DYNES Storage Infrastructure Artur Barczyk California Institute of Technology LHCOPN Meeting Geneva, October 07, 2010.

Slides:



Advertisements
Similar presentations
Client/Server Computing (the wave of the future) Rajkumar Buyya School of Computer Science & Software Engineering Monash University Melbourne, Australia.
Advertisements

Middleware Support for RDMA-based Data Transfer in Cloud Computing Yufei Ren, Tan Li, Dantong Yu, Shudong Jin, Thomas Robertazzi Department of Electrical.
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.
ATLAS Tier-3 in Geneva Szymon Gadomski, Uni GE at CSCS, November 2009 S. Gadomski, ”ATLAS T3 in Geneva", CSCS meeting, Nov 091 the Geneva ATLAS Tier-3.
Netscape Application Server Application Server for Business-Critical Applications Presented By : Khalid Ahmed DS Fall 98.
Network+ Guide to Networks, Fourth Edition
GridFTP: File Transfer Protocol in Grid Computing Networks
Network-Attached Storage
Protocols and the TCP/IP Suite Chapter 4 (Stallings Book)
Threads - Definition - Advantages using Threads - User and Kernel Threads - Multithreading Models - Java and Solaris Threads - Examples - Definition -
Protocols and the TCP/IP Suite
Embedded Transport Acceleration Intel Xeon Processor as a Packet Processing Engine Abhishek Mitra Professor: Dr. Bhuyan.
1 I/O Management in Representative Operating Systems.
Traffic shaping with OVS and SDN Ramiro Voicu Caltech LHCOPN/LHCONE, Berkeley, June
Network+ Guide to Networks, Fourth Edition Chapter 1 An Introduction to Networking.
Installing software on personal computer
Virtual Network Servers. What is a Server? 1. A software application that provides a specific one or more services to other computers  Example: Apache.
Protocols and the TCP/IP Suite Chapter 4. Multilayer communication. A series of layers, each built upon the one below it. The purpose of each layer is.
Christopher Bednarz Justin Jones Prof. Xiang ECE 4986 Fall Department of Electrical and Computer Engineering University.
Socket Programming -What is it ? -Why bother ?. Basic Interface for programming networks at transport level It is communication end point Used for inter.
CMS Data Transfer Challenges LHCOPN-LHCONE meeting Michigan, Sept 15/16th, 2014 Azher Mughal Caltech.
Performance Tradeoffs for Static Allocation of Zero-Copy Buffers Pål Halvorsen, Espen Jorde, Karl-André Skevik, Vera Goebel, and Thomas Plagemann Institute.
Network+ Guide to Networks, Fourth Edition Chapter 1 An Introduction to Networking.
Protocols for Wide-Area Data-intensive Applications: Design and Performance Issues Yufei Ren, Tan Li, Dantong Yu, Shudong Jin, Thomas Robertazzi, Brian.
©Kwan Sai Kit, All Rights Reserved Windows Small Business Server 2003 Features.
Technology Overview. Agenda What’s New and Better in Windows Server 2003? Why Upgrade to Windows Server 2003 ?  From Windows NT 4.0  From Windows 2000.
Pooja Shetty Usha B Gowda.  Network File Systems (NFS)  Drawbacks of NFS  Parallel Virtual File Systems (PVFS)  PVFS components  PVFS application.
M i SMob i S Mob i Store - Mobile i nternet File Storage Platform Chetna Kaur.
Boosting Event Building Performance Using Infiniband FDR for CMS Upgrade Andrew Forrest – CERN (PH/CMD) Technology and Instrumentation in Particle Physics.
Ramiro Voicu December Design Considerations  Act as a true dynamic service and provide the necessary functionally to be used by any other services.
VIRTUAL PRIVATE NETWORK By: Tammy Be Khoa Kieu Stephen Tran Michael Tse.
Csi315csi315 Client/Server Models. Client/Server Environment LAN or WAN Server Data Berson, Fig 1.4, p.8 clients network.
Swapping to Remote Memory over InfiniBand: An Approach using a High Performance Network Block Device Shuang LiangRanjit NoronhaDhabaleswar K. Panda IEEE.
Thoughts on Future LHCOPN Some ideas Artur Barczyk, Vancouver, 31/08/09.
