Joint US ATLAS Software and Computing / Physics Support Meeting

US/Pacific
Lawrence Berkeley National Laboratory

Lawrence Berkeley National Laboratory

1 Cyclotron Road Berkeley, CA 94720
Erich Ward Varnes (University of Arizona (US)), Jason Nielsen (University of California,Santa Cruz (US)), Mark Neubauer (Univ. Illinois at Urbana-Champaign (US)), Torre Wenaus (Brookhaven National Laboratory (US))
Description

This meeting combines what in the past have been three separate meetings:

  • The regular series of facilities/distributed software F2F meetings
  • The annual US ATLAS SW/PS planning meeting
  • The annual scrubbing of our activities by US ATLAS management

US ATLAS developers and technical experts are being invited to the meeting so that it can be a substantive technical discussion and planning session for key US activities. Towards making it substantive and technical, the topics will be limited.

The meeting will be F2F only, no Vidyo.

Meeting Etherpad
Meeting Rooms Map
Participants
  • Alexandr Zaytsev
  • Alexei Klimentov
  • Danila Oleynik
  • David Lesny
  • David Malon
  • Erich Ward Varnes
  • Jose Caballero Bejar
  • Lincoln Bryant
  • Mark Neubauer
  • Mark Sosebee
  • Michael Ernst
  • Nils Erik Krumnack
  • Nurcan Ozturk
  • Paolo Calafiura
  • Patrick McGuigan
  • Paul Nilsson
  • Peter Van Gemmeren
  • Richard Philip Mount
  • Robert William Gardner Jr
  • Saul Youssef
  • Scott Snyder
  • Shigeki Misawa
  • Srini Rajagopalan
  • Tadashi Maeno
  • Taylor Childers
  • Vakho Tsulaia
  • Wei Yang
  • Xin Zhao
    • 08:50 09:00
      Welcome
      Conveners: Paolo Calafiura (Lawrence Berkeley National Lab. (US)), Dr Torre Wenaus (Brookhaven National Laboratory (US))
      slides
    • 09:00 21:00
      xAOD and ROOT analysis ecosystem 2-100B-CR

      2-100B-CR

      Lawrence Berkeley National Laboratory

      EventLoop, ‘pROOT’-like functionality, xAOD analysis environment/framework, systematics framework, etc

      Conveners: Dr David Malon (Argonne National Laboratory (US)), Paolo Calafiura (Lawrence Berkeley National Lab. (US))
      • 09:00
        Session Introduction 10m
        Speaker: Paolo Calafiura (Lawrence Berkeley National Lab. (US))
        Slides
      • 09:10
        What is pROOT? 20m
        Speaker: Doug Benjamin (Duke University (US))
        Slides
      • 09:30
        pROOT use cases for Run II (Discussion) 20m
        Speaker: Zachary Louis Marshall (Lawrence Berkeley National Lab. (US))
        Slides
      • 09:50
        xAOD persistence considerations: size, versioning, schema evolution 20m
        Speaker: Peter Van Gemmeren
        Slides
      • 10:10
        metadata and metadata tools for analysis 20m
        Speaker: Dr Jack Cranshaw (Argonne National Laboratory (US))
        Slides
      • 10:30
        Coffee Break 20m
      • 10:50
        Design issues arising from CP tools migration 20m
        Speakers: Steven Andrew Farrell (University of California-Unknown-Unknown), Steven Andrew Farrell (Lawrence Berkeley National Lab. (US))
        Slides
      • 11:10
        Supertools for systematics and analysis steering (Discussion) 30m
        Speaker: Nils Erik Krumnack (Iowa State University (US))
        Slides
      • 11:40
        Data Access and Dual-use Tools (Discussion) 20m
        Speaker: Paolo Calafiura (Lawrence Berkeley National Lab. (US))
        Slides
      • 12:00
        Wrap-up discussion: US priorities in analysis support, discussion subjects for afternoon splinter session 30m
        Speaker: Dr David Malon (Argonne National Laboratory (US))
        Slides
    • 12:30 13:30
      Lunch 54 (Cafeteria)

      54 (Cafeteria)

      Lawrence Berkeley National Laboratory

      1 Cyclotron Road Berkeley, CA 94720
    • 13:30 17:30
      Splinter 6-1105

      6-1105

      Lawrence Berkeley National Laboratory

    • 13:30 17:30
      Supporting US analysis 2-100B-CR

      2-100B-CR

      Lawrence Berkeley National Laboratory

      This meeting comes shortly after the US ATLAS annual meeting where the Tier 3 white paper will be presented and discussed, so LBNL will be a good time to discuss moving forward to a focus on implementation. Topics can encompass things like data access for Tier 3s, physics support experience with FAX, ATLAS Connect, ... selectively, to make good use of 1/2 day

