Research Internships 2005-2006 (Sujets de MASTER recherche)

Immediate Services for the GRID

While Grid computing is becoming mainstream, calling for higher level of abstraction for distributed programming, low-level communication primitives have been used to optimize communications:
        active messages, one-sided communications (MPI and the likes), async {} construct in X10,  immediate services.
The overall objective of this research Master is to propose solutions for gaining the potential efficiency of active messages communications within the framework of high-level Grid environment, such as the ProActive library.

The OASIS INRIA Sophia Antipolis project has been designing and implementing  ProActive, a Java library for the GRID. In the framework of an active object model, the library features Asynchronous Typed Messages, Future Synchronizations (Wait-by-necessity), Group Communications, Mobility, Security, and a generic component model. At the infrastructure level, XML Deployment Descriptors provide the capacity to deploy on many kind of GRIDs.
In the absence of any syntactical extension, ProActive programmers write standard code. The library is itself extensible by the programmers, using a MOP (Meta-Object Protocol), making the system open for adaptations and optimizations.

The main objective of the research is to study, and propose original and sound solutions, for the integration of  immediate services in such a Grid-oriented platform. Underlying theory of asynchronous distributed objects, such as defined in TDO, should be taken into account. Overall, the work will include conceptual analysis, proposition of models, and practical experiments. The following steps could be followed:

One solution for maintaining ProActive properties could be to provide a high-level framework or abstraction, featuring immediate services in a tamed manner.

Depending of the student wish and skills, the research can be more or less experimental, more or less theoretical.
The prototype implementation and experimentations will take advantage of the ProActive platform, in the framework of asynchronous and group communication, high-level synchronization (wait-by-necessity).

This research should lead to a PhD program (Thèse de Doctorat).


This research is a joint work between INRIA-I3S Sophia Antipolis and Università di Pisa:

Nice Sophia Antipolis Advisor :  Denis Caromel
Téléphone : 04 92 38 76 31 Email : Denis.Caromel@inria.fr
Laboratoire ou équipe : INRIA Sophia Antipolis -- I3S -- CNRS



Prerequisite : Some knowledge in Object-oriented languages, Distributed Programming.

Hardware and software to be used : Networks of PCs, Clusters, Intranet and Internet P2P machines

Internship location:
       Sophia Antipolis, between Nice and Cannes, France



Bibliography:
 
A Theory of Distributed Objects
 
Asynchrony - Mobility - Groups - Components  
Denis Caromel - Ludovic Henrio
Springer, 2005, XXXII, 346 p., Hardcover, ISBN: 3-540-20866-6
 

The X10 Programming Language

Philippe Charles, Christopher Donawa, Kemal Ebcioglu, Christian Grothoff, Allan Kielstra, Christoph von Praun, Vijay Saraswat, Vivek Sarkar


Towards Seamless Computing and Metacomputing in Java
D. Caromel, W. Klauser, J. Vayssiere,
pp. 1043--1061 in Concurrency Practice and Experience,
September-November 1998, 10(11--13), Editor Geoffrey C. Fox, Published by Wiley & Sons.

Efficient, Flexible and Typed Group Communications for Java
Laurent Baduel, Francoise Baude, Denis Caromel
Joint ACM Java Grande - ISCOPE 2002 Conference, Seattle, Washington, November 3-5, 2002. bibtex

From Distributed Objects to Hierarchical Grid Components
Francoise Baude and Denis Caromel and Matthieu Morel
International Symposium on Distributed Objects and Applications (DOA), Catania, Sicily, Italy, 3-7 November 2003. bibtex

Asynchronous and Deterministic Objects
Denis Caromel, Ludovic Henrio, Bernard Serpette,
POPL'04, Proceedings of the 31st ACM Symposium on Principles of Programming Languages, 2004, Venice, Italy. bibtex

ProActive Home page

Active Messages from NOE team

High performance MPI-2 one-sided communication over InfiniBand.
Weihang Jiang, Jiuxing Liu, Hyun-Wook Jin, Dhabaleswar K. Panda, William Gropp, and Rajeev Thakur.
In Proc. of 4th IEEE/ACM Int'l Symp. on Cluster Computing and the Grid, April 2004.


Send-receive considered harmful: Myths and realities of message passing
Sergei Gorlatch
ACM Transactions on Programming Languages and Systems (TOPLAS)
Volume 26 ,  Issue 1  (January 2004), Pages: 47 - 56 , 2004, ISSN:0164-0925

Performance and Experience with LAPI -- A New High-Performance Communication Library for the IBM RS/6000 SP
G. Shah C. Bender
Proceedings of the 12th. International Parallel Processing Symposium,
Page: 260 , Year of Publication: 1998, ISSN:1063-7133

Low-Latency Communication on the IBM RISC System/6000 SP,
C. Chang, G. Czajkowski, C. Hawblitzel, and T. von Eicken,
SuperComputing'96