Abstract
Grid computing promises access to large amounts of computing power, but so far adoption of Grid computing has been limited to highly specialized experts for three reasons. First, users are used to batch systems, and interfaces to Grid software are often complex and different to those in batch systems. Second, users are used to having transparent file access, which Grid software does not conveniently provide. Third, efforts to achieve wide-spread coordination of computers while solving the first two problems is hampered when clusters are on private networks. Here we bring together a variety of software that allows users to almost transparently use Grid resources as if they were local resources while providing transparent access to files, even when private networks intervene. As a motivating example, the BaBar Monte Carlo production system is deployed on a truly distributed environment, the European DataGrid, without any modification to the application itself. Copyright (c) 2005 John Wiley & Sons, Ltd.
Original language | English |
---|---|
Pages (from-to) | 787-801 |
Number of pages | 15 |
Journal | Concurrency and Computation-practice & Experience |
Volume | 18 |
Issue number | 7 |
DOIs | |
Publication status | Published - 1 Jun 2006 |
Externally published | Yes |
Keywords
- Grid computing
- high-energy physics
- virtual file system
- user interface
- Monte Carlo production