论文信息 - Running A Seismic Workflow Application on Distributed Resources

Running A Seismic Workflow Application on Distributed Resources

In this poster, we present an approach to running workflow applications on distributed resources, including systems without support for remote job submission. We show how this approach extends the benefits of scientific workflows, such as job and data management, to large-scale applications on open-science HPC resources such as Blue Waters, Stampede, and USC HPCC. We demonstrate this approach with SCEC CyberShake, a physicsbased seismic hazard application, to run over 470 million tasks via 32,000 jobs submitted to Blue Waters and Stampede.

[1] Jun Zhou,et al. Physics-based seismic hazard analysis on petascale heterogeneous supercomputers , 2013, 2013 SC - International Conference for High Performance Computing, Networking, Storage and Analysis (SC).

[2] Keith Beattie,et al. Metrics for heterogeneous scientific workflows: A case study of an earthquake science application , 2011, Int. J. High Perform. Comput. Appl..

[3] C. Kesselman,et al. CyberShake: A Physics-Based Seismic Hazard Model for Southern California , 2011 .

[4] Philip J. Maechling,et al. Enabling large-scale scientific workflows on petascale resources using MPI master/worker , 2012, XSEDE '12.