Skip to main content
Data Science and Learning

OSG: Open Science Grid

OSG
Consortium-built, production-scale, shared cyber infrastructure

The OSG (Open Science Grid), which initially began with joint efforts from the Grid Physics Network (GriPhyN), Particle Physics Data Grid, and the International Virtual Data Grid Laboratory projects, is a production-scale, shared cyberinfrastructure built through an application-driven, multi-disciplinary science consortium. The middleware infrastructure is built from the Virtual Data Toolkit which packages and integrates software from Globus, Condor, and a suite of virtual data services developed by the GriPhyN project. The virtual organization stakeholders include members from the high energy physics, gravity wave physics, genomics, brain imaging research, and computer science communities, among others.

The University of Chicago/Computation Institute group has played leadership roles in several activities of the OSG, including education and outreach, consortium governance, and the OSG Integration Testbed (ITB). The ITB provides a platform for pre-production service validation, service integration, and application validations over the OSG software stack. The contributions, coordinated by the University of Chicago, come from a dozen or so external sites and virtual organizations which together address a wide range of middleware-site fabric integration issues in a hetereogenous environment.

To date, the group has provided a framework for OSG service release specifications, validation procedures, introduction of new services into the OSG computing environment, and site functional testing through deployment of a development and integration cluster comprised of computer, storage and OSG-edge service nodes.