We propose a parallel and distributed component framework for building Grid applications, adapted to the hierarchical, highly distributed, highly heterogeneous nature of Grids. This framework is based on ProActive, a middleware (programming model and environment) for object oriented parallel, mobile, and distributed computing. We have extended ProActive by implementing a hierarchical and dynamic component model, named Fractal, so as to master the complexity of composition, deployment, re-usability, and e#ciency of grid applications. This defines a concept of Grid components, that can be parallel, made of several activities, and distributed. These components communicate using typed one-to-one or collective invocations.

this paper investigates the current suitability of such object-oriented middleware for High-Performance and Grid programming

This paper presents two contributions to the study of component deployment for distributed real-time and embedded (DRE) systems. First, it uses an inventory tracking systems (ITS) as a case study to elicit challenges involved in deploying DRE systems to account for their quality of service requirements. Second, it describes how we designed and implemented the Deployment And Configuration Engine (DAnCE), which is QoS-enabled middleware that addresses the challenges that arose in the context of our ITS case study. Our experience shows that DAnCE provides an e#ective platform for deploying DRE system components using a standard runtime environment and metadata.

Group communication is a crucial feature for high-performance and Grid computing. While previous works and libraries proposed such a characteristic (e.g. MPI, or object-oriented frameworks), the use of groups imposed specific constraints on programmers -- for instance the use of dedicated interfaces to trigger group communications. We aim at a more flexible mechanism...

ABSTRACT. Software components turn out to be a convenient model to build complex applications for scientific computing and to run them on a computational grid. However, deploying complex, component-based applications in a grid environment is particularly arduous. To prevent the user from directly dealing with a large number of execution hosts and their heterogeneity within a grid, the application deployment phase must be as automatic as possible. This paper describes an architecture for automatic deployment of component-based applications on computational grids. In the context of the CORBA Component Model (CCM), this paper details all the steps to achieve an automatic deployment of components as well as the entities involved: a grid access middleware and its grid information service (like OGSI), a component deployment model, as specified by CCM, an enriched application description and a deployment planner in order to select resources and map components onto computers. RÉSUMÉ. Les composants logiciels sont une solution bien adaptée pour construire des applications complexes de calcul scientifique destinées à être exécutées sur une grille de calcul. Cependant, le déploiement d’applications complexes à base de composants sur une grille est une tâche particulièrement ardue. Pour éviter d’avoir à faire face directement au grand nombre d’ordinateurs de la grille et à leur hétérogénéité, la phase de déploiement d’application doit être automatisée. Cet article décrit une architecture de déploiement automatique d’applications à base de composants sur grille de calcul. En partant du modèle de composants CORBA (CCM), ce papier détaille les étapes du déploiement de composants et les acteurs en présence: un intergiciel d’accès aux ressources de la grille (à l’instar de OGSI), un modèle de déploiement de composants, une description étendue de l’application et un planificateur de déploiement.

This paper presents a Peer-to-Peer (P2P) infrastructure that supports a large scale grid. The P2P infrastructure is implemented in Java and federates Java Virtual Machines (JVMs) for computation. The management of shared JVMs is decentralized, self-organized, and configurable. The P2P infrastructure was deployed as a permanent desktop grid, with which we have achieved a computation record by solving the NQueens problem for 25 queens. Thereafter, we have mixed this desktop grid with five highly heterogeneous clusters from the Grid’5000 platform. We analyze the behavior of this thousand CPU grid with two communicating applications: NQueens and Flow-Shop.

we present the design and implementation of a numerical simulation for electromagnetic waves propagation. A sequential Java design and implementation is first presented. Further, a distributed and parallel version is derived from the first, using an active object pattern. In addition, benchmarks are presented on this non embarrassingly parallel application.

This thesis describes the design and implementation of Babylon v2.0. Babylon v2.0 is a 100% Java compatible framework for building parallel, distributed and mobile applications in Java. Babylon v2.0 incorporates features like object migration, asynchronous method invocation and remote class loading while providing an easy-to-use interface that enables seamless interaction with remote objects and hides the complexities of remote messaging protocols that are normally large part of distributed systems programming. The potential cluster computing benefits of Babylon v2.0 are demonstrated by the evaluation results which show that sequential Java applications can achieve significant performance gains by using Babylon v2.0 to parallelize their work across a cluster of workstations. Intuitive interfaces, ease of use, support for multiple simultaneous users, and services and features that facilitate the development and administration of distributed systems make Babylon v2.0 a unique and powerful system for distributed systems programmers.

Grid applications must be able to cope with large variations in deployment: from intra-domain to multiple domains, going over private, to virtually-private, to public networks. As a consequence, the security should not be tied up in the application code, but rather easily configurable in a flexible, and abstract manner. Moreover, any large scale Grid application using hundreds or thousands of nodes will have to cope with migration of computations, for the sake of load balancing, change in resource availability, or just node failures.

This article presents an evolution of classical SPMD programming for clusters and grids.