A Grid architectural approach applied for backward compatibility to a production system for events simulation.

Distributed systems paradigm gained in popularity during the last 15 years, thanks also to the broad diffusion of distributed frameworks proposed for the Internet plat form. In the late ’90s a new concept started to play a main role in the field of distributed computing: the Grid. This thesis presents a study related to the integration between the BaBar’s framework, an experiment belonging to the High Energy Physics field, and a grid system like the one implemented by the Italian National Institute for Nuclear Physics (INFN), the INFNGrid project, which provides support for several research domains. The main goal was to succeed in adapt an already well established system, like the one implemented into the BaBar pipeline and based on local centers not interconnected between themselves, to a kind of technology that was not ready by the time the experiment’s framework was designed. Despite this new approach was related just to some aspects of the experiment, the production of simulated events by using MonteCarlo methods, the efforts here described represent an example of how an old experiment can bridge the gap toward the Grid computing, even adopting solutions designed for more recent projects. The complete evolution of this integration will be explained starting from the earlier stages until the actual development to state the progresses achieved, presenting results that are comparable with production rates gained using the conventional BaBar’s approach, in order to examine the potentially benefits and drawbacks on a concrete case study.