u.s. atlas computing project: budget profiles, milestones jim shank boston university physics and...
TRANSCRIPT
U.S. ATLAS Computing Project: Budget Profiles, Milestones
Jim ShankJim Shank
Boston UniversityBoston University
Physics and Computing Advisory Panel ReviewPhysics and Computing Advisory Panel Review
LBNL LBNL
14-16 Nov., 200214-16 Nov., 2002
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 2
The last 2 days…
• Workshop to prepare the 2003 NSF ITR proposalWorkshop to prepare the 2003 NSF ITR proposal• Large ITR: $15M over 5 years
• Joint ATLAS/CMS + Computing Sciences
• The Workshop was charged with identifying a critical area of computing The Workshop was charged with identifying a critical area of computing
needed for LHC that was not being funded by exsting programsneeded for LHC that was not being funded by exsting programs• Participants from a wide spectrum
• LCG project• EDG project• Trillium (GriPhyN, iVDGL, PPDG)• CMS/ATLAS communities • Some other physics experiments…
• 2 Working groups: 2 Working groups: • The BIG Picture group
• The ITR group
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 3
Categories of missing pieces
• Transition to production level grids (middleware support, error recovery, Transition to production level grids (middleware support, error recovery,
robustness, 24x7, robustness, 24x7, monitoring and system usage optimizationmonitoring and system usage optimization, , strategy strategy
and policy for resource allocationand policy for resource allocation, authentication and authorization, , authentication and authorization,
simulation of grid operationssimulation of grid operations, , tools for optimizing distributed systemstools for optimizing distributed systems))
• Globally Enabled Analysis Communities (WG2)Globally Enabled Analysis Communities (WG2)
• Enabling Global Collaboration (a medium ITR?)Enabling Global Collaboration (a medium ITR?)
Slide from WG1 Summary
(I. Gaines)
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 4
Who Fills in Missing Pieces
• Experiments (from existing budgets) (but remember gap between bare Experiments (from existing budgets) (but remember gap between bare bones funding level and originally proposed leadership funding level)bones funding level and originally proposed leadership funding level)
• LCG (from existing funding)LCG (from existing funding)• Current grid projects (both connected to HEP and more general Current grid projects (both connected to HEP and more general
projects)projects)• Near future grid projectsNear future grid projects• This large ITR: Globally Enabled Analysis CommunitiesThis large ITR: Globally Enabled Analysis Communities• Additional medium ITR (?): Enabling Global CollaborationAdditional medium ITR (?): Enabling Global Collaboration
Slide from WG1 Summary
(I. Gaines)
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 5
Globally Empowered Analysis Communities: think globally, act locally
• User Grid Interactions (Mike, Bolek, Craig, Shaowen)User Grid Interactions (Mike, Bolek, Craig, Shaowen)• data browsing tools• User Interfaces • Visualisation Tools• Education outreach• Interactive tools• Developing automation (Higher level services, AI)
• Dynamic Resource Control (Sridahara, Kaushik, John)Dynamic Resource Control (Sridahara, Kaushik, John)• Resource scheduling• job scheduling• environment control• resource auditing, priority and priv.
• Data Provenance and workflow (Mike, Rick, David Adams)Data Provenance and workflow (Mike, Rick, David Adams)• Community sharing and collaboration
• metadata management and tools (Greg and Jaideep, David Malon)metadata management and tools (Greg and Jaideep, David Malon)• specification of data sets• spec. of user analysis• queries (metadata browser)• Equivalence
• Data Management (Ian, Torre)Data Management (Ian, Torre)• storage and data management• data management optimisations (Obj. level)
Slide from WG2 Summary
(R. Cavanaugh)
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 6
Template for summarising the different Topics
• Make a strong connection to the use casesMake a strong connection to the use cases
• how it is different from existing projectshow it is different from existing projects• why it is new and revolutionary
• Status of TopicStatus of Topic
• Description of abilitiesDescription of abilities
• Generalisation beyond HEPGeneralisation beyond HEP
• How it ties into TitleHow it ties into TitleSlide from WG2 Summary
(R. Cavanaugh)
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 7
Still need...
