Distributed Computing Systems Group (DCSG)

The DCSG at the University of Minnesota is led by Professor Jon Weissman. The group performs applications-driven systems research in the area of Computational Grids. In particular, we are working on systems, middleware, and techniques to support high-end Grid applications and end-users. We work in "layer 3" of the Grid protocol stack, layer 1 being the physical resources, layer 2 the enabling Grid infrastructure or Grid OS, and layer 4 is the Grid application.


Focus of Research

The goal of our research is to advance the theory and practice of Grid Computing. In our view, Grids will not be successful until they are invisible and effortless to use. Towards this vision, we are working in several areas: scheduling and resource management (high-performance must be automatically delivered to Grid applications with little programmer effort), and Grid-based service infrastructures (using Grid technology to build high-level, easy-to-use, environments, that hide their underlying Grid implementation).

To achieve our goal, we do a mixture of system building, middleware design, and performance evaluation (both in simulation and in live experiments). We also produce running prototypes that serve as testbeds for our research and proof-of-concept demonstrations.

Projects

  • Community Services

    The Community Services project is constructing next-generation middleware and systems software for dynamic Grid services. A focus of this project is the definition of an adaptive Grid service and a system architecture to support a fully dynamic Grid service lifecyle. Dynamic Grid services are an important substrate to support collaboration. This work is funded by NSF and DOE.

  • Data Mining on the Grid

    This collaborative project (with University of Illinois at Chicago and University of Florida) is focused on exploiting Grid technology to enable mining of distributed datasets, e.g. distributed network intrusion analysis. Our work is to support efficient scheduling of data mining computations by considering the placement and possible replication of data and mining computations in a Grid to achieve user-level QoS. This work is funded by NSF.

  • Intelligent Storage

    This collaborative project between Computer Science and Electrical and Computer Engineering is investigating new models of storage to support the explosion of data and information at many scales: personal, enterprise, and beyond. Our work concerns how the concept of active storage can be realized, how "files" can be located by attributes or content, and how data provenance can be captured by intelligent storage. This work is funded by the Digital Technology Center and industry participants.

  • Faculty
    Students
    Alumni

    Current Support

  • NSF
  • DOE
  • DTC