      Conveners: Erich Ward Varnes (University of Arizona (US)), Jason Nielsen (University of California,Santa Cruz (US))
      ATLAS Connect doc
      • 13:30
        Storage of output from jobs that run remotely 30m
        Speaker: Doug Benjamin (Duke University (US))
        Slides
      • 14:00
        ATLAS Connect from Tier3 30m
        Speaker: Robert William Gardner Jr (University of Chicago (US))
        Slides
      • 14:30
        FAX tests 20m
        Speakers: Erich Ward Varnes (University of Arizona (US)), Fred Luehring (Indiana University (US))
        Slides
      • 14:50
        Remote analysis test package 20m
        Speaker: Erich Ward Varnes (University of Arizona (US))
        Slides
      • 15:10
        Break 30m
      • 15:40
        ATLAS Infrastructure: LS1 Upgrade Challenges 30m
        Speaker: Dr Alexander Undrus (Brookhaven National Laboratory (US))
        Slides
      • 16:10
        PROOF/PROOF-on-demand within new Tier3 model 20m
        Speaker: Shuwei Ye (Brookhaven National Laboratory (US))
        Slides
      • 16:30
        Discussion 1h
    • 19:00 21:00
      Dinner

      http://greatchinaberkeley.com

      I heard they do catering and host dinners for the Chinese Consulate in San Francisco. Enough said!

    • 08:59 12:30
      Event Service 70A-3377

      70A-3377

      Lawrence Berkeley National Laboratory

      1 Cyclotron Road Berkeley, CA 94720
      Conveners: Dr Torre Wenaus (Brookhaven National Laboratory (US)), Vakho Tsulaia (Lawrence Berkeley National Lab. (US))
      • 09:00
        PART 1: Status survey with emphasis on outstanding issues and near/midterm plans (presentation/discussion ~50/50) 1m
      • 09:01
        Intro, plans, issues and objectives for today 14m
        Speaker: Dr Torre Wenaus (Brookhaven National Laboratory (US))
        Slides
      • 09:15
        PanDA/JEDI’s support for fine grained dynamic job definition, event-level bookkeeping, and the event service 15m
        Speaker: Tadashi Maeno (Brookhaven National Laboratory (US))
        Slides
      • 09:30
        Event service in the pilot: refactoring, payload integration & interaction, input/output management, object store usage 25m
        Speakers: Paul Nilsson (Brookhaven National Laboratory (US)), Wen Guan (University of Wisconsin (US))
        Slides
      • 09:55
        athenaMP payload, event I/O, event index integration 25m
        Speakers: Dr Peter Van Gemmeren (Argonne National Laboratory (US)), Vakho Tsulaia (Lawrence Berkeley National Lab. (US))
        Slides
      • 10:20
        Break 15m
      • 10:35
        PART 2: Next steps for integrating, operating and scaling the ES in the early deployment targets (mainly discussion, guided by a few slides) 1m
      • 10:36
        ES issues, plan and timeline: pre-production on the grid 14m
        Speakers: Paul Nilsson (Brookhaven National Laboratory (US)), Wen Guan (University of Wisconsin (US))
        Slides
      • 10:50
        ES issues, plan and timeline: operating and scaling up on the Amazon spot market 30m
        Speakers: Michael Ernst, Dr Torre Wenaus (Brookhaven National Laboratory (US))
        Slides
      • 11:20
        ES issues, plan and timeline: event service on HPCs 1h
        Status, plans, technical discussion, deployment targets, timeline for the ‘Yoda’ MPI scheme
        Speaker: Vakho Tsulaia (Lawrence Berkeley National Lab. (US))
        Slides
      • 12:20
        SC14 demo objectives and plans, in light of today’s discussion 10m
        Speaker: Dr Torre Wenaus (Brookhaven National Laboratory (US))
    • 12:30 13:30
      Lunch 54 (Cafeteria)

      54 (Cafeteria)

      Lawrence Berkeley National Laboratory

      1 Cyclotron Road Berkeley, CA 94720
    • 13:25 17:10
      High Performance Computing (HPC) 70A-3377

      70A-3377

      Lawrence Berkeley National Laboratory

      1 Cyclotron Road Berkeley, CA 94720

      Including common approach to PanDA integration and event service integration, athena porting, software access, ...

