Contact Person
Name Andrew McNab
Address Department of Physics and Astronomy,
Schuster Laboratory,
University of Manchester,
M13 9PL.
Phone 0161 275 4227
Fax 0161 273 5867
Email mcnab@hep.man.ac.uk
Title A Technical Overview of the GridPP Testbed
Author 1 Andrew McNab
The University of Manchester
Author is PC Member No
Main Fields 1. Middleware
4. Science Application
Other Main Fields [None Given]

Abstract 

A Technical Overview of the GridPP Testbed

Keywords: High Energy Physics, EU DataGrid, Job Submission, Data 
Management, Security, Monitoring

The future of Particle Physics is dominated by the Large Hadron Collider 
that is under construction at CERN. At a centre of mass energy of 14 TeV, 
the LHC will be the most powerful accelerator in the world for many years 
if not decades to come. Due to the enormously high energy, high luminosity 
and complexity of the detectors, the LHC experiments will produce 
unprecedented amounts of data, estimated to be several PetaBytes per year, 
for offline analysis by teams of physicists all over the world. To analyse 
this data and to generate the Monte Carlo simulated data necessary to 
understand it will require huge amounts of computing and data storage 
facilities. These computing resources will be distributed across the 
world, linked together as a Grid.

GridPP is currently building a prototype UK Grid that will enable the four 
LHC experiments, ATLAS, LHCb, CMS and ALICE, to generate large amounts of 
Monte Carlo simulated data. This is currently being tested by running 
experiments in the USA in which the UK is involved, BaBar at SLAC and CDF 
and D0 at the Tevatron, FNAL. In doing this, the largest Grid testbed in 
the UK has been created, consisting of more than 100 servers across 16 
Institutes, incorporating a functional Grid job submission system.

This testbed is largely built upon the common core software base deployed 
as part of the EU-wide European DataGrid middleware development programme.

Components of this include a distributed monitoring infrastructure of
resources, services and networks; data and metadata catalogues, 
replication and management tools; mass disk and tape storage management 
interfaces to the Grid; distributed authentication and authorization 
infrastructures for multiple virtual organisations; a job submission and 
brokering system, with dynamic allocation of jobs to resources using the 
monitoring and data location services; fabric installation, management and 
monitoring systems at the scale of thousands of hosts per site.

Remarks [None Given]