CDF Monte Carlo Production on LCG Grid via LcgCAF Portal

The improvements of the luminosity of the Tevatron Collider require large increases in computing requirements for the CDF experiment which has to be able to increase proportionally the amount of Monte Carlo data it produces. This is, in turn, forcing the CDF collaboration to move beyond the use of dedicated resources and to exploit grid resources. CDF has been running a set of CDF Analysis Farm (CAFs), which are submission portals to dedicated pools, and LcgCAF is basically a reimplementation of the CAF model in order to access grid resources by using the LCG/EGEE middleware components. By mean of LcgCAF CDF users can submit analysis jobs with the same mechanism adopted for the dedicated farms and at the same time the grid resources are accessed without any specific software requirements for the sites. This is obtained using Parrot for the experiment code distribution and Frontier for the run condition database availability on the worker nodes. Currently many sites in Italy and in Europe are accessed through this portal in order to produce Monte Carlo data and in one year of operations we expect about 100,000 grid jobs submitted by the CDF users. We review here the setup used to submit jobs and retrieve the output, including the grid components CDF-specific configuration. The batch and interactive monitor tools developed to allow users to verify the jobs status during their lifetimes in the grid environment are described. We analyze the efficiency and typical failure modes of the current grid infrastructure reporting the performances of different parts of the used system.