High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery1 University of Florida High Energy Physics and Grid Projects at U of Florida Dell Visit to University of Florida Dec. 13, 2002
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery2 High Energy Physics HEP data Collected in large facilities: Fermilab, Brookhaven, CERN Record collisions of opposite moving beams Each collision stored & analyzed independently 100M – 1000M collisions/year collected
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery3 Today: High Energy Physics at Fermilab CDF experiment International experiment, 600 physicists, several countries
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery4 Tomorrow: High Energy Physics at LHC “Compact” Muon Solenoid at the LHC (CERN) Smithsonian standard man
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery5 CMS Data Complexity “Events” resulting from beam-beam collisions: Signal event is obscured by 20 overlapping uninteresting collisions in same crossing (1 MB per event stored) CPU time to analyze each event rises dramatically
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery events/sec, selectivity: 1 in CMS Analysis: Higgs Decay to 4 muons
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery Physicists 150 Institutes 32 Countries LHC Computing Challenges Complexity:Millions of detector channels, complex events Scale:PetaOps (CPU), Petabytes (Data) Distribution:Global distribution of people & resources
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery8 Experiment (e.g., CMS) Global LHC Data Grid Online System CERN Computer Center > 20 TIPS USA Korea Russia UK Institute MBytes/s 2.5 Gbits/s Gbits/s 2.5 Gbits/s ~0.6 Gbits/s Tier 0 Tier 1 Tier 3 Tier 4 Tier0/( Tier1)/( Tier2) ~ 1:1:1 Tier 2 Physics cache PCs, other portals Institute Tier2 Center Florida
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery9 Florida Tier2 Center (2003) Router GEth/FEth Switch GEth Switch Data Server >1 RAID WAN “Hierarchical” switching topology Switch GEth/FEth Florida in 2003 300 CPUs >2.5 GHz P4 7 TBytes RAID
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery10 “Trillium”: US Data Grid Projects GriPhyN Grid research, toolkits $12M, 15 institutions iVDGL Deploy Global Grid lab $14M, 17 institutions PPDG Data Grid for HEP experiments $9.5M, 12 institutions Data intensive experiments Physicists + computer scientists Infrastructure development & deployment = Florida leads
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery11 Goal: PetaScale Virtual-Data Grids Virtual Data Tools Request Planning & Scheduling Tools Request Execution & Management Tools Transforms Distributed resources (code, storage, CPUs, networks) è Resource è Management è Services Resource Management Services è Security and è Policy è Services Security and Policy Services è Other Grid è Services Other Grid Services Interactive User Tools Production Team Individual Investigator Workgroups Raw data source ~1 Petaflop ~100 Petabytes
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery12 US-iVDGL Data Grid (Spring 2003) UF Wisconsin Fermilab BNL Indiana Boston U SKC Brownsville Hampton PSU J. Hopkins Caltech Tier1 Tier2 Tier3 FIU FSU Arlington Michigan LBL Oklahoma Argonne Vanderbilt UCSD/SDSC NCSA
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery13 Florida-Led US-CMS Testbed
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery14 CMS Production Simulations Remote Site 2 Master Site Remote Site 1 IMPALA mop_submitter DAGMan Condor-G GridFTP Batch Queue GridFTP Batch Queue GridFTP Remote Site N Batch Queue GridFTP Several productions in 2002 Sites in US & Europe Uncovered many Grid problems 1M events almost complete
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery15 WorldGrid Joint US - Europe effort Resources from both sides (15 sites) Use several visualization tools (Nagios, MapCenter, Ganglia) Use several monitoring tools (Ganglia, MDS, NetSaint, …) Applications CMS:CMKIN, CMSIM ATLAS:ATLSIM Submit jobs from US or EU Jobs can run on any cluster Demonstrated at IST2002 (Copenhagen) Demonstrated at SC2002 (Baltimore) Brochures available describing Grid projects
High Energy Physics and Grids at UF (Dec. 13, 2002)Paul Avery16 Collaborative Possibilities with Dell Raw computing power for production simulations 100s of CPUs, Terabytes of RAID High performance I/O Need to move ~ 5 Gbytes/sec between remote sites Bottlenecks: network protocols, clusters, components, software,… Managed clusters Goal is high efficiency use Cluster management tools, automatic operation, fewer people, … Campus Grid operations Large scale operation permits many interesting “stress” tests of cluster, network and Grid software components International Grid operations Tools for monitoring dozens of sites Automatic operation, fewer people, high throughput, …