Distributed analysis in the ATLAS experiment.
D. L. Adams, R. W. L. Jones · Lancaster EPrints (Lancaster University) · 2005
The ATLAS experiment, based at the Large Hadron Collider at CERN, Geneva, is currently developing a grid-based distributed system capable of supporting it's data analysis activities which require the processing of data volumes of the order of petabytes per year. The distributed analysis system aims to bring the power of computation on a global scale to thousands of physicists by enabling them to easily tap into the vast computing resource of various grids such as LCG, gLite, OSG and Nordugrid, for their analysis activities whilst shielding them from the complexities of the grid environment. This paper outlines the ATLAS distributed analysis model, the ATLAS data management system, and the multi-pronged ATLAS strategy for distributed analysis in a heterogeneous grid environment. Various frontend clients and backend submission systems will be discussed before concluding with a status update of the system.