INRIA -- CNRS - I3S -- Univ. of Nice Sophia-Antipolis, IUF 3 rd ProActive User Group, Nov Model.
Grid’5000 Grid' DAS-3 workshop 104/12/06 Grid’5000 * DAS-3 – Grid'5000 workshop December 4th, *5000 CPUs Pierre NEYRON - INRIA.
Overview of grid / cloud research in France Michel DAYDÉ Scientific Delegate at INS2/CNRS in charge of HPC / Grid / cloud Université de Toulouse - IRIT.
Design and Implementation of a Single System Image Operating System for High Performance Computing on Clusters Christine MORIN PARIS project-team, IRISA/INRIA.
SOA, BPM, BPEL, jBPM.
Optimized Java computing as an application for Desktop Grid Olejnik Richard 1, Bernard Toursel 1, Marek Tudruj 2, Eryk Laskowski 2 1 Université des Sciences.
JuxMem: An Adaptive Supportive Platform for Data Sharing on the Grid Gabriel Antoniu, Luc Bougé, Mathieu Jan IRISA / INRIA & ENS Cachan, France Workshop.
ICOM 5995: Performance Instrumentation and Visualization for High Performance Computer Systems Lecture 7 October 16, 2002 Nayda G. Santiago.
Abstractions: Programming and deploying apps. on Grids Franck Cappello INRIA* (*this is my own opinion!) CCGRID’08 - Panel.
1 CCA Meeting, Januray 25th 2007 Supporting the Master-Worker Paradigm in the Common Component Architecture Hinde Lilia Bouziane, Christian Pérez, Thierry.
The Grid Component Model: an Overview “Proposal for a Grid Component Model” DPM02 “Basic Features of the Grid Component Model (assessed)” -- DPM04 CoreGrid.
Hinde Bouziane – CBHPC’08 – October 2008 Marco ALDINUCCI and Marco DANELUTTO UNIPI - University of Pisa (Italy) Hinde Lilia BOUZIANE and Christian.
Architecting Web Services Unit – II – PART - III.
The Grid Component Model and its Implementation in ProActive CoreGrid Network of Excellence, Institute on Programming Models D.PM02 “Proposal for a Grid.
Large-scale Deployment in P2P Experiments Using the JXTA Distributed Framework Gabriel Antoniu, Luc Bougé, Mathieu Jan & Sébastien Monnet PARIS Research.
Large Scale Sky Computing Applications with Nimbus Pierre Riteau Université de Rennes 1, IRISA INRIA Rennes – Bretagne Atlantique Rennes, France
G-JavaMPI: A Grid Middleware for Distributed Java Computing with MPI Binding and Process Migration Supports Lin Chen, Cho-Li Wang, Francis C. M. Lau and.
Peer-to-Peer Distributed Shared Memory? Gabriel Antoniu, Luc Bougé, Mathieu Jan IRISA / INRIA & ENS Cachan/Bretagne France Dagstuhl seminar, October 2003.
Master Worker Paradigm Support in Software Component Models Hinde Bouziane, Christian Pérez PARIS Research Team INRIA/IRISA Rennes ANR CIGC LEGO (ANR-05-CICG-11)
Heavy and lightweight dynamic network services: challenges and experiments for designing intelligent solutions in evolvable next generation networks Laurent.
Programming High Performance Applications using Components Outline High-Performance applications and code coupling The CORBA Component Model CCM in the.
Issues in (Financial) High Performance Computing John Darlington Director Imperial College Internet Centre Fast Financial Algorithms and Computing 4th.
JuxMem: An Adaptive Supportive Platform for Data Sharing on the Grid Gabriel Antoniu, Luc Bougé, Mathieu Jan IRISA / INRIA & ENS Cachan, France Grid Data.
Building Hierarchical Grid Storage Using the GFarm Global File System and the JuxMem Grid Data-Sharing Service Gabriel Antoniu, Lo ï c Cudennec, Majd Ghareeb.
The JuxMem-Gfarm Collaboration Enhancing the JuxMem Grid Data Sharing Service with Persistent Storage Using the Gfarm Global File System Gabriel Antoniu,
OpenCCM MdC Philippe Merle LIFL - INRIA (soon)
Refining middleware functions for verification purpose Jérôme Hugues Laurent Pautet Fabrice Kordon
