DocumentCode :
2934219
Title :
iGrid: Interactive Grid
Author :
Meoni, Marco
Author_Institution :
Ecole Polytech. Fed. de Lausanne, Lausanne, Switzerland
fYear :
2011
fDate :
5-8 Dec. 2011
Firstpage :
189
Lastpage :
196
Abstract :
Grid computing is traditionally applied for batch jobs in the scientific and academic computing. This is also the case at WLCG, the global computing infrastructure providing the production and analysis environments for the LHC experiments at CERN. In this paper we envision the next generation Grid computing systems to support High Energy Physics (HEP) online sessions. We refine a resource management framework for enabling on-demand private virtual clusters on the AliEn Grid middleware and run applications on an online computing environment. We describe the software components distributed among the submission node and the site front-end and execution nodes. Our experiment evaluates the scalability of the agents deployed at the execution nodes and measures the performance of PROOF - the de-facto standard software for HEP data analysis - on our Grid environment rather than on a local cluster. Because PROOF generates time-sensitive I/O and requires to handle potentially thousands of network connections, we describe a multiplexed I/O approach as the ideal solution for scalable and high-performing proxy services.
Keywords :
data analysis; grid computing; middleware; AliEn Grid middleware; CERN; HEP data analysis; LHC experiments; PROOF; academic computing; batch jobs; de-facto standard software; global computing infrastructure; high energy physics online sessions; high-performing proxy services; iGrid; interactive grid; multiplexed IO approach; network connections; next generation grid computing systems; on-demand private virtual clusters; online computing environment; resource management framework; scientific computing; software components; time-sensitive IO; Grid computing; Instruction sets; Java; Middleware; Monitoring; Sockets; I/O multiplexing; grid; interactivity; job agent; latency; online computing; pool thread;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
E-Science (e-Science), 2011 IEEE 7th International Conference on
Conference_Location :
Stockholm
Print_ISBN :
978-1-4577-2163-2
Type :
conf
DOI :
10.1109/eScience.2011.34
Filename :
6123277
Link To Document :
بازگشت