Enabling e-Research over GridPP Dan Tovey University of Sheffield.
-
Upload
valentina-gentleman -
Category
Documents
-
view
214 -
download
2
Transcript of Enabling e-Research over GridPP Dan Tovey University of Sheffield.
Enabling e-Research over GridPP
Dan ToveyUniversity of Sheffield
28th March 2006 University of Sheffield2
ATLAS
• Large Hadron Collider (LHC) under construction at CERN in Geneva.
• When commences operation in 2007 will be the world’s highest energy collider.
• Sheffield key member of ATLAS collaboration building one of two General Purpose Detectors on LHC ring.
• Main motivations for building LHC and ATLAS:– Finding the Higgs boson– Finding evidence for
Supersymmetry – believed to be next great discovery / layer in our understanding of the universe.
28th March 2006 University of Sheffield3
ATLAS @ Sheffield
• Sheffield leads Supersymmetry (SUSY) searches at ATLAS
• Also coordinates all ATLAS physics activities in the UK including Higgs and SUSY searches.
• Sheffield responsible for building ATLAS Semiconductor Tracker (SCT) detector, and writing event reconstruction software.
SUSY (= Nobel Prize)SM
NB: This is a simulation!
28th March 2006 University of Sheffield4
ConstructionConstruction
28th March 2006 University of Sheffield5
Event Selection
LEVEL-1 Trigger Hardwired processors (ASIC, FPGA) Pipelined massive parallel
HIGH LEVEL Triggers Farms of
processors
10-9 10-6 10-3 10-0 103
25ns 3µs hour yearms
Reconstruction&ANALYSIS TIER0/1/2
Centers
ON-lineOFF-line
sec
Giga Tera Petabit
9 or
ders
of
mag
nitu
de
28th March 2006 University of Sheffield6
The Data Deluge
Understand/interpret data via numerically intensive simulations• e.g. 1 SUSY event (ATLAS Monte Carlo Simulation) =
20 mins/3.5 MB on 1 GHz PIII
16 Million channels
100 kHzLEVEL-1 TRIGGER
1 MegaByte EVENT DATA
200 GigaByte BUFFERS500 Readout memories
3 Gigacell buffers
500 Gigabit/s
Gigabit/s SERVICE LAN PetaByte ARCHIVE
Energy Tracks
Networks
1 Terabit/s(50000 DATA CHANNELS)
20 TeraIPS
EVENT BUILDER
EVENT FILTER
40 MHzCOLLISION RATE
Charge Time Pattern
Detectors
Grid Computing Service 300 TeraIPS
• Many events– ~109 events/experiment/year– >~1 MB/event raw data– several passes required
Worldwide LHC computing requirement (2007):– 100 Million SPECint2000
(=100,000 of today’s fastest processors)
– 12-14 PetaBytes of data per year (=100,000 of today’s highest capacity HDD).
28th March 2006 University of Sheffield7
LCG
• Aim to use Grid techniques to solve this problem• CERN LHC Computing Grid (LCG) project coordinating activities
in Europe.• Similar projects in US (Grid3/OSG) and Nordic countries
(NorduGrid).• LCG prototype went live in September 2003 in 12 countries
including UK. • Extensively tested by the LHC experiments
28th March 2006 University of Sheffield8
What is GridPP?
• 19 UK Universities, CCLRC (RAL & Daresbury) and CERN
• Funded by the Particle Physics and Astronomy Research Council (PPARC)
• GridPP1 - 2001-2004 £17m "From Web to Grid"
• GridPP2 - 2004-2007 £16m "From Prototype to Production"
• UK contribution to LCG.
28th March 2006 University of Sheffield9
UK Core e-Science
Programme
Institutes
Tier-2 Centres
CERNLCG
EGEE
GridPP
GridPP in Context
Tier-1/A
Middleware, Security,
Networking
Experiments
GridSupportCentre
Not to scale!
