On-Line Application Performance Monitoring of Blood Flow Simulation in Computational Grid Architectures
18th IEEE Symposium on Computer-Based Medical Systems (CBMS'05) (2005)
June 23, 2005 to June 24, 2005
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/CBMS.2005.79
Alfredo Tirado-Ramos , University of Amsterdam
Derek Groen , University of Amsterdam
Peter Sloot , University of Amsterdam
We report on our findings after running a number of on-line performance monitoring experiments with a biomedical parallel application to investigate levels of performance at hardware resources distributed across a computational Grid network. We use on-line application monitoring for improved computational resource selection and application optimization. We used a number of user-defined performance metrics within the European CrossGrid Project?s G-PM tool together with a blood flow simulation application based on the lattice Boltzmann method for fluid dynamics. We found that the performance results observed during our on-line experiments give us a more accurate view of computational resource status than the regular resource information provided by standard information services to resource brokers, and that on-line monitoring has good potential for optimizing our biomedical application for more efficient runs.
P. Sloot, A. Tirado-Ramos and D. Groen, "On-Line Application Performance Monitoring of Blood Flow Simulation in Computational Grid Architectures," 18th IEEE Symposium on Computer-Based Medical Systems (CBMS'05)(CBMS), Dublin, Ireland, 2005, pp. 511-516.