In this paper the technical solutions, the usage and the future development of the Swiss ATLAS Grid are presented. In 2009 the Swiss ATLAS Grid consists of four clusters with about 2000 shared computing cores and about 250 TB of disk space. It is based on middlewares provided by the NorduGrid Collaboration and the EGEE project. It supports multiple virtual organisations and uses additional middleware, developed by the ATLAS collaboration, for data management. The Swiss ATLAS grid is interconnected with both NorduGrid and the Worldwide LHC Grid. This infrastructure primarly serves Swiss research institutions working within the ATLAS experiment at LHC, but is open for about two thousand users on lower priority. The last three years about 80 000 wall clock time days have been processed by ATLAS jobs on the Swiss ATLAS Grid.
[1]
R. W. Hamming.
State of the art in scientific computing
,
1963,
AFIPS '63 (Spring).
[2]
G. Aad,et al.
The ATLAS Experiment at the CERN Large Hadron Collide
,
2008
.
[3]
David E. Culler,et al.
The ganglia distributed monitoring system: design, implementation, and experience
,
2004,
Parallel Comput..
[4]
Farid Ould-Saada,et al.
Data Management for the World's Largest Machine
,
2006,
PARA.
[5]
Anders Wäänänen,et al.
Advanced resource connector middleware for lightweight computational Grids
,
2007
.