SuperComputing 2015
High Speed Scientific Data Transfers using Software Defined Networking

Austin, Texas - During the SuperComputing 2015 (SC15) conference November 16-20, an International team of high energy physicists, computer scientists, and network engineers led by the California Institute of Technology (Caltech), SPRACE Sao Paulo and University of Michigan, together with teams from FIU, Vanderbilt and support from vendors including Dell, Mangstor, Mellanox, QLogic, SGI and Spirent worked together to demosntrate large data flow transfers across an highly intelligent SDN network. This work is being supported by SCinet Network Research Exhibition (NRE) and presented in the ESnet organized INDIS and SDN workshops. The paper is also submitted to IEEE ACM for it's publication. Click here to read the paper.

SDN Demonstrations

SDN demonstration revolves around an OpenFlow ring connecting 7 different booths and the WAN connections. Some of the WAN connections are built using the NSI and then stitched using the custom SDN application. This helps in creating an intelligent design where large flows traverse across various paths based on either the shortest or fastest routes availability. Remote NSI paths are shows in the picture below. An interesting aspect is that all the remote switches will be controlled by a single controller in the Caltech booth at the show floor.

The group has just released first working library to control the SDN controller OpenDaylight (Lithium). The library is written in Python and is available at pypi and github:

1.5 Tbit/sec Transfer


High Speed DTN demonstrations at 100GE

With the recent scientific innovations and continuous outcomes in the major fields like LHC, Genomis, Astronomy, Oceanlogy, the data volumes are increasing and require efficient protocols and mechanisms to transport across the globe. The group will present such technologies and demonstrate data transfers at 100GE from disk to disk.

A separate demonstration will show the future challenges on the systems design for the 400GE and beyond between a pair of systems.

Named Data Networking (NDN)

NDN project aims to develop a new Internet architecture that can capitalize on strengths — and address weaknesses — of the Internet’s current host-based, point-to-point communication architecture in order to naturally accommodate emerging patterns of communication. By naming data instead of their locations, NDN transforms data into a first-class entity. At SC15, a separate NDN node will be announcing LHC data objects while the visualization software from the NDN group will show case the data flows to the shortest routes.

This work is made possible by the strong support of the US Department of Energy Office of Science and the National Science Foundation.

                  Science with a Mission