• $15M spread over 5 years$15M spread over 5 years• scope needs to be consistent with this
• Need deliverables
• Need designated editors for textNeed designated editors for text• Rob Ed., Ruth assists
• Need list of participantsNeed list of participants
Slide from WG2 Summary
(R. Cavanaugh)
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 8
Project Core SW FTE
10.5
10
643
8
6U.S.FranceU.K.CERNItalyOtherNeeded
22%
21%
13%8%6%
17%
13% U.S.FranceU.K.CERNItalyOtherNeeded
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 9
ATLAS Subsystem/Task Matrix
Offline Offline
CoordinatorCoordinator
ReconstructionReconstruction SimulationSimulation DatabaseDatabase
ChairChair N. McCubbinN. McCubbin D. RousseauD. Rousseau A. Dell’AcquaA. Dell’Acqua D. MalonD. Malon
Inner DetectorInner Detector D. BarberisD. Barberis D. RousseauD. Rousseau F. LuehringF. Luehring S. Bentvelsen /S. Bentvelsen /
D. CalvetD. Calvet
Liquid ArgonLiquid Argon J. CollotJ. Collot S. RajagopalanS. Rajagopalan M. LeltchoukM. Leltchouk H. MaH. Ma
Tile CalorimeterTile Calorimeter A. SolodkovA. Solodkov F. MerrittF. Merritt V.TsulayaV.Tsulaya T. LeCompteT. LeCompte
MuonMuon J.ShankJ.Shank J.F. LaporteJ.F. Laporte A. RimoldiA. Rimoldi S. GoldfarbS. Goldfarb
LVL 2 Trigger/ LVL 2 Trigger/
Trigger DAQTrigger DAQ
S. GeorgeS. George S. TapproggeS. Tapprogge M. WeilersM. Weilers A. Amorim /A. Amorim /
F. TouchardF. Touchard
Event FilterEvent Filter V. VercesiV. Vercesi F. TouchardF. Touchard
Computing Steering Group members/attendees: 4 of 19 Computing Steering Group members/attendees: 4 of 19 from US (Malon, Quarrie, Shank, Wenaus)from US (Malon, Quarrie, Shank, Wenaus)
Physics Coordinator: F.Gianotti
Chief Architect: D.Quarrie
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 10
Budget Profile Overview
• What has happened since last reviewWhat has happened since last review• The Bare-Bones profile from last summer
• The construction project stretch-outThe construction project stretch-out
• The NSF M&O/Computing proposalThe NSF M&O/Computing proposal
• New profile estimateNew profile estimate
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 11
Bare Bones Budget from June 2002
Item FY 02 FY 03 FY 04 FY05 FY 06 FY 07 TotalPhysics 100 196 196 210 215 230 1147SW 2158 2430 2600 3000 3100 3200 16488Tier 1 832 1275 1701 3392 5972 10615 23787IVDGL 290 532 550 449 457 0 2278Tier 2+Dist. IT 570 1243 2513 3000 7326Project Support 0 50 300 300 300 300 1250Project Reserve 127 300 500 500 500 1927
0Total w/o iVDGL 3090 4078 5667 8645 12600 17845 51925Total with iVDGL 3380 4610 6217 9094 13057 17845 54203
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
FY 02 FY 03 FY 04 FY05 FY 06 FY 07
FY
AY
$k
iVDGL
Reserve
Support
Physics
Tier 2
Tier 1
Software
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 12
Recent BCP (approved recently)
U.S. ATLAS PROJECT BASELINE CHANGE PROPOSAL (BCP)
17) BCP Number: 60 18) BCP Title: Project Complet ion Date Extension
19) Impact on Cost Baseline:
Baseline (AY$ in
Thousands)
Proposed (AY$ in
Thousands)
AY$ Change
DOE NSF Total
102,950 60,800 163,750
102,950 60,800 163,750
0 0 0
20) Impact on Funding Profile: U.S. ATLAS NSF Project Funding Profile
Prior Years FY03 FY04 FY05 FY06 FY07 FY08 Total Approved (AY$M) 53.510 7.290 - - - - 60.800
There is no change proposed to the NSF funding profile.
