avery ivdgl 2004may20

50 %
50 %
Information about avery ivdgl 2004may20

Published on March 25, 2008

Author: Sigismondo

Source: authorstream.com

Slide1:  Paul Avery University of Florida avery@phys.ufl.edu International Virtual Data Grid Laboratory Trillium Steering Meeting University of Chicago May 20, 2004 iVDGL Goals:  iVDGL Goals Deploy a Grid laboratory Support research mission of data intensive experiments Provide computing and personnel resources at university sites Provide platform for computer science technology development Prototype and deploy a Grid Operations Center (iGOC) Integrate Grid software tools Into computing infrastructures of the experiments Support delivery of Grid technologies Hardening of the Virtual Data Toolkit (VDT) and other middleware technologies developed by GriPhyN and other Grid projects Education and Outreach Lead and collaborate with Education and Outreach efforts Provide tools and mechanisms for underrepresented groups and remote regions to participate in international science projects iVDGL Institutions:  iVDGL Institutions U Florida Boston Caltech Chicago Harvard Hampton Indiana Johns Hopkins Penn State Salish Kootenai UC San Diego Southern California Texas, Austin Texas, Brownsville Wisconsin, Milwaukee Wisconsin, Madison Argonne Brookhaven Fermilab LBL SDSC SLAC New sites Vanderbilt (Summer 2003) Iowa (Fall 2003) Michigan (May 2004?) FIU (May 2004?) U Buffalo (May 2004?) STAR, etc. (May 2004?) (Funded by iVDGL) iVDGL Sites (February 2004):  iVDGL Sites (February 2004) UF UW Madison BNL Indiana Boston U SKC Brownsville Hampton PSU J. Hopkins Caltech FIU Austin Michigan LBL Argonne Vanderbilt UCSD Fermilab Partners EU Brazil Korea Iowa Chicago UW Milwaukee ISI New Experiments and Institutions:  New Experiments and Institutions Vanderbilt (BTEV) (added summer 2003) Fermilab experiment, startup in 2008 Intriguing idea of using Grid for online “Level 4” trigger Iowa (added summer 2003) US-CMS Buffalo (considered tomorrow) Biology, other apps Florida International University (considered tomorrow) US-CMS, AMPATH, major expansion of education outreach Michigan (considered tomorrow) US-ATLAS, networking, School of Information Science HENP experiments within PPDG (considered tomorrow) STAR iVDGL Budget (9/2001 – 9/2006):  iVDGL Budget (9/2001 – 9/2006) iVDGL Budget: Major Activities:  iVDGL Budget: Major Activities Recent Supplemental Funds:  Recent Supplemental Funds $387K for some “new” activities $250K for VDT support of LCG (U Wisconsin) $ 82K for Tom Greene (outreach, reach new communities) $ 30K for demos at Dec. 2003 WSIS (CERN + FNAL) $ 25K for E/O workshop in Miami (Jan. 2004) $418K previously requested to funnel NSF funds to CMS Rules permit <20% in supplements In theory, this permits $1.9M more to be requested (< 1 year) Funds all come from Physics Division Although no money now, allows quick funding of new initiatives Think about GriPhyN supplement also (CISE)? (~ $2M) $25K in new funds obtained May 19 for Grid Tutorial New proposal from Soma Mukherjee Students:  Students Involved in many aspects of iVDGL J. Zamora (B.S., UT Brownsville) Outreach, LIGO A. Gupta (M.S. UFL) Web pages A. Zahn (B.S. UFL) Info. systems, web pages L. Sorrillo (B.S. Hampton) Grid2003 S. Morriss (B.S. UT Brownsville) LIGO, Grid2003 A. Martinez (B.S. UT Brownsville) LIGO, web pages P. Peiris (B.S. UT Brownsville) LIGO, web pages Theses Sonal Patil (M.S., USC) Pegasus 10/2003 Sarp Oral (Ph.D, UFL) Grid networking 12/2003 S. Pena (M.S. UT Brownsville) LIGO analysis 9/2004 R. Subramaniyan (Ph.D, UFL) Grid monitoring 12/2005 R. Balasubramanian (Ph.D, UFL) Grid networking 12/2006 Students (2):  Students (2) GriPhyN students contributing (using iVDGL resources) K. Ranagnathan (Ph.D, Chicago) Data placement C. Dumitrescu (Ph.D, Chicago) Policy models, enforcement Y. Zhou (Ph.D Chicago) Grid portal (outreach) J. Uk In (Ph.D, Florida) Scheduling, policy algorithms Students from other projects A. Rodriguez (Ph.D, Illinois) Bioinformatics QuarkNet Paid for 3 students to spend time at Fermilab Slide11:  Grid2003: An Operational National Grid 28 sites: Universities + national labs (Fermilab, Brookhaven, Argonne, Berkeley) 2800 CPUs, 400-1300 jobs Running since October 2003 HEP, gravity-wave, digital astronomy, Genomics Korea http://www.ivdgl.org/grid2003 LSU? JLAB? UTB FIU Grid Software Packaging: Pacman:  Language: define software environments Interpreter: create, install, configure, update, verify environments LCG/Scram ATLAS/CMT CMS DPE/tar/make LIGO/tar/make OpenSource/tar/make Globus/GPT NPACI/TeraGrid/tar/make D0/UPS-UPD Commercial/tar/make Combine and manage software from arbitrary sources. % pacman –get iVDGL:Grid3 “1 button install”: Reduce burden on administrators Remote experts define installation/ config/updating for everyone at once Grid Software Packaging: Pacman iVDGL, GriPhyN Education / Outreach:  iVDGL, GriPhyN Education / Outreach Basics $200K/yr Led by UT Brownsville Workshops, portals Partnerships with CHEPREO, QuarkNet, … CHEPREO: Center for High Energy Physics Research and Educational Outreach Florida International University:  CHEPREO: Center for High Energy Physics Research and Educational Outreach Florida International University Physics Learning Center CMS Research iVDGL Grid Activities AMPATH network (S. America) Funded September 2003 $4M initially (3 years) Outreach Events:  Outreach Events Grid Outreach mini-summit April 8 in DC iVDGL, CHEPREO, QuarkNet, Exploratorium, NSF, DOE, etc. Discussed programs, build collaborations Look for new funding