SC13 Denver, CO

The International Conference for High Performance Computing, Networking, Storage and Analysis

Research Activities at Fermilab for Big Data Movement.

Authors: Parag Mhashilkar (Fermi National Accelerator Laboratory), Wenji Wu (Fermi National Accelerator Laboratory), Hyun Woo Kim (Fermi National Accelerator Laboratory), Gabriele Garzoglio (Fermi National Accelerator Laboratory), David Dykstra (Fermi National Accelerator Laboratory), Marko Slisz (Fermi National Accelerator Laboratory), Phil Demar (Fermi National Accelerator Laboratory)

Abstract: Fermilab is the US Tier-1 Center for the Large Hadron Collider’s (LHC) Compact Muon Solenoid (CMS) experiment and the central data center for several other large-scale research collaborations. These experiments generate a significant amount of data every year-- which need to be transferred, stored and analyzed. To deal with the scaling and wide-area distribution challenges of the data, the Laboratory is now connected to next generation of networking infrastructure, ESnet’s 100G-backbone. Current Fermilab R&D efforts include optimization of network I/O handling in multi-core systems, identification of gaps in the middleware, modification of middleware to improve application performance in 100G network environments, and end-to-end data and network aware storage selection mechanism. This poster describes the multi-prong approach in network-related R&D activities at Fermilab to facilitate big data processing and movement.

Poster: pdf
Two-page extended abstract: pdf

Poster Index