The OSG plans for year 9

Wisconsin_Institutes_for_DiscoveryIt is a tradition by now. In May, the OSG executive team and area coordinators meet in Madison, WI, to take stock of the program to date and, most importantly, discuss the vision and plans for the upcoming year.

The OSG project has started its 9th year of funding. All areas are aligning to continue and extend the support for our scientific stakeholders with an eye toward the next 5 year grant cycle. Coordination across areas and with partner organizations, such as WLCG and XD, continues to be essential to achieve our goals.


Wisconsin Institutes for Discovery — Photo courtesy of Wikimedia Commons

Maintaining access to the dedicated resources of established VOs such as LHC, Intensity Frontier, etc. is central to the program, as is remaining a valuable service provider to XSEDE. For opportunistic VOs, the goal for year 9 is to maintain and possibly expand the pool of opportunistic resources to 150 million hours per year. The strategy will include demonstrating access to Supercomputers, such as Comet at SDSC and Cori at NERSC. Eventually, all these resources will be available through both the OSG Direct and OSG Connect interfaces to enable science for VO’s and campus users. Improvements to HTCondor will enable and simplify the provisioning of compute resources at higher scales.

Increasingly, reliable and scalable data access is fundamental to the use of the computing resources for many of the science communities on OSG. The project is focusing on strengthening data management-related technologies, such as StashCache for input data and OASIS for software distribution. In addition, the project wants to lower the barriers for resource access through improved identity management processes that rely on broadly-available systems, such as CILogon, or that do not require globally-trusted credentials, such as x509 certificates, whenever possible.

As data and job management depend on reliable networks, the plan also includes transitioning the OSG networking dashboard into production operations to collect, display, and analyze the inter-institution connectivity data for OSG and WLCG.

At the same time, the project will continue its work to support upcoming platforms, such as EL7, streamlining internal processes for software and operations management, and providing operational support for existing and new services at the agreed service levels or beyond.

All in all, OSG has a challenging program ahead to realize a great vision with our scientific stakeholders.

Gabriele Garzoglio