Pan: A High-Level Configuration Language
Lionel Cons, Piotr Poznański · 2002
The computational requirements for the new Large Hadron Collider are enormous: 5-8 PetaBytes of data generated annually with analysis requiring 10 more PetaBytes of disk storage and the equivalent of 200,000 of today’s fastest PC processors. This will be a very large and complex computing system, with about two thirds of the computing capacity installed in ‘‘regional computing centres’ ’ across Europe, America, and Asia. Implemented as a global computational grid, the goal of integrating the large geographically distributed computing fabrics presents challenges in many areas, including: distributed scientific applications; computational grid middleware, automated computer system management; high performance networking; object database management; security; global grid operations. This paper describes our approach to one of these challenges: the configuration management of a large number of machines, be they nodes in large clusters or desktops in large organizations.