DiRAC-II Technical working group project support

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Physics and Astronomy

Abstract

Distributed Research utilising Advanced Computing (DiRAC) is an integrated set of high-performance computing facilities supporting theoretical research in particle physics, astrophysics and nuclear physics, which is using computer simulation to obtain a deeper understanding of standard models, to seek signals of new physics by confronting them with experiment and observation, and to develop new models that provide better descriptions of nature.

The integration and smooth operation and monitoring of this distributed facility is
role of the Technical Working Group (TWG). Services the TWG must provide include
a unified Systems Administration Framework, giving a single web presence for the acquisition
of user accounts, management of project membership, generation of usage reports, help desk and so on.
Central to this is a single DiRAC-ID given to each user, and this is also used to give access
to community web pages documenting the DiRAC systems and enabling the research communities to perform their research.

This work to enable integration, operation and efficient utilisation of all DiRAC facilities by RAC-selected projects will be carried out
according to priorities set by the DiRAC Technical Working Group.

Planned Impact

The University of Edinburgh contribution to the design, prototyping and testing of BlueGene/Q enabled IBM to bring a better product to market and created this major computing platform for DiRAC. The ongoing software development and exploitation for some of the most computationally demanding science in the world will help to ensure these machines are utilised effectively for a wide range of applications that benefit the economy, the environment and our health, and, by training members of the DiRAC Consortium in using the most advanced supercomputer technology, it will enhance computing skills in the workforce.

The UK HPC community needs tools for managing and integrating the multiplicity of machines at Tier-2 and below. The SAFE system has been developed by the
University of Edinburgh for management of multiple supercomputer architectures, and has been successfully used to manage the UK's Tier-1 national resources
such as HeCTOR and HPCX for some time. Another significant impact of this project is demonstrating and hardening this tool for the integration of multiple
Tier-2 resources in multiple installation sites.

Publications

10 25 50