U.S. ATLAS DOE Project Funding Profile Prior Years FY03 FY04 FY05 FY06 FY07 FY08 Total
Approved (AY$M) 65.935 17.416 14.690 4.909 - - 102.950 Proposed (AY$M) 65.935 17.416 8.990 5.490 3.239 1.880 102.950 Change (AY$M) 0 0 (5.700) 0.581 3.239 1.880 0
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 13
NSF M&O/Computing proposal budget
Submitted Oct. 02
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 14
Nov 2002 Profile Estimate
U.S. ATLAS Research Program EstimateHigh Level SummaryFY 03 FY 04 FY 05 FY 06 FY 07 FY 08
Description (k$) (k$) (k$) (k$) (k$) (k$)Computing 5,875 8,216 9,488 11,863 13,452 16,650M&O 2,090 5,305 9,306 10,082 11,735 9,933Upgrade R&D 0 238 859 2,263 2,210 3,234Education 50 50 50 50 50 50Project Office 415 304 1,095 1,099 1,099 1,099Management Reserve 87 1,704 2,718 3,725 4,326 4,791Total FY02$ 8,517 15,818 23,516 29,082 32,872 35,757Total AY$ 8,603 16,607 25,398 32,223 37,343 41,620DOE Guidance 3,350 9,680 13,420 21,300 22,620 22,620NSF Guidance 2,500 5,000 7,500 10,000 10,000 10,000Balance (2,753) (1,927) (4,478) (923) (4,723) (9,000)
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 15
Nov02 Profile Estimate Breakout
Item FY 03 FY 04 FY 05 FY 06 FY 07 FY 08Physics 253 200 200 200 200 200SW 4,242 4,830 5,230 5,330 5,430 5,630Tier 1 1,070 1,794 2,445 3,771 4,908 6,111IVDGL 532 550 449 457 0 0Tier 2+Dist. IT 310 892 1,613 2,562 2,914 4,709Project SupportProject ReserveTotal w/o iVDGL 5,875 7,716 9,488 11,863 13,452 16,650Total with iVDGL 6,407 8,266 9,937 12,320 13,452 16,650
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 16
Profile comparison
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
FY 03 FY 04 FY 05 FY 06 FY 07 FY 08
Nov02 Estimate
Bare Bones July 02
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 17
Software Project Developments
• Recent software progress has been driven by the ATLAS Recent software progress has been driven by the ATLAS
data challenges.data challenges.• This will be mentioned in many other talks at this review
• Some details of the ongoing and soon to be ongoing DC’sSome details of the ongoing and soon to be ongoing DC’s
• The US ATLAS TestbedThe US ATLAS Testbed
• Review of ATLAS MilestonesReview of ATLAS Milestones
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 18
ATLAS DC1 Phase 1 : July-August 2002 (A. Putzer)
1. Australia2. Austria3. Canada4. CERN5. Czech Republic6. France7. Germany8. Israel9. Italy10. Japan11. Nordic12. Russia13. Spain14. Taiwan15. UK16. USA
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 19
ATLAS DC1 Phase 2 : October-November 2002ATLAS DC1 Phase 2 : October-November 2002
• Pile-Up Production (High and Low Luminosity)Pile-Up Production (High and Low Luminosity)• About the same CPU neeed as for phase 1• 70 Tbyte• 100 000 files
• Additional countries/institutes will joinAdditional countries/institutes will join• Large scale Grid test foreseen end NovemberLarge scale Grid test foreseen end November
• As many sites involved as possible• Stability test : ~ 1-2 weeks• Test of the Worldwide Computing Model
• Next Steps (2003)Next Steps (2003)•Reconstruction (scheduled activities)•Analysis (`chaotic access‘: most demanding)
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 20
ATLAS DC2 : October 2003 - March 2004ATLAS DC2 : October 2003 - March 2004
• Use Geant4Use Geant4• Perform large scale physics analysisPerform large scale physics analysis• Use LCG common softwareUse LCG common software• Use widely Grid middlewareUse widely Grid middleware• Further test of the computing modelFurther test of the computing model• ~ same amount of data as for DC1~ same amount of data as for DC1
ATLAS DC3 : End 2004 - Begin 2005ATLAS DC3 : End 2004 - Begin 2005• 5 times more data than for DC25 times more data than for DC2
ATLAS DC4 : End 2005 - Begin 2006ATLAS DC4 : End 2005 - Begin 2006• 2 times more data than for DC32 times more data than for DC3
•
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 21
Summary Major Milestones
0 1 2 3 4 5 6 1 Tbyte database prototype (Done)Release of Athena pre-alpha version (Done)Athena alpha release (Done)Geant3 digi data available (Done)Athena beta release (Done)Athena accepted (ARC concluded) (Done)Athena Lund release (Done)Event store architecture design document (Done)DC0 production release (Done)Decide on database product (Done)DC0 Completed - continuity test (Done)Full validation of Geant4 physics DelayDC1 Completed DelayComputing TDR Finished (Align with LCG) DelayDC2 Completed (followed by annual DCs) DelayDC3 Completed (Exercise LCG-3) NewPhysics readiness report completed DelayDC4 Completed (Align with wedge test) NewFull chain in real environment (DC5) Delay
Green: Done Gray: Original date Blue: Current date
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 22
Major Milestones
One DC per year until startup
Nov 2002Nov 2002Jim ShankJim Shank , PCAP Review LBNL 23
Summary
• Great progress/success with the Data Challenges.Great progress/success with the Data Challenges.
• US ATLAS Testbed has become MUCH more functionalUS ATLAS Testbed has become MUCH more functional• Driven by DC and the SuperComputing 2002 demonstrations taking
place next week
• Interactions with CERNInteractions with CERN• ATLAS interaction with LCG needs strengthening although US is
quite active.
• International ATLAS responding positively to our pressure: SIT