CMS computing performance on the GRID during the second year of LHC collisions

The CMS Computing Model deployed on the Worldwide LHC Computing GRID (WLCG) has successfully met its milestones of efficiently processing the first year LHC data and serving them to thousands of physicists worldwide as input to their searches for new physical phenomena. The second year of LHC collision running is characterized by a spectacular expected increase in integrated luminosity by similar to 3 orders of magnitude (to O(5) fb(-1)), resulting in a data volume of O(2) billion events. This requires the experiment to ramp up resources, adapt computing services and move towards sustained computing operations. In parallel, the experience gained on the GRID and emerging technologies lead to a natural evolution of the CMS Computing Model. We review the main computing achievements by CMS including the host laboratory processing, data distribution and re-processing on the GRID, data serving, and analysis by CMS physicists during the second year of LHC operations. We also present the evolution of the usage of GRID resources by CMS to adapt to constantly growing data volume and processing demands. We conclude with prospects for upcoming years CMS data processing.