Spectrum of Support for Data Movement and Analysis in Big Data Science Network Management and Control E-Center & ESCPS Network Management and Control E-Center.

Slides:



Advertisements
Similar presentations
Network Resource Broker for IPTV in Cloud Computing Lei Liang, Dan He University of Surrey, UK OGF 27, G2C Workshop 15 Oct 2009 Banff,
Advertisements

Big Data over a 100G Network at Fermilab Gabriele Garzoglio Grid and Cloud Services Department Computing Sector, Fermilab CHEP 2013 – Oct 15, 2013.
Experience and proposal for 100 GE R&D at Fermilab Interactomes – May 22, 2012 Gabriele Garzoglio Grid and Cloud Computing Department Computing Sector,
High Throughput Data Program at Fermilab R&D Parag Mhashilkar Grid and Cloud Computing Department Computing Sector, Fermilab Network Planning for ESnet/Internet2/OSG.
Database Architectures and the Web
Current Testbed : 100 GE 2 sites (NERSC, ANL) with 3 nodes each. Each node with 4 x 10 GE NICs Measure various overheads from protocols and file sizes.
 Contributing >30% of throughput to ATLAS and CMS in Worldwide LHC Computing Grid  Reliant on production and advanced networking from ESNET, LHCNET and.
Towards a Virtual European Supercomputing Infrastructure Vision & issues Sanzio Bassini
Kathy Benninger, Pittsburgh Supercomputing Center Workshop on the Development of a Next-Generation Cyberinfrastructure 1-Oct-2014 NSF Collaborative Research:
Data Grids Darshan R. Kapadia Gregor von Laszewski
DWDM-RAM: DARPA-Sponsored Research for Data Intensive Service-on-Demand Advanced Optical Networks DWDM RAM DWDM RAM BUSINESS WITHOUT BOUNDARIES.
Office of Science U.S. Department of Energy Grids and Portals at NERSC Presented by Steve Chan.
Milos Kobliha Alejandro Cimadevilla Luis de Alba Parallel Computing Seminar GROUP 12.
International Workshop APAN 24, Current State of Grid Computing Researches and Applications in Vietnam Nguyen Thanh Thuy 1, Nguyen Kim Khanh 1,
Simo Niskala Teemu Pasanen
23 September 2004 Evaluating Adaptive Middleware Load Balancing Strategies for Middleware Systems Department of Electrical Engineering & Computer Science.
GridFTP Guy Warner, NeSC Training.
GlobusWorld 2012: Experience with EXPERIENCE WITH GLOBUS ONLINE AT FERMILAB Gabriele Garzoglio Computing Sector Fermi National Accelerator.
Big Data: Movement, Crunching, and Sharing Guy Almes, Academy for Advanced Telecommunications 13 February 2015.
Fermi National Accelerator Laboratory 3 Fermi National Accelerator Laboratory Mission Advances the understanding of the fundamental nature of matter.
ARGONNE  CHICAGO Ian Foster Discussion Points l Maintaining the right balance between research and development l Maintaining focus vs. accepting broader.
An Integrated Instrumentation Architecture for NGI Applications Ian Foster, Darcy Quesnel, Steven Tuecke Argonne National Laboratory The University of.
100G R&D at Fermilab Gabriele Garzoglio (for the High Throughput Data Program team) Grid and Cloud Computing Department Computing Sector, Fermilab Overview.
100G R&D at Fermilab Gabriele Garzoglio Grid and Cloud Computing Department Computing Sector, Fermilab Overview Fermilab Network R&D 100G Infrastructure.
100G R&D at Fermilab Gabriele Garzoglio Grid and Cloud Computing Department Computing Sector, Fermilab Overview Fermilab Network R&D 100G Infrastructure.
GlobusWorld 2012: Experience with EXPERIENCE WITH GLOBUS ONLINE AT FERMILAB Gabriele Garzoglio Computing Sector Fermi National Accelerator.
File and Object Replication in Data Grids Chin-Yi Tsai.
Current Testbed : 100 GE 2 sites (NERSC, ANL) with 3 nodes each. Each node with 4 x 10 GE NICs Measure various overheads from protocols and file sizes.
Development Timelines Ken Kennedy Andrew Chien Keith Cooper Ian Foster John Mellor-Curmmey Dan Reed.
Data Transfers in the Grid: Workload Analysis of Globus GridFTP Nicolas Kourtellis, Lydia Prieto, Gustavo Zarrate, Adriana Iamnitchi University of South.
DataTAG Research and Technological Development for a Transatlantic Grid Abstract Several major international Grid development projects are underway at.
Communicating Security Assertions over the GridFTP Control Channel Rajkumar Kettimuthu 1,2, Liu Wantao 3,4, Frank Siebenlist 1,2 and Ian Foster 1,2,3 1.
