PSC. BigBen Features Compute Nodes 2068 nodes running Catamount (QK) microkernel Seastar interconnect in a 3-D torus configuration No external.

Slides:



Advertisements
Similar presentations
1 Routing Protocols I. 2 Routing Recall: There are two parts to routing IP packets: 1. How to pass a packet from an input interface to the output interface.
Advertisements

Copyright © 2014 EMC Corporation. All Rights Reserved. Basic Network Configuration for File Upon completion of this module, you should be able to: Configure.
AT LOUISIANA STATE UNIVERSITY CCT: Center for Computation & LSU Stork Data Scheduler: Current Status and Future Directions Sivakumar Kulasekaran.
Protocols and software for exploiting Myrinet clusters Congduc Pham and the main contributors P. Geoffray, L. Prylli, B. Tourancheau, R. Westrelin.
Institute of Computer Science Foundation for Research and Technology – Hellas Greece Computer Architecture and VLSI Systems Laboratory Exploiting Spatial.
PROMISE: Peer-to-Peer Media Streaming Using CollectCast Mohamed Hafeeda, Ahsan Habib et al. Presented By: Abhishek Gupta.
IP –Based SAN extensions and Performance Thao Pham CS 622 Fall 07.
Socket Programming.
Multiprocessors ELEC 6200: Computer Architecture and Design Instructor : Agrawal Name: Nam.
An Introduction to Parallel Computing Dr. David Cronk Innovative Computing Lab University of Tennessee Distribution A: Approved for public release; distribution.
Learning Objectives Understanding the difference between processes and threads. Understanding process migration and load distribution. Understanding Process.
1 Lecture 24: Interconnection Networks Topics: communication latency, centralized and decentralized switches (Sections 8.1 – 8.5)
CS335 Networking & Network Administration Tuesday, April 20, 2010.
TCP: Software for Reliable Communication. Spring 2002Computer Networks Applications Internet: a Collection of Disparate Networks Different goals: Speed,
Jennifer Rexford Princeton University MW 11:00am-12:20pm Wide-Area Traffic Management COS 597E: Software Defined Networking.
Pooja Shetty Usha B Gowda.  Network File Systems (NFS)  Drawbacks of NFS  Parallel Virtual File Systems (PVFS)  PVFS components  PVFS application.
Textbook  “Data Communications and Networking” 2 nd Edition by Behrouz A. Forouzan  “Data and Computer Communication” 6 th Edition by William Stallings.
Profiling Grid Data Transfer Protocols and Servers George Kola, Tevfik Kosar and Miron Livny University of Wisconsin-Madison USA.
Robert Fourer, Jun Ma, Kipp Martin Copyright 2006 An Enterprise Computational System Built on the Optimization Services (OS) Framework and Standards Jun.
The MPC Parallel Computer Hardware, Low-level Protocols and Performances University P. & M. Curie (PARIS) LIP6 laboratory Olivier Glück.
QoS Support in High-Speed, Wormhole Routing Networks Mario Gerla, B. Kannan, Bruce Kwan, Prasasth Palanti,Simon Walton.
NETWORKING. OBJECTIVES Identify network topologies Identify hardware components of a network.
Oracle Tuning Ashok Kapur Hawkeye Technology, Inc.
CCNA 2 Week 1 Routers and WANs. Copyright © 2005 University of Bolton Welcome Back! CCNA 2 deals with routed networks You will learn how to configure.
Example: Sorting on Distributed Computing Environment Apr 20,
Parallel Programming on the SGI Origin2000 With thanks to Igor Zacharov / Benoit Marchand, SGI Taub Computer Center Technion Moshe Goldberg,
Storage Interconnect Requirements Chen Zhao, Frank Yang NetApp, Inc.
1 CMPE 511 HIGH PERFORMANCE COMPUTING CLUSTERS Dilek Demirel İşçi.
Messages asynchronously reads or writes a block of data to another module on a network. Controller communication done using local or remote messages. local.
Washington WASHINGTON UNIVERSITY IN ST LOUIS CP and Full MSR Test Status.
Lecture (Mar 23, 2000) H/W Assignment 3 posted on Web –Due Tuesday March 28, 2000 Review of Data packets LANS WANS.
1 Novell NetWare Protocol Stack Media Access Protocols (Ethernet, Token Ring, WAN, others) Physical Data Link Network Session Transport Presentation.
Intro to Distributed Systems and Networks Hank Levy.
Message Passing Computing 1 iCSC2015,Helvi Hartmann, FIAS Message Passing Computing Lecture 2 Message Passing Helvi Hartmann FIAS Inverted CERN School.
Module 16: Distributed System Structures Silberschatz, Galvin and Gagne ©2005 Operating System Concepts – 7 th Edition, Apr 4, 2005 Distributed.
Parallel IO for Cluster Computing Tran, Van Hoai.
SPEAKER: MUHAMMAD REZA ZULMAN DATE: NOVEMBER 17, 2014 OPENFLOW SPECIFICATION.
PIDX PIDX - a parallel API to capture the data models used by HPC application and write it out in an IDX format. PIDX enables simulations to write out.
CDA-5155 Computer Architecture Principles Fall 2000 Multiprocessor Architectures.
DAQ & ConfDB Configuration DB workshop CERN September 21 st, 2005 Artur Barczyk & Niko Neufeld.
CHAPTER -II NETWORKING COMPONENTS CPIS 371 Computer Network 1 (Updated on 3/11/2013)
TeraGrid Data Transfer Joint EGEE and OSG Workshop on Data Handling in Production Grids June 25, Monterey, CA Derek Simmel
LonWorks Introduction Hwayoung Chae.
Transmission Control Protocol A Reliable, Connection-Oriented, Byte-Stream Service Lab 9.
CHAPTER 3 Router CLI Command Line Interface. Router User Interface User and privileged modes User mode --Typical tasks include those that check the router.
Chapter 10 Congestion Control in Data Networks and Internets 1 Chapter 10 Congestion Control in Data Networks and Internets.
Bandwidth estimation: metrics, measurement techniques, and tools Presenter: Yuhang Wang.
Mini-Workshop on multi-core joint project Peter van Gemmeren (ANL) I/O challenges for HEP applications on multi-core processors An ATLAS Perspective.
Towards a High Performance Extensible Grid Architecture Klaus Krauter Muthucumaru Maheswaran {krauter,
1 Chapter 1 INTRODUCTION TO WEB. 2 Objectives In this chapter, you will: Become familiar with the architecture of the World Wide Web Learn about communication.
Instructor Materials Chapter 7: EIGRP Tuning and Troubleshooting
Hello everyone I am rajul and I’ll be speaking on a case study of elastic slurm Case Study of Elastic Slurm Rajul Kumar, Northeastern University
The Distributed Application Debugger (DAD)
California Institute of Technology
Lecture 23: Interconnection Networks
Routers Multiport connectivity device
TYPES OFF OPERATING SYSTEM
Large Scale Test of a storage solution based on an Industry Standard
Chapter 16: Distributed System Structures
Parallel Programming in C with MPI and OpenMP
CprE 458/558: Real-Time Systems
Pluggable Architecture for Java HPC Messaging
MPI-Message Passing Interface
Overview of Computer Architecture and Organization
Indiana University, Bloomington
PVFS: A Parallel File System for Linux Clusters
Optical communications & networking - an Overview
Ch 17 - Binding Protocol Addresses
Counter Mode, Output Feedback Mode
Presentation transcript:

PSC

BigBen Features Compute Nodes 2068 nodes running Catamount (QK) microkernel Seastar interconnect in a 3-D torus configuration No external connectivity (no TCP) All Inter-node communication is over Portals Applications use MPI which is based on Portals Service & I/O Nodes (SIO) Nodes 22 nodes running Suse Linux Also on the Seastar interconnect SIO nodes can have PCI-X hardware installed, defining unique roles for each 2 SIO nodes are externally connected to ETF with 10GigE cards (currently)

Portals Direct I/O (PDIO) Details Portals-to-TCP routing –PDIO daemons aggregate hundreds of portals data streams into a configurable number of outgoing TCP streams –Heterogenous portals (both QK + Linux nodes) Explicit Parallelism –Configurable # of Portals receivers (on SIO nodes) Distributed across multiple 10GigE-connected Service & I/O (SIO) nodes –Corresponding # of TCP streams (to the WAN) one per PDIO daemon –A Parallel TCP receiver in the Goodhue booth Supports a variable/dynamic number of connections

Portals Direct I/O (PDIO) Details Utilizing the ETF network –10GigE end-to-end –Benchmarked >1Gbps in testing Inherent flow-control feedback to application –Aggregation protocol allows TCP transmission or even remote file system performance to throttle the data streams coming out of the application (!) Variable message sizes and file metadata supported Multi-threaded ring buffer in the PDIO daemon –Allows the Portals receiver, TCP sender, and computation to proceed asynchronously

Portals Direct I/O (PDIO) Config User-configurable/tunable parameters: –Network targets Can be different for each job –Number of streams Can be tuned for optimal host/network utilization –TCP network buffer size Can be tuned for maximum throughput over the WAN –Ring buffer size/length Controls total memory utilization of PDIO daemons –Number of portals writers Can be any subset of the running application’s processes –Remote filename(s) File metadata are propagated through the full chain, per write

ETF network Compute Nodes I/O Nodes Steering iGRIDPSC HPC resource and renderer waiting…

pdiod recv ETF network Compute Nodes I/O Nodes Steering iGRIDPSC Launch PPM job, PDIO daemons, and iGRID recv’ers

pdiod recv ETF network Compute Nodes I/O Nodes Steering iGRIDPSC Aggregate data via Portals

pdiod recv ETF network Compute Nodes I/O Nodes Steering iGRIDPSC Route traffic to ETF net

pdiod recv ETF network Compute Nodes I/O Nodes Steering iGRIDPSC Recv iGRID

pdiod recv ETF network render Compute Nodes I/O Nodes Steering iGRIDPSC Render real-time data

pdiod recv ETF network render Compute Nodes I/O Nodes Steering iGRIDPSC Send steering data back to active job input

pdiod recv ETF network render Compute Nodes I/O Nodes Steering iGRIDPSC Dynamically update rendering input