Apps Dev
AppsInt
GridPP
28th March 2006 University of Sheffield10
ARDA
Ex
pm
tsEG
EE LCG
Deployment Board
Tier1/Tier2,Testbeds,
Rollout
Servicespecification& provision
User Board
Requirements
ApplicationDevelopment
Userfeedback
Metadata
Workload
Network
Security
Info. Mon.
PMB
CB
Storage
28th March 2006 University of Sheffield11
Tier Structure
Tier-1Tier-1
Tier-0 (CERN)
Tier-1 (Lyon) Tier-1 (BNL)Tier-1 (RAL)
NorthGrid SouthGrid
ScotGrid ULGridTier-2
28th March 2006 University of Sheffield12
UK Tier-1/A Centre Rutherford Appleton
Laboratory• High quality data services• National and international
role• UK focus for international
Grid development1400 CPU80 TB Disk60 TB Tape
(Capacity 1PB)
Grid Resource Discovery Time = 8 Hours
2004 CPU Utilisation
28th March 2006 University of Sheffield13
UK Tier-2 Centres
ScotGridDurham, Edinburgh, Glasgow NorthGridDaresbury, Lancaster, Liverpool,Manchester, Sheffield (WRG)
SouthGridBirmingham, Bristol, Cambridge,Oxford, RAL PPD, Warwick
LondonGridBrunel, Imperial, QMUL, RHUL, UCL
28th March 2006 University of Sheffield14
NorthGrid
• Tier-2 collaboration between Sheffield (WRG), Lancaster, Liverpool, Manchester and Daresbury Lab.
28th March 2006 University of Sheffield15
WRG & NorthGrid
• White Rose Grid contributing to NorthGrid and GridPP with new SRIF2 funded machine at Sheffield (Iceberg).
• LCG component to Iceberg provides a base of 230kSI2k and on demand up to 340kSI2k, with state-of-the-art 2.4 GHz Opteron cpus.
• Delivered 2nd highest GridPP Tier-2 throughput for ATLAS in 2005.
http://lcg.shef.ac.uk/ganglia
28th March 2006 University of Sheffield16
GridPP Deployment Status
Three Grids on Global scale in HEP (similar functionality)
sites CPUs• LCG (GridPP) 228 (19) 17820 (3500)• Grid3 [USA] 29 2800• NorduGrid 30 3200
• GridPP deployment is part of LCG
• Currently the largest Grid in the world
28th March 2006 University of Sheffield17
ATLAS Data Challenges
ATLAS DC2 - LCG - September 71%
2%
0%
1%
2%
14%
3%
1%
3%
9%
8%
3%2%5%1%4%
1%
1%
3%
0%
1%
1%
4%1%
0%
12%
0%
1%
1%
2%
10%
1% 4%
at.uibk
ca.triumf
ca.ualberta
ca.umontreal
ca.utoronto
ch.cern
cz.golias
cz.skurut
de.fzk
es.ifae
es.ific
es.uam
fr.in2p3
it.infn.cnaf
it.infn.lnl
it.infn.mi
it.infn.na
it.infn.na
it.infn.roma
it.infn.to
it.infn.lnf
jp.icepp
nl.nikhef
pl.zeus
ru.msu
tw.sinica
uk.bham
uk.ic
uk.lancs
uk.man
uk.rl
ATLAS DC2 - CPU usage
LCG41%
NorduGrid30%
Grid329%
LCG
NorduGrid
Grid3
Total:
~ 1350 kSI2k.months~ 95000 jobs~ 7.7 Million events fully simulated (Geant4)~ 22 TB
• DC2 (2005): 7.7 M GEANT4 events and 22 TB• DC3/CSC (2006): > 20M G4 events• UK ~20% of LCG
• Ongoing..• (3) Grid
Production• Largest
total computing requirement
• Small fraction of what ATLAS needs..Now in Grid Production Phase
LCG now reliably used for production
28th March 2006 University of Sheffield18
Further Info
http://www.gridpp.ac.uk