O AK R IDGE N ATIONAL L ABORATORY U.S. D EPARTMENT OF E NERGY Facilities and How They Are Used ORNL/Probe Randy Burris Dan Million – facility administrator.
GVis: Grid-enabled Interactive Visualization State Key Laboratory. of CAD&CG Zhejiang University, Hangzhou
The Earth System Grid (ESG) Computer Science and Technologies DOE SciDAC ESG Project Review Argonne National Laboratory, Illinois May 8-9, 2003.
LEGS: A WSRF Service to Estimate Latency between Arbitrary Hosts on the Internet R.Vijayprasanth 1, R. Kavithaa 2,3 and Raj Kettimuthu 2,3 1 Coimbatore.
Practical Distributed Authorization for GARA Andy Adamson and Olga Kornievskaia Center for Information Technology Integration University of Michigan, USA.
GridFTP GUI: An Easy and Efficient Way to Transfer Data in Grid
May 2004NTUA1 National Technical University of Athens EGEE Project 3 rd Parties Kick off Meeting, Athens, May 27-28, 2004 Dr. Costis Christogiannis Telecommunications.
GridFTP Richard Hopkins
Introduction to Grids By: Fetahi Z. Wuhib [CSD2004-Team19]
6/23/2005 R. GARDNER OSG Baseline Services 1 OSG Baseline Services In my talk I’d like to discuss two questions:  What capabilities are we aiming for.
1 Grid Activity Summary » Grid Testbed » CFD Application » Virtualization » Information Grid » Grid CA.
Globus online Software-as-a-Service for Research Data Management Steve Tuecke Deputy Director, Computation Institute University of Chicago & Argonne National.
BNL Service Challenge 3 Status Report Xin Zhao, Zhenping Liu, Wensheng Deng, Razvan Popescu, Dantong Yu and Bruce Gibbard USATLAS Computing Facility Brookhaven.
Securing the Grid & other Middleware Challenges Ian Foster Mathematics and Computer Science Division Argonne National Laboratory and Department of Computer.
DOE UltraScience Net The Need –DOE large-scale science applications on supercomputers and experimental facilities require high-performance networking Petabyte.
TeraPaths: A QoS Enabled Collaborative Data Sharing Infrastructure for Petascale Computing Research The TeraPaths Project Team Usatlas Tier 2 workshop.
Super Computing 2000 DOE SCIENCE ON THE GRID Storage Resource Management For the Earth Science Grid Scientific Data Management Research Group NERSC, LBNL.
EGI-Engage Data Services and Solutions Part 1: Data in the Grid Vincenzo Spinoso EGI.eu/INFN Data Services.
High Throughput Data Program (HTDP) at FNAL Mission: investigate the impact of and provide solutions for the scientific computing challenges in Big Data.
Parag Mhashilkar Computing Division, Fermi National Accelerator Laboratory.
1 (Brief) Introductory Remarks On Behalf of the U.S. Department of Energy ESnet Site Coordinating Committee (ESCC) W.Scott Bradley ESCC Chairman
GridFTP Guy Warner, NeSC Training Team.
Sampling and Analysis Tools for E-Center for Multi-domain Internet Performance Measurement Prasad Calyam, Ph.D. Winter.
CRISP WP18, High-speed data recording Krzysztof Wrona, European XFEL PSI, 18 March 2013.
BDTS and Its Evaluation on IGTMD link C. Chen, S. Soudan, M. Pasin, B. Chen, D. Divakaran, P. Primet CC-IN2P3, LIP ENS-Lyon
1 Network related topics Bartosz Belter, Wojbor Bogacki, Marcin Garstka, Maciej Głowiak, Radosław Krzywania, Roman Łapacz FABRIC meeting Poznań, 25 September.
100G R&D for Big Data at Fermilab Gabriele Garzoglio Grid and Cloud Computing Department Computing Sector, Fermilab ISGC – March 22, 2013 Overview Fermilab.
Run - II Networks Run-II Computing Review 9/13/04 Phil DeMar Networks Section Head.
100GE Upgrades at FNAL Phil DeMar; Andrey Bobyshev CHEP 2015 April 14, 2015.
100G R&D at Fermilab Gabriele Garzoglio (for the High Throughput Data Program team) Grid and Cloud Computing Department Computing Sector, Fermilab Overview.
Big Data over a 100G Network at Fermilab Gabriele Garzoglio Grid and Cloud Services Department Computing Sector, Fermilab CHEP 2013 – Oct 15, 2013 Overview.
1 WAN Network Researches Wenji Wu, Phil Demar Fermilab Nov 30, 2007.
DOE Facilities - Drivers for Science: Experimental and Simulation Data
Introduction to Data Management in EGI
Establishing End-to-End Guaranteed Bandwidth Network Paths Across Multiple Administrative Domains The DOE-funded TeraPaths project at Brookhaven National.
