In 2002 Pegasus was shown to produce any virtual data products present in the LIGO pulsar search.  The figure below shows the visualization of the search shown at SC 2002. This configuration of Pegasus is based on AI-planning technologies. To find out more about planning process, click here.

Pegasus uses the Metadata Catalog Service (MCS), developed at ISI, to perform the mapping between application-specific attributes and logical file names of existing data products. AI-based planning technologies are used to construct both the abstract and concrete workflows.  Pegasus also contains a Virtual Data Language generator that can populate the Chimera catalog with newly constructed derivations.  The LIGO pulsar search is performed using the Pegasus system, LDAS and compute and storage resources at Caltech, University of Southern California, University of Wisconsin Milwaukee, University of Florida, and NCSA.

To see some of the applications using Pegasus and Chimera, click here.

Pegasus has been used in demonstration of earthquake analysis: As part of the demonstrations at the annual SCEC (Southern California Earthquake Center) meeting, GriPhyN and iVDGL-developed software suite the Virtual Data Toolkit (http://www.cs.wisc.edu/vdt/) was used to support SCEC analysis.  Three different Seismic Hazard-related Calculations (Pathway 1, Pathway 2, and Pathway 4) were created and executed using the SCEC portal, in a secure manner. The system produced workflow templates (that identify the necessary computations and their sequence), populated them with data producing the abstract workflows, and then handed those to the GriPhyN Virtual Data System (VDS). Pegasus, one of the VDS components, mapped the workflows consisting of both MPI-based and single processor jobs onto the SCEC and USC grid. DAGMan and Condor-G executed the workflow components. One impressive aspect of the system was that there was a consistent workflow system under different seismic hazard programs. Previously, each SCEC pathway ran in an independent and dissimilar manner. The following software was used in the demonstrations: NMI, VDT (including VDS), and MCS (the Metadata Catalog Service.)

This work represents the collaboration of scientists and developers from SCEC at USC and USC/ISI.
Slides about this work can be found here.