GridBased HPC Computing on University Campuses

1

• Grid-Based HPC Computing on University Campuses

Paul Avery University of Florida avery_at_phys.ufl.ed
u
Phase II HPC MeetingJune 14, 2005
2
Examples Discussed Here

• Four campuses, in different states of readiness
• University of Wisconsin GLOW
• University of Michigan MGRID
• University of Florida UF Research Grid
• University of Texas TACC Grid (not discussed
  here)
• Not complete, by any means
• Goal illustrate factors that go into creating
  campus Grid facilities

3
Grid Laboratory of Wisconsin (GLOW)

• 2003 Initiative funded by NSF/UW Six GLOW Sites
• Computational Genomics, Chemistry
• Amanda, Ice-cube, Physics/Space Science
• High Energy Physics/CMS, Physics
• Materials by Design, Chemical Engineering
• Radiation Therapy, Medical Physics
• Computer Science
• Deployed in two Phases
• University funding plus 1.1M from MRI grant
• Managed locally at 6 sites
• Shared globally across all sites
• Higher priority for local jobs

http//www.cs.wisc.edu/condor/glow/
4
GLOW Deployment

• GLOW Phase-I and II are commissioned
• CPU
• 66 nodes each _at_ ChemE, CS, LMCG, MedPhys
• 60 nodes _at_ Physics
• 30 nodes _at_ IceCube
• 50 extra nodes _at_ CS (ATLAS)
• Total CPU 800
• Storage
• Head nodes _at_ at all sites
• 45 TB each _at_ CS and Physics
• Total storage 100 TB
• GLOW resources used at 100 level
• Key is having multiple user groups

5
Resource Sharing in GLOW

• Six GLOW sites
• Equal priority ? 17 average
• Chemical Engineering took 33
• Others scavenge idle resources
• Yet, they got 39

Efficient users can realize more than they put in
6
Adding New GLOW Members

• GLOW team membership
• System support person joins GLOW-tech
• PI joins GLOW-exec
• Proposed minimum involvement 50 CPUs?
• Sponsored by existing GLOW members
• ATLAS physics group proposed by CMS physics team
• Condensed matter group proposed by CS
• Expressions of interest from other groups
• Agreements to adhere to GLOW policies

7
Simulations on Condor/GLOW (1)
All of INFN
8
Simulations on Condor/GLOW (2)
9.5M events generated in 2004
9
MGRID at Michigan

• MGRID
• Michigan Grid Research and Infrastructure
  Development
• Develop, deploy, and sustain an institutional
  grid at Michigan
• Group started in 2002 with initial U Michigan
  funding
• Many groups across the University participate
• Compute/data/network-intensive research grants
• ATLAS, NPACI, NEESGrid, Visible Human Project,
  NFSv4, NMI
• 1300 CPUs in main clusters

http//www.mgrid.umich.edu
10
MGRID Center

• Central core of technical staff (3FTEs, new
  hires)
• Faculty and staff from participating units
• Exec. committee from participating units
  provost office
• Collaborative grid research and development with
  technical staff from participating units

11
MGrid Research Project Partners

• College of LSA
• www.lsa.umich.edu
• Center for Information Technology Integration
• www.citi.umich.edu
• Michigan Center for BioInformatics
• www.ctaalliance.org
• Visible Human Project
• vhp.med.umich.edu
• Center for Advanced Computing
• cac.engin.umich.edu
• Mental Health Research Institute
• www.med.umich.edu/mhri
• ITCom
• www.itcom.itd.umich.edu
• School of Information
• si.umich.edu

12
MGRID Goals

• For participating units
• Knowledge, support and framework for deploying
  Grid technologies
• Exploitation of Grid resources both on campus and
  beyond
• A context for the University to invest in
  computing resources
• Provide test bench for existing, emerging Grid
  technologies
• Make significant contributions to general grid
  problems
• Sharing resources among multiple VOs
• Network monitoring and QoS issues for grids
• Integration of middleware with domain specific
  applications
• Grid filesystems

13
MFRID Authorization Groups
14
MGRID Job Portal
15
MGRID Job Status
1300 CPUs
16
Major MGRID Users (Snapshot)
17

• Grid3 A National Grid Infrastructure
• 32 sites, 4000 CPUs Universities 4 national
  labs
• UF is a founding member
• Sites in US, Korea, Brazil, Taiwan
• Running since October 2003

Brazil
www.ivdgl.org/grid3
18
Grid3 Applications
www.ivdgl.org/grid3/applications
19
Grid3 Shared Use Over 6 months
Usage CPUs
Sep 10, 2004
20
Grid3 ? Open Science Grid

• Iteratively extend Grid3 to Open Science Grid
• Grid3 ? OSG-0 ? OSG-1 ? OSG-2 ?
• Shared resources, benefiting broad set of
  disciplines
• Grid middleware based on Virtual Data Toolkit
  (VDT)
• Further develop OSG
• Partnerships with other sciences, universities
• Incorporation of advanced networking
• Focus on general services, operations, end-to-end
  performance
• Aim for July 20, 2005 deployment
• Major press announcement

21
http//www.opensciencegrid.org