CHAPEL HILL, Dec. 9, 2011–RENCI at the University of North Carolina, Chapel Hill and Duke University in partnership with IBM will lead a new project to build a nationwide test bed for networking and networked cloud computing.

The project is part of NSF’s Global Environment for Network Innovation (GENI) initiative, which enables researchers to explore networks of the future.

The National Science Foundation (NSF) awarded just over $2 million to the three-year ExoGENI project, led by Ilia Baldine, director of RENCI’s networking research group and Jeff Chase, a Duke University computer science professor.

The project will deploy and operate 13 ExoGENI sites at research universities and labs across the U.S.  The project will use software based on the Open Resource Control Architecture (ORCA) to control the networked cloud infrastructure. The project team developed the ORCA platform in earlier NSF-funded research and extended it for use in the GENI initiative.

Each ExoGENI site will receive a rack of equipment with multiple IBM x3650 servers featuring dual socket Intel Westmere and Sandy Bridge CPUs, each with 6 to 8 cores and 48 gigabytes of RAM. The sites will connect to a variety of advanced research networks offering dynamic circuit capabilities and programmable control.

ExoGENI sites in North Carolina’s Research Triangle area will connect to each other using RENCI’s Breakable Experimental Network (BEN), a networking test bed that links RENCI, Duke, NC State University and UNC-Chapel Hill. The sites will link to national research networks such as National Lambda Rail (NLR), Internet2 and the Department of Energy’s Energy Sciences Network (ESnet).

The ORCA control software will enable experimenters to construct on demand private virtual networks spanning these research networks and ExoGENI sites. ExoGENI racks will use OpenFlow-enabled switches to link to OpenFlow-enabled campus infrastructures and national networks. OpenFlow technology separates a network switch’s packet forwarding, or data, path from its high-level routing decisions, or control path, thereby allowing researchers to easily deploy innovative routing and switching protocols.

When all the hardware is operational and all the sites are connected, ExoGENI will operate as a networked cloud infrastructure—a virtual laboratory for networking and computer science experiments that will help researchers advance the development of a faster, smarter and more reliable Internet, said Baldine.

“Future computer science and applied research must bring together computation, storage and network capabilities on a global scale to address emerging complex problems related to network science, large-scale distributed computations, large dataset mobility and future network architectures,” said Baldine. “With ExoGENI researchers will gain a global, elastic reconfigurable platform to conduct such research.”

ExoGENI will support a variety of experiments that will create network topologies consisting of nodes allocated from ExoGENI sites tied together with network connections that will be provisioned based on the bandwidth needed for the experiment. ExoGENI will support using custom kernels to experiment with different network protocols. To support research into high-speed protocols, some ExoGENI sites will be capable of transferring data at 10 gigabits per second (Gbp/s) and in the future at 40 Gbp/s and 100 Gbp/s.

Using ExoGENI, researchers will be able to allocate private networks spanning the continental U.S., allocate computing clusters and storage for use by scientists who collect and analyze data, and tie these experimental resources to production networks, devices or instruments. Because ExoGENI will interact with other networking and compute resources assembled through the GENI initiative, researchers will be able to create more powerful assemblies, or slices, of linked resources that include wireless, mobile and sensor networks.

In addition, ExoGENI will serve as a test environment for a global federated cloud infrastructure that can reconfigure collections of linked computational resources as needed and bring together diverse resources from multiple cloud providers. Such an environment could someday replace typical institutional computational resources, which today exist in a single lab or data center.

ExoGENI sites will be deployed over the course of the next 18-24 months and the facility will begin operation as soon as the first sites are deployed. The first four sites will be operational before the end of September 2012.

Reagan Moore, Ph.D., the principal investigator for the consortium, director of the DICE Center, SILS professor and domain scientist for data management at RENCI.

CHAPEL HILL, NC, Sept. 28, 2011–The National Science Foundation has funded the University of North Carolina at Chapel Hill to lead a multi-institutional team that will build and deploy a prototype national data management infrastructure that addresses some of the key data challenges facing scientific researchers in the digital age.

The infrastructure will support collaborative multidisciplinary research through shared collections, data publication within digital libraries and reference collections within persistent archives.

The NSF awarded nearly $8 million over five years to the DataNet Federation Consortium (DFC), a group that spans seven universities. The DFC will address the data management needs of six science and engineering disciplines: oceanography, hydrology, engineering design, plant biology, cognitive science, and social science. About half the award will support research and development at UNC-Chapel Hill.

The Data Intensive Cyber Environments (DICE) research group in UNC’s School of Information and Library Science (SILS) leads the consortium and RENCI (the Renaissance Computing Institute at UNC-Chapel Hill)  is responsible for  federating the consortium’s diverse data repositories to enable cross-disciplinary research. Federating data involves building a common name space for identifying files, providing a context for file meaning and relevance, providing a common access interface, and developing management policies across the distributed collection.

