Eric Pouyoul is a Senior System Engineer in ESnet's Advanced Network Technologies Group at Lawrence Berkeley National Laboratory (LBNL). His interests include all aspects of high performance big data movement, networking, hardware, software and distributed systems. He has been ESnet lead for designing Data Transfer Nodes (DTN) as defined in the Science DMZ architecture as well as ESnet's work in Software Defined Networking (OpenFlow). Mr Pouyoul joined ESnet in 2009 and his 25 years prior experience includes real-time operating system, supercomputing and distributed systems.
Nathan Hanford, Vishal Ahuja, Matthew Farrens, Dipak Ghosal, Mehmet Balman, Eric Pouyoul, Brian Tierney, “Improving network performance on multicore systems: Impact of core affinities on high throughput flows”, Future Generation Computer Systems, Vol 56., March 1, 2016,
- Download File: FGCS-Hanford.pdf (pdf: 1 MB)
M Kiran, E Pouyoul, A Mercian, B Tierney, C Guok, I Monga, “Enabling Intent to Configure Scientific Networks for High Performance Demands”, 3nd International Workshop on Innovating the Network for Data Intensive Science (INDIS) 2016, SC16., November 10, 2016,
- Download File: indis-final-2016.pdf (pdf: 745 KB)
Adrian Lara, Byrav Ramamurthy, Eric Pouyoul, Inder Monga, “WAN Virtualization and Dynamic End-to-End Bandwidth Provisioning Using SDN”, Optical Fiber Communication Conference 2015, March 22, 2015,
- Download File: OFC-2015-Th1A.2.pdf (pdf: 304 KB)
We evaluate a WAN-virtualization framework in terms of delay and scalability and demonstrate that adding a virtual layer between the physical topology and the end-user brings significant advantages and tolerable delays
Nathan Hanford, Vishal Ahuja, Matthew Farrens, Dipak Ghosal, Mehmet Balman, Eric Pouyoul, Brian Tierney., “Analysis of the effect of core affinity on high-throughput flows”, Proceedings of the Fourth International Workshop on Network-Aware Data Management (NDM '14), November 16, 2014,
- Download File: ndm2013-paper1Hanford2.pdf (pdf: 909 KB)
Henrique Rodriguez, Inder Monga, Abhinava Sadasivarao , Sharfuddin Sayed, Chin Guok, Eric Pouyoul, Chris Liou,Tajana Rosing, “Traffic Optimization in Multi-Layered WANs using SDN”, 22nd Annual Symposium on High-Performance Interconnects, Best Student Paper Award, August 27, 2014,
- Download File: hoti2014cam-1-1.pdf (pdf: 1.1 MB)
Wide area networks (WAN) forward traffic through a mix of packet and optical data planes, composed by a variety of devices from different vendors. Multiple forwarding technologies and encapsulation methods are used for each data plane (e.g. IP, MPLS, ATM, SONET, Wavelength Switching). Despite standards defined, the control planes of these devices are usually not interoperable, and different technologies are used to manage each forwarding segment independently (e.g. OpenFlow, TL-1, GMPLS). The result is lack of coordination between layers and inefficient resource usage. In this paper we discuss the design and implementation of a system that uses unmodified OpenFlow to optimize network utilization across layers, enabling practical bandwidth virtualization. We discuss strategies for scalable traffic monitoring and to minimize losses on route updates across layers. We explore two use cases that benefit from multi-layer bandwidth on demand provisioning. A prototype of the system was built open using a traditional circuit reservation application and an unmodified SDN controller, and its evaluation was per-formed on a multi-vendor testbed.
Nathan Hanford, Vishal Ahuja, Mehmet Balman, Matthew Farrens, Dipak Ghosal, Eric Pouyoul and Brian Tierney, “Characterizing the Impact of End-System Affinities On the End-to-End Performance of High-Speed Flows”, The 3rd International Workshop on Network-aware Data Management, in conjunction with SC'13, November 17, 2013,
- Download File: ndm2013-paper1Hanford.pdf (pdf: 909 KB)
Ezra Kissel, Martin Swany, Brian Tierney and Eric Pouyoul, “Efficient Wide Area Data Transfer Protocols for 100 Gbps Networks and Beyond”, The 3rd International Workshop on Network-aware Data Management, in conjunction with SC'13:, November 17, 2013,
- Download File: ndm2013paper3Kissel.pdf (pdf: 590 KB)
Yufei Ren, Tan Li, Dantong Yu, Shudong Jin, Thomas Robertazzi, Brian L. Tierney, Eric Pouyoul, “Protocols for Wide-Area Data-intensive Applications: Design and Performance Issues”, Proceedings of IEEE Supercomputing 2012, November 12, 2012,
- Download File: rftp-sc12-final.pdf (pdf: 280 KB)
Providing high-speed data transfer is vital to various data-intensive applications. While there have been remarkable technology advances to provide ultra-high-speed network band- width, existing protocols and applications may not be able to fully utilize the bare-metal bandwidth due to their inefficient design. We identify the same problem remains in the field of Remote Direct Memory Access (RDMA) networks. RDMA offloads TCP/IP protocols to hardware devices. However, its benefits have not been fully exploited due to the lack of efficient software and application protocols, in particular in wide-area networks. In this paper, we address the design choices to develop such protocols. We describe a protocol implemented as part of a communication middleware. The protocol has its flow control, connection management, and task synchronization. It maximizes the parallelism of RDMA operations. We demonstrate its performance benefit on various local and wide-area testbeds, including the DOE ANI testbed with RoCE links and InfiniBand links.
