The DYNES Instrument: A Description and Overview

Abstract

<!--HTML-->Scientific innovation continues to increase requirements for the \ncomputing and networking infrastructures of the world. Collaborative \npartners, instrumentation, storage, and processing facilities are often \ngeographically and topologically separated, as is the case with LHC \nvirtual organizations. These separations challenge the technology used \nto interconnect available resources, often delivered by Research and \nEducation (R&E) networking providers, and leads to complications in the \noverall process of end-to-end data management.\n\nCapacity and traffic management are key concerns of R&E network \noperators; a delicate balance is required to serve both long-lived, high \ncapacity network flows, as well as more traditional end-user activities. \nThe advent of dynamic circuit services, a technology that enables the \ncreation of variable duration, guaranteed bandwidth networking channels, \nallows for the efficient use of common network infrastructures. These \ngains are seen particularly in locations where overall capacity is \nscarce compared to the (sustained peak) needs of user communities. \nRelated efforts, including those of the LHCOPN operations group and the \nemerging LHCONE project, may take advantage of available resources by \ndesignating specific network activities as a "high priority", allowing \nreservation of dedicated bandwidth or optimizing for deadline scheduling \nand predicable delivery patterns.\n\nThis paper presents the DYNES instrument, an NSF funded \ncyberinfrastructure project designed to facilitate end-to-end dynamic \ncircuit services. This combination of hardware and software innovation \nis being deployed across R&E networks in the United States at selected end-sites \nlocated on University Campuses. DYNES is peering with international \nefforts in other countries using similar solutions, and is increasing \nthe reach of this emerging technology. This global data movement \nsolution could be integrated into computing paradigms such as cloud and \ngrid computing platforms, and through the use of APIs can be integrated \ninto existing data movement software.