Remote Access Using Citrix Presentation Server December 6, 2006 Matthew Granger IT665.
Hands-On Microsoft Windows Server Implementing Microsoft Internet Information Services Microsoft Internet Information Services (IIS) –Software included.
Tool Integration with Data and Computation Grid GWE - “Grid Wizard Enterprise”
Servicii distribuite Alocarea dinamică a resurselor de reea pentru transferuri de date de mare viteză folosind servicii distribuite Distributed Services.
Operating Systems David Goldschmidt, Ph.D. Computer Science The College of Saint Rose CIS 432.
Reconsidering Internet Mobility Alex C. Snoeren, Hari Balakrishnan, M. Frans Kaashoek MIT Laboratory for Computer Science.
Latest news on JXTA and JuxMem-C/DIET Mathieu Jan GDS meeting, Rennes, 11 march 2005.
1 Public DAFS Storage for High Performance Computing using MPI-I/O: Design and Experience Arkady Kanevsky & Peter Corbett Network Appliance Vijay Velusamy.
INTRODUCTION TO DBS Database: a collection of data describing the activities of one or more related organizations DBMS: software designed to assist in.
Intel Research & Development ETA: Experience with an IA processor as a Packet Processing Engine HP Labs Computer Systems Colloquium August 2003 Greg Regnier.
 The End to the Means › (According to IBM ) › 03.ibm.com/innovation/us/thesmartercity/in dex_flash.html?cmp=blank&cm=v&csr=chap ter_edu&cr=youtube&ct=usbrv111&cn=agus.
Paperless Timesheet Management Project Anant Pednekar.
CSI 3125, Preliminaries, page 1 SERVLET. CSI 3125, Preliminaries, page 2 SERVLET A servlet is a server-side software program, written in Java code, that.
ALCF Argonne Leadership Computing Facility GridFTP Roadmap Bill Allcock (on behalf of the GridFTP team) Argonne National Laboratory.
The Intelligent Infrastructure John Pollard – 29 th April 2008
ANSE: Advanced Network Services for Experiments Institutes: –Caltech (PI: H. Newman, Co-PI: A. Barczyk) –University of Michigan (Co-PI: S. McKee) –Vanderbilt.
1 R. Voicu 1, I. Legrand 1, H. Newman 1 2 C.Grigoras 1 California Institute of Technology 2 CERN CHEP 2010 Taipei, October 21 st, 2010 End to End Storage.
2.2 Interfacing Computers MR JOSEPH TAN CHOO KEE TUESDAY 1330 TO 1530
Recent experience with PCI-X 2.0 and PCI-E network interfaces and emerging server systems Yang Xia Caltech US LHC Network Working Group October 23, 2006.
Amazon Web Services. Amazon Web Services (AWS) - robust, scalable and affordable infrastructure for cloud computing. This session is about:
SDN controllers App Network elements has two components: OpenFlow client, forwarding hardware with flow tables. The SDN controller must implement the network.
Jean-Philippe Baud, IT-GD, CERN November 2007
The Distributed Application Debugger (DAD)
Project Target Develop a Web Based Management software suit that will enable users to control Hardware using standard HTTP & Java Applet compatible web.
WP18, High-speed data recording Krzysztof Wrona, European XFEL
Netscape Application Server
Securing the Network Perimeter with ISA 2004
Introduction to Data Management in EGI
Integration of Network Services Interface version 2 with the JUNOS Space SDK
Enabling High Speed Data Transfer in High Energy Physics
GGF15 – Grids and Network Virtualization
Telnet/SSH Connecting to Hosts Internet Technology.
An Introduction to Computer Networking
Network+ Guide to Networks, Fourth Edition
A Network Operating System Edited By Maysoon AlDuwais
MPJ: A Java-based Parallel Computing System
Presentation transcript:

DYNES Storage Infrastructure Artur Barczyk California Institute of Technology LHCOPN Meeting Geneva, October 07, 2010

DYNES Instrument at Tier2 & 3 DYNES instrument comes with a storage server and attached disk array DYNES instrument allows connecting other (e.g. existing) storage elements!