mechanisms Grid Tutorial: South Padre Island, Texas June 21-25 ~30 undergraduate students Slide16:  25-30 students Grids and the Digital Divide Rio de Janeiro, Feb. 16-20, 2004:  NEWS: Bulletin: ONE TWO WELCOME BULLETIN General Information Registration Travel Information Hotel Registration Participant List How to Get UERJ/Hotel Computer Accounts Useful Phone Numbers Program Contact us: Secretariat Chairmen Grids and the Digital Divide Rio de Janeiro, Feb. 16-20, 2004 Background World Summit on Information Society HEP Standing Committee on Inter-regional Connectivity (SCIC) Themes Global collaborations, Grids and addressing the Digital Divide Next meeting: 2005 (Korea) http://www.uerj.br/lishep2004 iVDGL Review 2/11/2004:  iVDGL Review 2/11/2004 Geoffrey Fox (U. Indiana) Edward Walker (TACC, U. Texas at Austin) Anders Ryd (Cornell U.) Alec Habig (U. Minnesota at Duluth) Vijay Samalam (SDSC, U. Cal. San Diego), Chair Importance of iVDGL:  Importance of iVDGL iVDGL (international Virtual Data Grid Laboratory) has to continue to succeed! It is important for the LHC (Large Hadron Collider), LIGO (Laser Interferometer Gravitational-Wave Observatory), and SDSS (Sloan Digital Sky Survey) It is a show case for how grids can be used High Energy Physics (HEP) has experience working in large groups using distributed resources The problem frequently parses naturally for distributed resources If grids do not work for HEP it is hard to see them working in other areas Project Accomplishments:  Project Accomplishments Have done an outstanding job of meeting project milestones and achieving deliverables. In particular we appreciate the following: Laboratory concept is a great success and this implementation is very good The approach taken by the VDT (Virtual Data Toolkit) team is excellent. The software process is good and deployment methodology is also very good Project Accomplishments – Cont’d.:  Project Accomplishments – Cont’d. Documenting lessons learned is a very good idea iGOC (the iVDGL Grid Operations Center) – the concept is excellent and it looks like implementation with a suite of tools is very good The potential for integrating both network monitoring and grid monitoring is a golden opportunity The SuperCcomputing2003 demo was a demonstration of a very successful first example of a persistent grid Project Accomplishments – Cont’d.:  Project Accomplishments – Cont’d. Liked the presentation template of LIGO. Recommended it in the future for other applications. Outreach activities is particularly impressive Outreach activities have real teeth to them with demonstrable results. Recommendations:  Recommendations Keep focused on the original goals and try to avoid mission creep especially if it involves significant resources Worried about what Open Science entails The team needs to define the concept of federation and interoperability better The team needs to actively collaborate with Europe and integrate with other US developments Recommendations - Cont’d:  Recommendations - Cont’d Procedures should be put in place to partition the test bed into a production part and an experimental part The result of establishing a production lab should be a better understanding of the difference between grid specific issues and application specific issues The iVDGL team should be more aggressive in publicizing their SC2003 grid demo success iVDGL should find a new name for the SC2003 persistent grid Recommendations – Cont’d :  Recommendations – Cont’d The team should start a process to define application requirements, define application metrics and monitor application performance against these metrics Establish procedures for outsiders to use the test bed and also advertise the test bed better Establish milestones and deliverables that are also demo independent Recommendations – Cont’d :  Recommendations – Cont’d For new users, especially students, a web based orientation course on grid computing, clusters etc. The course should especially concentrate on describing the physics at the right level Should also look into contacting other umbrella minority organizations , such as AN-MSI (Advanced Networking With Minority-Serving Institutions) Recommendations - Technical:  Recommendations - Technical Portlet architecture for portals be evaluated Evaluate to what extent heterogeneity can be supported Evaluate fabric management software Evaluate other monitoring software Recommendations for NSF:  Recommendations for NSF Should seriously think about how the project will be supported beyond the formal end period Should coordinate with other agencies funding other grids to come up with common win strategy A possible approach is to modularize all software development and share the costs resulting in a single software stack Recommendations to NSF:  Recommendations to NSF Suggest exploring funding models that will facilitate putting resources and funds into producing production quality code Should be an active partner in defining federation and interoperability with other grids

Add a comment

Related presentations

Related pages

Your Favourite Online StoreHome - Your Favourite Online Store

Copyright © 2014 PPDG . Designed by Care Vitamins Online. feedback Your Name * Email * Subject . Message. Top ...
Read more


avery ivdgl 2004may20. Education. Tweet. 24. 01. 2008 0 views czerniawska slides. Education. Tweet. 07. 04. 2008 0 views ...
Read more

Voting Behavior

Voting Behavior: Voting Behavior By: Cameron Moyer, Chelsea Radford, and Jami Moffett, Kyle Schmidt Words you need to know:: Words you need to know ...
Read more