The BaBar experiment is taking data since 1999, investigating the
violation of charge and parity (CP) symmetry in the field of High Energy
Physics. Event simulation is an intensive computing task, due to the complexity
of algorithm based on Monte-Carlo method implemented using the GEANT engine.
Data needed as input for the simulation, stored in the ROOT format, are
classified into two categories: conditions data for describing the detector
status when data are recorded, and background triggers data for including noise
signal necessary to obtain a realistic simulation. In order to satisfy these
requirements, in the traditional BaBar computing model events are distributed
over several sites involved in the collaboration where each site manager
centrally manages a private farm dedicated to simulation production. The new
grid approach applied to the BaBar production framework is discussed along with
the schema adopted for data deployment via Xrootd servers, including
data management using grid middleware on distributed storage facilities spread
over the INFN-GRID network. A comparison between the two models is provided,
describing also the custom application developed for performing the whole
production task on the grid and showing results achieved.