Session FR: CEU Poster Session
1:00 PM–1:00 PM, Wednesday, September 21, 2005
Ritz-Carlton Hotel Room: Ballroom Foyer 1:00pm - 3:00pm
Sponsoring Units:
DNP JPS
Abstract ID: BAPS.2005.HAW.FR.89
Abstract: FR.00089 : Information Download: Data Management Monitoring and Computing Analysis for STAR at PDSF
Preview Abstract
MathJax On | Off 
Abstract 
Authors:
Ariel Fleming
(Tennesse State University)
Doug Olson
Eric Hjort
(Lawrence Berkeley National Laboratory)
Data management and computing analysis for the Solenoidal Tracker at RHIC (STAR) is an essential effort to ensure efficient use of the limited computing resources. Trying to keep account of the usage of the Parallel Distributed Systems Facility (PDSF) computing facility and the number of files kept at the two different facilities, RHIC Computing Facility (RCF) and PDSF, becomes a very intensive job. The purpose of this research is to find a way to monitor the usage of the PDSF cluster and to keep account of the number of files each facility carries so that the datasets are the same. In analyzing the workload characteristics on PDSF (length of the batch jobs) over 3 years it was apparent that the most jobs were fairly long in the early years. As time progressed the jobs were shorter due to the fact that much of the activity was due to the analysis of the data. The software developed to count the files at RCF and PDSF has been an asset to researchers because now they have access to identifying the files that are needed at the facilities. This development prints the file monitoring results automatically to a web page making it easy to monitor the replication of files from RCF and PDSF.
To cite this abstract, use the following reference: http://meetings.aps.org/link/BAPS.2005.HAW.FR.89