Table of Contents Author Guidelines Submit a Manuscript
Scientific Programming
Volume 15 (2007), Issue 4, Pages 213-234
http://dx.doi.org/10.1155/2007/516174

GRIDCC: A Real-Time Grid Workflow System with QoS

A. Stephen McGough,1 Asif Akram,1 Li Guo,1 Marko Krznaric,1 Luke Dickens,1 David Colling,2 Janusz Martyniak,2 Roger Powell,3 Paul Kyberd,3 Chenxi Huang,3 Constantinos Kotsokalis,4 and Panayiotis Tsanakas4

1London e-Science Centre, Imperial College London, London, UK
2High Energy Physics Group, Imperial College London, London, UK
3Brunel University, School of Engineering & Design, Uxbridge, UK
4Computing Systems Laboratory, National Technical University of Athens, Athens, Greece

Received 30 November 2007; Accepted 30 November 2007

Copyright © 2007 Hindawi Publishing Corporation. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

The over-arching aim of Grid computing is to move computational resources from individual institutions where they can only be used for in-house work, to a more open vision of vast online ubiquitous `virtual computational' resources which support individuals and collaborative projects. A major step towards realizing this vision is the provision of instrumentation – such as telescopes, accelerators or electrical power stations – as Grid resources, and the tools to manage these resources online. The GRIDCC project attempts to satisfy these requirements by providing the following four co-dependent components; a flexible wrapper for publishing instruments as Grid resources; workflow support for the orchestration of multiple Grid resources in a timely manner; the machinery to make reservation agreements on Grid resources; and the facility to satisfy quality of service (QoS) requirements on elements within workflows. In this paper we detail the set of services developed as part of the GRIDCC project to provide the last three of these components. We provide a detailed architecture for these services along with experimental results from load testing experiments. These services are currently deployed as a test-bed at a number of institutions across Europe, and are poised to provide a 'virtual lab' to production level applications.