HENP Grids and the Networks They Depend Upon

HENP, Grids and the Networks They Depend Upon Shawn Mc. Kee (smckee@umich. edu) March 2004 National Internet 2 Day March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics

Outline • HENP: Why do physicist’s care about the network? • GRIDs and networks in HENP • Doing physics at the LHC • Future and Conclusions March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 2

Physics and Networks So, why do physicists care about networks? • I will try to explain how physics will be done at LHC and the corresponding implications for the network needs • Networks, like Internet 2, are critical for the globally distributed, data intensive e-Science collaborations, like physics at the LHC • Details to follow… March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 3

Four LHC Experiments: The Petabyte to Exabyte Challenge ATLAS, CMS, ALICE, LHCB Higgs + New particles; Quark-Gluon Plasma; CP Violation Data stores ~40 Petabytes/Year and UP; CPU 0. 3 Petaflops and UP 0. 1 to 1. 0 Exabytes (1 EB = 1018 Bytes) (2007) (~2012 ? ) for the LHC Experiments March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 4

How Much Data is Involved? High Level-1 Trigger (1 MHz) Level 1 Rate (Hz) 106 LHCB 105 Hans Hoffman High No. Channels High Bandwidth (500 Gbit/s) ATLAS CMS HERA-B KLOE 104 Te. V II DOE/NSF High Data Archive (Peta. Byte) CDF/D 0 Review, Nov 00 103 H 1 ZEUS NA 49 UA 1 102 104 105 LEP March 18, 2004 ALICE 106 107 Event Size (bytes) Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 5

The Problem March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 6

The Solution March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 7

What is “The Grid”? • There are many answers and interpretations • The term was originally coined in the mid 1990’s (in analogy with the power grid) and can be described thusly: “The grid provides flexible, secure, coordinated resource sharing among dynamic collections of individuals, institutions and resources (virtual organizations: VOs)” March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 8

Grid Perspectives • Users Viewpoint: – A virtual computer which minimizes time to completion for my application while transparently managing access to inputs and resources • Programmers Viewpoint: – A toolkit of applications and API’s which provide transparent access to distributed resources • Administrators Viewpoint: – An environment to monitor, manage and secure access to geographically distributed computers, storage and networks. March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 9

Network Exponentials • Network vs. computer performance – Computer speed doubles every 18 months – Network speed doubles every 9 months – Difference = order of magnitude per 5 years • 1986 to 2000 – Computers: x 500 – Networks: x 340, 000 • 2001 to 2010 – Computers: x 60 – Networks: x 4000 March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 10

The Network • As can be seen in the previous transparency, it can be argued it is the evolution of the network which has been the primary motivator for the Grid. • Ubiquitous, dependable worldwide networks have opened up the possibility of tying together geographically distributed resources • The success of the WWW for sharing information has spawned a push for a system to share resources • The network has become the “virtual bus” of a virtual computer. March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 11

Doing Physics at the LHC ATLAS as an example March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics

ATLAS • A Torroidal LHC Apparatus • Collaboration – 150 institutes – 1850 physicists • Detector – – Inner tracker Calorimeter Magnet Muon • United States ATLAS – 29 universities, 3 national labs – 20% of ATLAS March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 13

March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 14

ATLAS March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 15

Discovery Potential for SM Higgs Boson • Good sensitivity over the full mass range from ~100 Ge. V to ~ 1 Te. V • For most of the mass range at least two channels available • Detector performance is crucial: b-tag, leptons, g, E resolution, g / jet separation, . . . March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 16

HEP Data Analysis • Raw data – hits, pulse heights • Reconstructed data (ESD) – tracks, clusters… • Analysis Objects (AOD) – Physics Objects – Summarized – Organized by physics topic • Ntuples, histograms, statistical data March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 17

Data Flow from ATLAS leve l 1 40 M Hz (~P B/se s peci 7 c) 5 al h K leve H z (75 ardwa l 2 5 KH embedde GB/sec) re d pr z( oces 5 G B sors / sec) l 3 (200 100 H PCs -400 z data MB/ sec) offl record i i leve ne a n naly g & sis March 18, 2004 ATLAS: 10 PB/y ~ one million PC hard drives! Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 18

HENP Grid/Network Projects • Grid Physics Network (Gri. Phy. N) – Enabling R&D for advanced data grid systems, focusing in particular on Virtual Data concept • i. VDGL: A Global Grid Laboratory – A global grid laboratory to conduct grid test “at scale” • There a numerous other projects focused on various aspects of grids and networks in support of HENP physics… March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 19

Ultra. Light: Exploring Future Networks for e-Science • Ultra. Light is a program to explore the integration of cutting-edge network technology with the grid computing and data infrastructure of HEP/Astronomy • The program intends to explore network configurations from common shared infrastructure (current IP networks) thru dedicated optical paths point-to-point. • A critical aspect of Ultra. Light is its integration with two driving application domains in support of their national and international e. Science collaborations: LHC-HEP and e. VLBI-Astronomy • The Collaboration includes: – – – Caltech Florida Int. Univ. MIT Univ. of Florida Univ. of Michigan March 18, 2004 ― UC Riverside ― BNL ― FNAL ― SLAC ― UCAID/Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 20

Increased functionality, standardization The Move to OGSA and then Managed Integration Systems ~Integrated Systems Web services + … X. 509, LDAP, FTP, … App-specific Services Open Grid Web Services Arch Resrc Framwk Stateful; Managed GGF: OGSI, … (+ OASIS, W 3 C) Globus Toolkit Multiple implementations, including Globus Toolkit Custom solutions March 18, 2004 Defacto standards GGF: Grid. FTP, GSI Time Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 21

Managing Global Systems: Dynamic Scalable Services Architecture Mon. ALISA: http: //monalisa. cacr. caltech. edu March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 22

Grid Analysis Environment CLARENS: Web Services Architecture Analysis Client u. Analysis Clients talk standard protocols to a simple API HTTP, SOAP, XML/RPC Grid Services Web Server Scheduler Catalogs Fully. Abstract Planner Metadata Partially. Abstract Planner Fully. Concrete Planner u. The secure Clarens portal hides the complexity Virtual Data Management Monitoring Replica Execution Priority Manager Applications u. Key features: Global Scheduler, Catalogs, Monitoring, and Gridwide Execution service u. The network underlies and enables this model Grid Wide Execution Service March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 23

Conclusions • Networks form the critical basis for the future of e-Science • LHC Physics will depend heavily on globally distributed resources => the NETWORK is critical! • Future requirements for grids and networking in support of HENP physics is an open question which will need investigation to define, develop and deploy the needed infrastructure in a timely manner. March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 24

For More Information… • HENP Internet 2 SIG – henp. internet 2. edu • Global Grid Forum – www. ggf. org • International Virtual Data Grid Laboratory – www. ivdgl. org • Grid Physics Network – www. griphyn. org • Ultra. Light: ultralight. caltech. edu Questions? March 18, 2004 Internet 2 Day - Shawn Mc. Kee - University of Michigan Physics 25