Middleware 2003

ACM/IFIP/USENIX International Middleware Conference

Rio Othon Palace Hotel

Rio de Janeiro, Brazil

16-20 June 2003

[ Home ] [ Program ] [ Registration ] [ Organization ] [ Sponsors ] [ CFP ] [ Important Dates ] [ Submissions ] [ Travel ]

 

Tutorials
Keynote Speech
Posters
Workshops
Work in Progress
Student Travel Grants
Student Volunteer Program

 

 

Website Mirror at the University of São Paulo

Tutorial 4: Object-Oriented Middleware and Components for the Grid

Abstract

Object technology and middleware are increasingly used within parallel and distributed computing. At the same time, components are becoming a very effective tool for the composition and deployment of business applications. While the difficulties to actually program and deploy on the Grid have been, to a great extend, acknowledged, components technology is in the process of bringing many advances for such parallel programming. The aim of this course is to explain and detail this evolution.

The course objectives are:

  • to explain the main principles of component technology,
  • to explain how object-oriented middleware can be used for parallel and distributed programming,
  • to state and detail how object-oriented middleware, together with components turn out to be very effective for the Grid.

The course is based on widespread, industrial languages and frameworks, such as Java, Corba, CCM. Real time demos should be featured, and the presentation will leave time for questions and interactive discussions.

Table of Contents

1. Software Components: from Business to Grid

  • Basics of Components
  • From Objects to Components
  • EJB, CCM standards
  • Main Contributions and use for the Grid

2. Java Parallel Objects and Components

2.1 Introduction

  • Main concepts
  • Main benefits
  • Main difficulties

2.2 A Panorama of Java-based frameworks

  • mpiJava,
  • Manta Object-Based Collective Comm.
  • MPJ
  • etc.

2.3 The ProActive middleware

  • Parallel and Active Objects
  • Group Communications (vs. MPI)
  • Mobility and P2P
  • Hierarchical Components, Deployment
  • Security
  • Use case and performances

3. CORBA Parallel Objects and Components

3.1 A Panorama of CORBA-like frameworks

  • Data Parallel Corba (DPC)
  • Common Component Architecture (CCA)
  • etc.

3.2 GridCCM: Parallel Corba Components

  • Main Concepts of Parallel Components
  • Data redistribution
  • Sequential/Parallel Component Interoperability
  • Legacy code
  • Deployment
  • Use case and performances

4. Conclusion

  • When to use those techniques
  • Perspectives:
  • Interaction with WebServices
  • Peer-to-peer Deployment

Target Audience

This tutorial is of interest to researchers, students, and professionals who are interested in learning how middleware and component technology can turn the Grid into an effective tool for parallel and distributed execution.

Speakers Profile

Denis Caromel is full professor at University of Nice-Sophia Antipolis (UNSA). He is also member of the Institut Universitaire de France (IUF), a multi-disciplinary national academia that select a few professors based on the excellence of their research records. His research interests include parallel, concurrent, and distributed object-oriented programming, the semantics of sequential and parallel languages. He spent two years and an half in Santa Barbara, California, as a research Engineer, and also working on his Ph.D. In February 1991, he received his Ph.D. from the University of Nancy, on Asynchronous and Imperative Parallel Programming. From 1990 to 1991, he was Assistant Professor at this same university. Then he moved to an Associate Professor position at the University of Nice - Sophia Antipolis in September 1991, and obtained his Full Professor position at that same university in 1998. Altogether, he has published more than 50 scientific papers in referred and international journals and conferences. He also edited 5 volumes of Lecture Notes. He gave many invited talks on Object and Distributed Computing at various universities around the world (including Jet Propulsion Laboratory, Berkeley, Stanford, ISI, USC, Electrotechnical Laboratory Tsukuba, Univ. of Sydney, Univ. of Adelaide, Univ. Federal de Rio, University College London, European Science Foundation). He was also an invited visiting scientist at various research institutions (including Digital System Research Center in Palo Alto, and NASA Langley Research Center in Hampton, Virginia).

Christian Perez is full time INRIA researcher at IRISA. His research interests include parallel and distributed object and component programming model, middleware for Grid computing and efficient execution of parallel languages. He is born in May 1972. He received a M.Sc. in Computer Science (1995) and his Ph.D. (1999) from the Ecole Normale Superieur de Lyon. His Ph.D. is on the integration of automatic load balancing support in data parallel program execution. In September 2000, he obtained a full researcher position. He is the coordinator of GRID-RMI (http://www.irisa.fr/Grid-RMI), a software project of the French Action Concertée Incitative (ACI), that aims at promoting object and component programming models for computing Grids.

Latest update: 13 June 2003 - Questions and Comments about the Site: fmc@inf.ufg.br