Towards high-performance communication layers for JXTA on grids Mathieu Jan GDS meeting, Lyon, 17 February 2006.
Issues Autonomic operation (fault tolerance) Minimize interference to applications Hardware support for new operating systems Resource management (global.
1 INRIA’s strategy Jean-Pierre Banâtre. 2 Some figures. About 3,500 persons including: –1,600 employed by INRIA –staff from partner institutions –personnel.
Latest news on JXTA and JuxMem-C/DIET Mathieu Jan GDS meeting, Rennes, 11 march 2005.
Grid programming with components: an advanced COMPonent platform for an effective invisible grid © 2006 GridCOMP Grids Programming with components. An.
GRID Overview Internet2 Member Meeting Spring 2003 Sandra Redman Information Technology and Systems Center and Information Technology Research Center National.
George Goulas, Christos Gogos, Panayiotis Alefragis, Efthymios Housos Computer Systems Laboratory, Electrical & Computer Engineering Dept., University.
Cluster Software Overview
 Copyright 2005 Digital Enterprise Research Institute. All rights reserved. Enabling Components Management and Dynamic Execution Semantic.
Distributed Components for Integrating Large- Scale High Performance Computing Applications Nanbor Wang, Roopa Pundaleeka and Johan Carlsson
ProActive components and legacy code Matthieu MOREL.
ORNL/IU Workshop on Computational Frameworks in Fusion Oak Ridge, TN, USA, 9 th 2005 Defining, Implementing, Executing and Deploying a Parallel Component.
DISCOGRID Sophia, mars 2006 The Padico Environment Christian Pérez PARIS Research Team, IRISA/INRIA, Rennes, France.
1 BBN Technologies Quality Objects (QuO): Adaptive Management and Control Middleware for End-to-End QoS Craig Rodrigues, Joseph P. Loyall, Richard E. Schantz.
International Symposium on Grid Computing (ISGC-07), Taipei - March 26-29, 2007 Of 16 1 A Novel Grid Resource Broker Cum Meta Scheduler - Asvija B System.
Making a DSM Consistency Protocol Hierarchy-Aware: An Efficient Synchronization Scheme Gabriel Antoniu, Luc Bougé, Sébastien Lacour IRISA / INRIA & ENS.
OpenCCM: Status and Work plan Dr. Philippe Merle LIFL - INRIA ObjectWeb Architecture Meeting, Grenoble, 21 – 22.
GRID ANATOMY Advanced Computing Concepts – Dr. Emmanuel Pilli.
ANR CIGC LEGO (ANR-CICG-05-11) Bordeaux, 2006, December 11 th Automatic Application Deployment on Grids Landry Breuil, Boris Daix, Sébastien Lacour, Christian.
Advanced Component Models ULCM & HLCM Julien Bigot, Hinde Bouziane, Christian Perez COOP Project Lyon, 9-10 mars 2010.
GridRPC Sources / Credits: IRISA/IFSIC IRISA/INRIA Thierry Priol et. al papers.
CBHPC’08: Component-Based High Performance Computing (16/10/08) 1 A GCM-Based Runtime Support for Parallel Grid Applications Elton Mathias, Françoise Baude.
High Performance Flexible DSP Infrastructure Based on MPI and VSIPL 7th Annual Workshop on High Performance Embedded Computing MIT Lincoln Laboratory
Toward a Distributed and Parallel High Performance Computing Environment Johan Carlsson and Nanbor Wang Tech-X Corporation Boulder,
Denis Caromel1 OASIS Team INRIA -- CNRS - I3S -- Univ. of Nice Sophia-Antipolis -- IUF IPDPS 2003 Nice Sophia Antipolis, April Overview: 1. What.
Towards a High Performance Extensible Grid Architecture Klaus Krauter Muthucumaru Maheswaran {krauter,
Grid Institute Scientific Council, September 10, 2008
Abstract Machine Layer Research in VGrADS
Inventory of Distributed Computing Concepts and Web services
Inventory of Distributed Computing Concepts
MPJ: A Java-based Parallel Computing System
The Grid Component Model and its Implementation in ProActive
Presentation transcript:

Programming Parallel and Distributed Systems for Large Scale Numerical Simulation Application Christian Perez INRIA researcher IRISA Rennes, France

2 IRISA INRIA Rennes CNRS Univ. of Rennes I INSA 600 people (march 06) 233 researchers 184 PhD students 120 Engineers, technicians

3 Project team composition Tenured personnel (11) F. André (Prof IFSIC) G. Antoniu (CR INRIA) J-P. Banâtre (Prof IFSIC) L. Bougé (Prof ENS) Y. Jégou (CR INRIA) D. Margery (IR INRIA) 50% C. Morin (DR INRIA) P. Morillon (IE IFSIC) J.L. Pazat (Prof. INSA) C. Perez (CR INRIA) T. Priol (DR INRIA) Post-docs (5) PhD students (11) Engineers (6)

4 PARIS objectives Objects of study: Cluster and Cluster of Clusters (aka Grids) Main objectives: Study and design operating systems, middleware and runtime systems to make the programming of such computing infrastructures easier with: High-performance High-availability Scalability Design advanced programming models for the programming of Clusters of Clusters Combining both parallel and distributed computing paradigms Evaluation of the proposed operating systems, runtimes and middleware through the development of advanced software (not only software prototype) Technology transfer through collaboration with industrial partners

5 Research activities Operating System and Runtime systems for Clusters and Cluster Federations Single System Image Operating System Grid-aware Operating System Middleware for Computational Grids Component-based middleware for computational Grids Communication framework Parallel components Deployment of parallel components within a grid Adaptive components Large-scale data management for Grids Coupling of Distributed Shared Memories Data sharing services for mutable data using a P2P approach Advanced Models for the Grids High-order Gamma Enactment of workflows based on a chemical metaphor Experimental Grid Infrastructures The Grid’5000 testbed

6 Orsay 1000 (684) Rennes 518 (658) Bordeaux 500 (96) Toulouse 500 (116) Lyon 500 (252) Grenoble 500 (270) Sophia Antipolis 500 (434) Lille 500 (106) Nancy: 500 (94) Grid’5000 testbed 10 Gbps Dark fiber Dedicated Lambda Fully isolated traffic! Provided by RENATER

7 Component-based middleware for computational Grids Apply modern software practices to scientific computing High-performance & Scalability Specialized component models for high performance computing (clusters & grids) Code coupling applications Parametric applications Increase the level of abstraction SPMD paradigm (MxN communications) Master-worker paradigm Data sharing paradigm High-performance communication Independence vis à vis of the networking technologies Adaptation Adaptation to the dynamic behavior of grids Deployment Map components to available resources Technology independent (CCM, CCA, Fractal, CoreGRID GCM) SPMD components Master-Slaves component ThermalDynamics Structural Mechanics Optics

8 High Performance Components for code coupling: SPMD paradigm SPMD component Parallelism is a non-functional property of a component It is an implementation issue Collection of sequential components SPMD execution model Support of distributed arguments API for data redistribution API for communication scheduling w.r.t. network properties Support of parallel exceptions Object Request Broker CORBA stub/skeleton Communication Library ( MPI) Application Application view management - Data distribution description Communication management - Comm. Matrix computation - Comm. Matrix scheduling - Communication execution Redistribution Library 1 Communication Library GridCCM runtime Scheduling Library Component AComponent B

9 High Performance Components for parametric codes: Master-worker paradigm Collection of components Simple model for application developers A request is delivered to an instance of the collection The selection of collection instance is delegated to a request transport policy Enable the use of existing MW environments (DIET, ….) Resources infrastructure independence No dealing with the number of workers No dealing with request delivery concerns Valid for ADL and non ADL based component models Preliminary results Fluid motion estimation Experiment on Grid’ components 974 processors 7 sites Speedup of 213 with a round robin pattern Round-Robin Programmer view binding master worker Exposed provided port RR Proxy Set of “request delivery” patterns Number of workers & policy pattern selection Resources Abstract ADL Concrete ADL XML collection definition worker

10 PadicoTM: Communication framework Provides an open integration platform to combine various communication middleware and runtimes Message based runtimes (MPI, PVM, …) DSM-based runtimes (TreadMarks, …) RPC and RMI based middleware (DCE, CORBA, Java, …) Allows several communication middleware and runtime to share various networking technologies Ethernet, Myrinet, Infiniband, Quadrics, SCI Last but not least: get the maximum performance of the network!  Available as an open source software under the GPL licence > 200 downloads since July 2002 Madeleine Portability across networks Marcel I/O aware multi-threading MyrinetSCI PadicoTM Core PadicoTM Services Multithreading Networks DSMJVMMPICORBA JXTA TCP Personality Layer Internal engine Mpich OmniORB MICO Orbacus Orbix/E KaffeJuxMem Mome

11 ADAGE: Automatic Deployment of Application in a Grid Environment Deploy a same application on any kind of resources from clusters to grids Support multi-middleware application MPI+CORBA+JUXMEM+... Planner as plugin round robin & random Some successes JXTA peers on ~400 nodes 4003 components on 974 processors on 7 sites Alpha support for dynamic application MPI Application Description CCM Application Description Resource Description Generic Application Description Control Parameters Deployment Planning Deployment Plan Execution Application Configuration