DYNES Storage  The storage part of the DYNES instrument will consist of (per deployment instance at Tier2/3 site)  One FDT server  One attached disk array (SAS)  FDT will be used as transport application  FDT/Hadoop  FDT/dCache

FDT – Fast Data Transfer  FDT is an open source application for efficient data transfers.  Easy to use: similar syntax with SCP, iperf/netperf  Written in java and runs on all major platforms.  Single.jar file (~800 KB)  Based on an asynchronous, multithreaded system  Uses the New I/O (NIO) interface and is able to:  stream continuously a list of files  use independent threads to read and write on each physical device  transfer data in parallel on multiple TCP streams, when necessary  use appropriate size of buffers for disk IO and networking  resume a file transfer session

FDT - Architecture Pool of buffers Kernel Space Pool of buffers Kernel Space Data Transfer Sockets / Channels Independent threads per device Restore the files from buffers Control connection / authorization Ramiro Voicu

FDT features  User defined loadable modules for Pre and Post Processing to provide support for dedicated Mass Storage system, compression, dynamic circuit setup, …  Pluggable file systems “providers” (e.g. non-POSIX FS)  Dynamic bandwidth limitations  Different transport strategies:  blocking (1 thread per channel)  non-blocking (selector + pool of threads)  On the fly MD5 checksum on the reader side  Configurable number of streams and threads per physical device (useful for distributed FS)  Automatic updates  Can be used as network testing tool (/dev/zero → /dev/null memory transfers, or –nettest flag)

FDT security  DYNES security is based on secure point-to-point connection setup  AA for circuit setup  In addition, FDT architecture allows to "plug-in" external security APIs and to use them for client authentication and authorization  Supports several security schemes :  IP based ACL filtering  SSH  GSI-SSH  Standalone Globus-GSI  Plain SSL

FDT performance: Memory-to-Memory WAN data transfers (CERN-Caltech) % CPU idle 50 % CPU idle CPU utilisation

FDT Performance: Storage Storage-to-storage performance between pair of servers: sustained 2.6 Gbps

40G  Recently received a pair of Mellanox 40GE NICs  Performance tests done in CERN Openlab and Ultralight environment  Example: Memory-to-Memory in LAN 25Gbps: hitting the PCIe v2 (8 lane) limit! Need PCIe v3 for full 40Gbps 25Gbps: hitting the PCIe v2 (8 lane) limit! Need PCIe v3 for full 40Gbps Unidirectional transfers

40G Bi-directional memory-to-memory transfers Currently investigating storage transfer performance

FDT with Dynamic Circuits: GLIF’09 Demo July 2010 Ramiro Voicu FDT can use IDC API to set up lightpaths. Example: Caltech Tier2 to compute cluster at CERN FDT can use IDC API to set up lightpaths. Example: Caltech Tier2 to compute cluster at CERN

Path setup  3 domains involved, all using DCN/ION (OSCARS+DRAGON)  Caltech  Internet2  USLHCNet  Path requested by FDT to USLHCNet IDC

Automatic path selection July 2010 Ramiro Voicu FDT automatically selects the correct interface to send data No dynamic circuit, use default 1GbE interface Successful setup of Lightpath, transfer speed limited by capability of server!

FDT-PhEDEx integration  Work ongoing in CMS  Will facilitate the integration of DYNES instrument in the CMS data operations  (Will be presented at CHEP’10)

FDT Summary & Future developments  FDT is a mature and a robust open source software  Key features:  Portability – runs on all major platforms  Simple to use and small size  Streams data over multiple channels  Pluggable security (SSH, GSI, GSI+SSH, …)  Can be used as a network testing tool (TCP only)  Pluggable user filters ( e.g. MS storage, compression, …)  Dynamic circuits capability  Future developments:  GUI interface  New features once Java7 will be released  NIO.2 (asynchronous I/O, new FS interface, SCTP, …)  FJ tasks

THANK YOU!