A test has been carried out to study the feasibility of running part of the raw data reconstruction on the Open Science Grid. The results of that test are presented, with a breakdown of what worked well and what did not. A description of follow-on work to address these issues is also given.