The DFC will use iRODS, the integrated Rule Oriented Data System, to implement a policy-based data management infrastructure. iRODS, developed by UNC’s DICE Center and DICE researchers at the University of California at San Diego, enforces policies as computer actionable rules to organize distributed data into sharable collections.  Procedures to automate data management functions are cast as computer executable workflows.  Policies control data access, sharing and archiving. Research groups worldwide, including the NASA Center for Climate Simulations, the National Optical Astronomy Observatory, the Australian Research Collaboration Service, and the Texas Digital Libraries, use iRODS technology to manage their research data grids, implement digital libraries, and build persistent archives.

“Excelling in the digital age requires that scientific disciplines and government agencies have the ability to manage the enormous amount of data that are generated each day,” said UNC-Chapel Hill Vice Chancellor for Research Barbara Entwisle. “Scientists can only solve the important problems of our times if they can easily access, share, analyze, and preserve data for future researchers and students. This award is important beyond its dollar amount because it establishes Carolina as the leader in the worldwide research community in taming the data deluge and as the data federation hub for collaborative research. It’s a role that is essential for future discoveries and innovations.”

Experts in the DICE group and at RENCI will work with six NSF-supported national consortia to federate their distributed data repositories and create policies for retention, distribution, access and validation of critical data properties. Those communities are:

• The Ocean Observatories Initiative (OOI), an NSF-funded program led by the University of California at San Diego and the Scripps Institution of Oceanography in San Diego. The OOI researchers use data from environmental sensors to study the physical, chemical, geological and biological variables in the ocean and seafloor.
• The Consortium of Universities for Advancement of Hydrologic Science, Inc.  (CUAHSI) an organization led by the University of South Carolina. CUAHSI includes more than 130 partner organizations, including UNC’s Institute for the Environment, working to advance water science.
• CIBER-U, the Cyber-Infrastructure-Based Engineering Repositories for Undergraduates, an initiative led by Drexel University, which uses digital design repositories to enhance engineering instruction and learning.
• The iPlant Collaborative, a community of researchers and students led by the University of Arizona that is developing an integrated cyberinfrastructure to advance studies of plant biology.
• The Odum Institute for Research in Social Science, an interdisciplinary institute at UNC-Chapel Hill that focuses on teaching and research in the social sciences.
• The Temporal Dynamics of Learning Center (TDLC), an NSF Science of Learning Center based at the University of California at San Diego that studies the role of time and timing in learning in order to improve educational practices.

Arizona State University researchers will participate in the DFC by collaborating on policy-based data management systems and Duke University researchers will collaborate on education and outreach initiatives to broaden the impact of the DFC.

“The data we will work with includes observational data from sensors, experimental and simulation data, engineering designs and both structured and unstructured data,” said Reagan Moore, Ph.D., the principal investigator for the consortium, director of the DICE Center, SILS professor and domain scientist for data management at RENCI.  “The infrastructure we develop will address all stages in the community-based data collection lifecycle, from initial collection formation for a single project, to shared collections across institutions, to formation of data processing pipelines, to publication and long term preservation. We see this as the first step to building a data infrastructure that will accommodate collaborative research, new educational approaches and innovative problem solving in academic institutions, in federal agencies and across national boundaries.”

During the first 18 months of the grant, the consortium will focus on federating the data management cyberinfrastructure for the OOI, CUASHI and CIBER-U. The work will include identifying federation requirements, integrating existing data management systems, deploying a federation hub, and developing policies and procedures for data sharing so that the data collections of these research communities can become the foundation of a national data cyberinfrastructure.

For more information:

iRODS website

DICE Center website

SILS website

CAMBRIDGE, MASS.—New funding was announced this week for the RENCI and Duke University project to develop a dark fiber network where researchers can create and manage network experiments.

RENCI and Duke will receive Spiral 2 funding of $150,000 from the National Science Foundation’s  Global Environment for Network Innovations (GENI) project. That money and a previous award will be used to continue developing the Breakable Experimental Network (BEN), a project of RENCI, Duke and Infinera, and to further refine the Open Resource Control Architecture (ORCA).  Originally developed by Duke computer scientist Jeff Chase, ORCA serves as the resource control system for BEN.

According to Ilia Baldine, manager of RENCI’s Network Research and Innovation Group, the new and previous NSF funding will be used to further develop the ORCA software framework so that it is easy to use and provides the stability, flexibility and assortment of tools needed to open up BEN to the network research community within the next year.

“Our goal will be to make ORCA a production system for creating and managing networking experiments on BEN by the summer of 2010,” Baldine said. “As part of this effort, we will enhance the architecture, develop user tools and new capabilities that will allow users to select appropriate resources and measurement capabilities within the network to include in their experiments.”

The GENI Spiral 2 awards were announced Oct. 12 by BBN Technolgies, the firm that manages the GENI initiative for the NSF.

For more on BEN, ORCA and GENI, see the RENCI feature story, Break this network.