Running a Production Grid Site at the London e-Science Centre
David Mcbride, Marko Krznarić, John Darlington, Olivier Van Der Aa, Mona Aggarwal, D. J. Colling · 2013
This paper describes how the London e-Science Centre cluster MARS, a production 400+ Opteron CPU cluster, was integrated into the production Large Hadron Collider Compute Grid. It describes the practical issues that we encountered when deploying and maintaining this system, and details the techniques that were applied to resolve them. Finally, we provide a set of recommendations based on our experiences for grid software development in general that we believe would make the technology more accessible. production Grid system sufficiently powerful to meet the demands of the LHC. Today, the worldwide LCG deployment currently spans some 136 sites in 36 different countries, offering to its users access to a total of nearly 14,000 CPUs and approximately 8PB of online storage. This paper describes how we integrated the LeSC MARS cluster into the production LHC Grid. It describes the practical issues that we encountered when deploying and maintaining this system as well as the techniques that applied to resolve them. 2. Overview of the LCG Architecture 1.