Enabling provenance on large scale e-Science applications

Abstract

Large-scale e-Science experiments present unprecedented data handling requirements with their multi-petabyte data storages. Complex software applications, such as the ATLAS High Energy Physics experiment at CERN, run throughout Grid computing sites around the world in a distributed environment, with scientists performing concurrent analysis on data and producing new data products shared among the collaboration. In this paper, we introduce a multi-phase infrastructure to achieve data provenance for an e-Science experiment. We propose an infrastructure to integrate provenance onto an existing legacy application with strong emphasis on scalability and explore the relationship between provenance and metadata introducing a model where data provenance is made available as metadata through a separate reasoning phase. © Springer-Verlag Berlin Heidelberg 2006.