Inder Monga, Eric Pouyoul, Chin Guok, “Software Defined Networking for big-data science (paper)”, SuperComputing 2012, November 11, 2012,
- Download File: ESnet-SRS-SC12-paper-camera-ready.pdf (pdf: 1 MB)
University campuses, Supercomputer centers and R&E networks are challenged to architect, build and support IT infrastructure to deal effectively with the data deluge facing most science disciplines. Hybrid network architecture, multi-domain bandwidth reservations, performance monitoring and GLIF Open Lightpath Exchanges (GOLE) are examples of network architectures that have been proposed, championed and implemented successfully to meet the needs of science. Most recently, Science DMZ, a campus design pattern that bypasses traditional performance hotspots in typical campus network implementation, has been gaining momentum. In this paper and corresponding demonstration, we build upon the SC11 SCinet Research Sandbox demonstrator with Software-Defined networking to explore new architectural approaches. A virtual switch network abstraction is explored, that when combined with software-defined networking concepts provides the science users a simple, adaptable network framework to meet their upcoming application requirements.
Brian Tierney, Ezra Kissel, Martin Swany, Eric Pouyoul, “Efficient Data Transfer Protocols for Big Data”, Proceedings of the 8th International Conference on eScience, IEEE, October 9, 2012,
- Download File: eScience-networks.pdf (pdf: 289 KB)
Abstract—Data set sizes are growing exponentially, so it is important to use data movement protocols that are the most efficient available. Most data movement tools today rely on TCP over sockets, which limits flows to around 20Gbps on today’s hardware. RDMA over Converged Ethernet (RoCE) is a promising new technology for high-performance network data movement with minimal CPU impact over circuit-based infrastructures. We compare the performance of TCP, UDP, UDT, and RoCE over high latency 10Gbps and 40Gbps network paths, and show that RoCE-based data transfers can fill a 40Gbps path using much less CPU than other protocols. We also show that the Linux zero-copy system calls can improve TCP performance considerably, especially on current Intel “Sandy Bridge”-based PCI Express 3.0 (Gen3) hosts.
Mehmet Balman, Eric Pouyoul, Yushu Yao, E. Wes Bethel Burlen Loring, Prabhat, John Shalf, Alex Sim, Brian L. Tierney, “Experiences with 100Gbps Network Applications”, The Fifth International Workshop on Data Intensive Distributed Computing (DIDC 2012), June 20, 2012,
- Download File: didc.LBNL.pdf (pdf: 8.8 MB)
100Gbps networking has finally arrived, and many research and educational institutions have begun to deploy 100Gbps routers and services. ESnet and Internet2 worked together to make 100Gbps networks available to researchers at the Supercomputing 2011 conference in Seattle Washington. In this paper, we describe two of the first applications to take advantage of this network. We demonstrate a visualization application that enables remotely located scientists to gain insights from large datasets. We also demonstrate climate data movement and analysis over the 100Gbps network. We describe a number of application design issues and host tuning strategies necessary for enabling applications to scale to 100Gbps rates.
M. Boddie, T. Entel, C. Guok, A. Lake, J. Plante, E. Pouyoul, B. H. Ramaprasad, B. Tierney, J. Triay, V. M. Vokkarane, On Extending ESnet's OSCARS with a Multi-Domain Anycast Service, IEEE ONDM 2012, December 2012,
I. Monga, E. Pouyoul, C. Guok, Software-Define Networking for Big-Data Science – Arthictectural Models from Campus to the WAN, SC12: IEEE HPC, November 2012,
Inder Monga, Eric Pouyoul, Chin Guok, Eli Dart, SDN for Science Networks, Summer Joint Techs 2012, July 17, 2012,
- Download File: Science-SDN-Monga-JT-07172012.pdf (pdf: 2.6 MB)
Eric Pouyoul, Inder Monga, Brian Tierney, Dynamic creation of end-to-end virtual networks for science and cloud computing leveraging OpenFlow/Software Defined Networking, TERENA 2012, May 2012,
Eric Pouyoul, Brian Tierney, Achieving 98Gbps of Cross-country TCP traffic using 2.5 hosts, 10 10G NICs, and 10 TCP streams, Winder 2012 Joint Techs, January 25, 2012,
Joe Breen, Eli Dart, Eric Pouyoul, Brian Tierney, Achieving a Science "DMZ", Winter 2012 Joint Techs, Full day tutorial, January 22, 2012,
There are several aspects to building successful infrastructure to support data intensive science. The Science DMZ Model incorporates three key components into a cohesive whole: a high-performance network architecture designed for ease of use; well-configured systems for data transfer; and measurement hosts to provide visibility and rapid fault isolation. This tutorial will cover aspects of network architecture and network device configuration, the design and configuration of a Data Transfer Node, and the deployment of perfSONAR in the Science DMZ. Aspects of current deployments will also be discussed.