      Conveners: Dr Alexei Klimentov (Brookhaven National Laboratory (US)), Kaushik De (University of Texas at Arlington (US)), Thomas Le Compte (Argonne National Laboratory (US))
      • 13:30
        Introduction 5m
        Speakers: Dr Alexei Klimentov (Brookhaven National Laboratory (US)), Kaushik De (University of Texas at Arlington (US)), Thomas Le Compte (Argonne National Laboratory (US))
        Slides
      • 13:35
        An overview and HPC strategy 30m
        Speaker: Thomas Le Compte (Argonne National Laboratory (US))
        Slides
      • 14:05
        Activities Highlights 1h
        • ALCF Mira 15m
          Speakers: Doug Benjamin (Duke University (US)), Taylor Childers (Argonne National Laboratory (US))
          Slides
        • Stampede - ATLAS Software access options & issues 15m
          Speakers: Peter Onyisi (University of Texas (US)), Robert William Gardner Jr (University of Chicago (US))
          Slides
        • Blue Waters plans 5m
          Speaker: Mark Neubauer (Univ. Illinois at Urbana-Champaign (US))
          Slides
      • 15:05
        Tea break 15m
      • 15:20
        Technical Discussion 1h 15m
        • HPC pilot 15m
          Speakers: Danila Oleynik (Joint Inst. for Nuclear Research (RU)), Paul Nilsson (Brookhaven National Laboratory (US))
          Slides
        • HPC data management 15m
          Speaker: Doug Benjamin (Duke University (US))
          Slides
        • ATLAS SW @HPC 10m
          Speakers: Doug Benjamin (Duke University (US)), Peter Onyisi (University of Texas (US))
          Slides
        • PanDA@HPC 15m
          Speaker: Kaushik De (University of Texas at Arlington (US))
          Slides
      • 16:35
        wrap up and next steps 30m
    • 13:30 17:30
      Splinter 6-2202

      6-2202

      Lawrence Berkeley National Laboratory

      1 Cyclotron Road Berkeley, CA 94720
    • 17:30 19:30
      Splinter - ES/HPC Yoda technical discussion

      Develop the technical specifics and workplan for the MPI based Yoda concept for event service HPC integration. Plan a timeline consistent with demoing some form of Yoda at SC14. Same room as HPC session. Duration: until we've had enough.

      Conveners: Dr Torre Wenaus (Brookhaven National Laboratory (US)), Vakho Tsulaia (Lawrence Berkeley National Lab. (US))
    • 09:00 12:30
      US ATLAS Facilities readiness for Run 2

      Processing, storage management, (federated) data access, facility operations optimization, ...

      Conveners: Michael Ernst (Brookhaven National Laboratory (US)), Robert William Gardner Jr (University of Chicago (US))
      • 09:00
        Introduction, goals for the meeting 10m
        Speakers: Michael Ernst, Robert William Gardner Jr (University of Chicago (US))
      • 09:10
        Priorities, adjustments for the new analysis model, derivation framework, analysis support requirements: discussion based on Wednesday & Thursday meetings 40m
      • 09:50
        Tier3 Implementation Committee recommendations: recap and summary, and facility impacts 20m
        Speaker: Mark Neubauer (Univ. Illinois at Urbana-Champaign (US))
        Slides
      • 10:10
        Flocking from Tier3 for analysis batch, ATLAS Connect to campus clusters, etc. 20m
        Speaker: Robert William Gardner Jr (University of Chicago (US))
      • 10:30
        Break 20m
      • 10:50
        WAN access via FAX by Panda 15m
        Speaker: Ilija Vukotic (University of Chicago (US))
        Slides
      • 11:05
        Autopyfactory performance and configuration 15m
        Speaker: Dr Jose Caballero Bejar (Brookhaven National Laboratory (US))
        Slides
      • 11:20
        HTCondor-CE production readiness 15m
        Speaker: Mr Xin Zhao (Brookhaven National Laboratory (US))
        Slides
      • 11:35
        Dynamic multicore provisioning issues 15m
        Speaker: William Strecker-Kellogg (BNL)
        Slides
      • 11:50
        Managing data in the facility (Group, User, LocalGroup, etc); deletions, subscriptions, etc., with focus on user capabilities 20m
      • 12:10
        Prioritizing access to CPU and storage resources for US physicists 20m
    • 12:30 13:30
      Lunch 54 (Cafeteria)

      54 (Cafeteria)

      Lawrence Berkeley National Laboratory

      1 Cyclotron Road Berkeley, CA 94720
    • 13:30 17:30
      Splinter 50B-4205

      50B-4205

      Lawrence Berkeley National Laboratory

      1 Cyclotron Road Berkeley, CA 94720
    • 13:30 17:30
      US ATLAS S&C/PS Management Scrubbing 50A-5132 Sessler

      50A-5132 Sessler

      Lawrence Berkeley National Laboratory

      This will be a closed session attended by S&C/PS and US ATLAS operations program management to examine the S&C/PS program, plan and budget, towards finalizing the FY15 program and budget.

      Conveners: Jim Cochran (Iowa State Univ.), Srini Rajagopalan (Brookhaven National Laboratory (US))