Operational Aspects of Dealing with the Large BaBar Data Set

Abstract

To date, the BaBar experiment has stored over 0.7PB of data in an Objectivity/DB database. Approximately half this data-set comprises simulated data of which more than 70% has been produced at more than 20 collaborating institutes outside of SLAC. The operational aspects of managing such a large data set and providing access to the physicists in a timely manner is a challenging and complex problem. We describe the operational aspects of managing such a large distributed data-set as well as importing and exporting data from geographically spread BaBar collaborators. We also describe problems common to dealing with such large datasets.