CI Collaborations

Current Collaborations

• Western New York Grid (WNY-Grid). Funding for this project is provided by an NSF CRI Grant, entitled "CRI: A Western New York Computational and Data Science Grid." The PI of this project is Dr. Miller, with co-PIs Dr. Charles M. Weeks (Hauptman-Woodward Medical Research Institute/HWI), Dr. Mary McCourt (Niagara University/NU), Prof. Homma Farian (SUNY-Geneseo), and Dr. Mark L. Green (SUNY-Buffalo).

  This project focuses on the design, analysis, and implementation of a heterogeneous integrated computational and data grid in Western New York (WNY). The Cyberinfrastructure Laboratory has reached out to HWI, NU, and Geneseo to provide education to scientists, system administrators, and students in the use of clusters and the establishment of grids for performing leading-edge science and engineering. These facilities have been incorporated into the classroom, have led to a spin-off company, have led to enabling new applications for the grid, and have increased the interactions between the Buffalo-based group and Open Science Grid.

  Geneseo has installed a Sun/AMD cluster with 128 compute cores. HWI has installed an Apple cluster with 100 G5 compute processors. NU installed a Dell cluster with 128 Intel Xeon compute processors. Buffalo is waiting on its Provost to allocate appropriate space and required matching funds for the project in order to acquire its system, house the equipment, and house the personnel in the PI's research group. The machines installed at Geneseo, HWI, and NU are the largest machines that have been installed to date at those institutions and has helped to raise the profile of the institutions in terms of computational science and engineering (research and education).

  The institutions involved with this project have benefited from the CI Labs advances, including the development and deployment of an efficient lightweight grid monitoring system, automated procedures for rolling portions of a cluster into and out of a grid, a predictive scheduler based on user and application profiling, an efficient backfill scheduling mechanism, a generic grid portal, and a methodology for porting applications to this portal through grid-enabling application templates.

• New York State Grid (NYS Grid). Due to the success of the WNY Grid, it has been expanded to include numerous institutions throughout New York State (NYS). The initial New York State Grid (NYS Grid) included SUNY-Albany, SUNY-Binghamton, SUNY-Buffalo, SUNY-Geneseo, Canisius College, Columbia University, the Hauptman-Woodward Medical Research Institute, and Niagara University. In fact, with NYS Grid serving as the basis for the NYSGrid.org grid, further expansion has occurred to include Marist College, the University of Rochester, New York University, Cornell University, the Rochester Institute of Technologoy, and Syracuse University. This work is funded by NSF Grants CRI CNS-0454114 and ITR ACI-0204918 (Miller is PI of both grants).

• NYSGrid.org. During the Summer of 2006, Cornell University organized and hosted the "NYS Workshop on Data-Driven Science and Cyberinfrastructure." Due to the significant interest in this workshop from institutions around New York State, a sequel was scheduled in September, which was hosted in Troy, New York, by RPI. At this meeting, a Governing Board of a New York State Cyberinfrastructure Initiative was created and it was requested that the NYS Grid serve as the basis for the underlying grid.

  Subsequent to the September meeting in Troy, it was determined that CI Lab's NYS Grid should serve as the foundation of the NYS Cyberinfrastructure Initiative.

• Open Science Grid. The Cyberinfrastructure Lab and CCR have worked closely with the Open Science Grid for a number of years on a wide range of projects. We appreciate the guidance and assistance they have provided over the years.

Previous Collaborations

• Grid2003 Grid Monitoring. The CI Lab's ACDC-Grid Monitor was deployed at SC03 on Grid2003.
• Grid2003 Dynamic Resource Allocations. The CI Lab's dynamic scheduling system for moving portions of a cluster into and out of a grid in an automated fashion was deployed at SC03 on Grid2003.
• HP supports the creation of post-doctoral research position. HP provided funds to support a post-doctoral student in who focused on critical aspects of grid computing. This student also will served as a liaison between Dr. Miller's research team and H-P in trems of the WNY Grid, porting Shake-and-Bake to the grid, and implementing various tools that were being developed in the CI Lab.

Disclaimer: Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.