An Overview of the ITTC Networking & Distributed Systems Laboratory
BigData Express: Toward Predictable, Schedulable, and High-performance Data Transfer Wenji Wu, April 4, 2019.
Presentation transcript:

Spectrum of Support for Data Movement and Analysis in Big Data Science Network Management and Control E-Center & ESCPS Network Management and Control E-Center & ESCPS Performance Optimization and Tools G-Netmon & MDTM Performance Optimization and Tools G-Netmon & MDTM 100GE Test and Evaluation Environment High Performance Data Program (HTDP)  Provides end-to-end and hop-by-hop network path data extending across multiple network domains.  Portal for end users to capture and discuss network path-related issues and interpretation of  Perfsonar-collected data with experts  Traffic forecasting capability for user-specified network paths  E-Center  Site integration with wide-area network circuit service  Local network configured for circuit-directed traffic Multicore-Aware Data Transfer Middleware (MDTM) - Harnessing Multicore Parallelism to Scale Data Movement Toolkits MDTM is a user-space resource scheduler that harness multicore parallelism to scale data movement toolkits at multicore systems.  Data Transfer-Centric Scheduling and Resource Management Capabilities.  NUMA Topology-Aware Scheduler  Supporting Core Affinity on Networking Processing  Supporting the QoS Mechanism to allow Differentiated Data Transfer G-NetMon: GPU-Accelerated Network Traffic Monitoring & Analysis Architecture  A combination of multicore and manycore technologies  Real-time Network Traffic Monitoring & Analysis  Handling 10,000,000+ pps capabilities Performance Analysis on ANI 100Gb/s Testbed 150ms+ latency between Globus Online servers and the testbed affects performance File size: 8MB Clients: 3000 for each of 3 hosts at NERSC File size: 8MB Clients: 3000 for each of 3 hosts at NERSC Dashed lines indicate scaled performance based on measurement using 10Gb/s connection Clients: 48 for each of 3 hosts at NERSC More clients ≠> Better performance Clients: 48 for each of 3 hosts at NERSC More clients ≠> Better performance Lots Of Small Files (LOSF) Problem on 1Gb/s v/s 100Gb/s At 100Gb/s, Medium Is The New Small Lots Of Small Files (LOSF) Problem on 1Gb/s v/s 100Gb/s At 100Gb/s, Medium Is The New Small At 100Gb/s, LOSF problem observed for medium files At 1Gb/s, LOSF problem observed for small files Reference: John Bresnahan, Michael Link, Rajkumar Kettimuthu, Dan Fraser, Ian Foster; “GridFTP Pipelining”, In the proceedings of Teragrid 2007 Conference, Madison, WI, USA ESnet Advanced Network Initiative (ANI) 100GE Testbed  Six hosts, three each at NERSC & ANL (RTT: 54ms). Each host has four 10 GE NICs.  Test Datasets: Files from 8KB to 8GB increasing in power of 2, split into three sets, small, medium and large. Each dataset contained files of different sizes. Future Plans 100GE at Fermilab  Dedicated 100 GE wave into Fermilab (ESnet ChiExpress MAN)  100 GE connection will be used for R&D, connecting Fermilab to ESnet ANI testbed.  Local testbed at Fermilab will include 10GE & 40GE host systems. HTDP Future Plans  Facilitate performance testing before moving 100 Gb/s to production - o Redo ANI tests using local testbed o Test the performance of other Grid Middleware tools like dCache with NFS4, CVMFS, IRODS, etc., based on the stakeholders needs Mission Mission of the High Throughput Data Program (HTDP) at Fermilab is to prepare the Laboratory and its stakeholders for 100GE infrastructure. Focus Test the performance and identify gaps in the Grid Middleware used by Fermilab stakeholders on the 100GE testbed operated by ESnet’s Advanced Networking Initiative (ANI) Phil Demar, David Dykstra, Gabriele Garzoglio, Parag Mhashilkar, Anupam Rajendran, Wenji Wu Big Data Networking at Fermilab Acknowledgement: This material is based upon work supported by the National Science Foundation under Grant No , ExTENCI: Extending Science Through Enhanced National Cyberinfrastructure. This research used resources of the ESnet Testbed, which is supported by the Office of Science of the U.S. Department of Energy under contract DE-AC02-05CH11231 Spectrum of Support for Data Movement and Analysis in Big Data Science End Site Control Plane System Forecasting network traffic conditions Configuration rules define routing modifications Social Portal on Site Network Issues Network path performance weather map End site part of end-to-end circuit service Multicore-Aware Data Transfer Middleware