January 2017 DMLT




  • LSST Project Management Office, Tucson - Room 27 (Tuesday/Wednesday) and Main Conference Room (Thursday)
  • Remote access via BlueJeans at ls.st/jmc or dial one of the following numbers:     +1.408.740.7256       +1.888.240.2560      +1.408.317.9253
          Connecting from a room system dial or bjn.vc and enter meeting ID  293724745 followed by # 

Hotel / Per Diem

  • Erin has booked a block of rooms in ALOFT
  • Per Diem M&IE for Tucson is $59; first/last day of travel is $44.25; no meals provided other than Wednesday night so deduct $26 from your meals on this day when doing your TER.

Planning to attend


Chance for me (Wil) to get to know you all a bit better, understand your AGILE process, and learn about the current long term Plan.

Other Topics for the agenda (please add here any points you would like to discuss)

  • To cover: Agile and long-term Roadmaps
  • To cover: ITIL and DevOps - and how they ARE compatible.
  • The Data model - separation of processing, defining interfaces intra sytem and subsystems
  • Use of Cloud Infrastructure .. further study needed ?
  • Documentation
    • Documentation tree .. if there is one - LDM-493 does not seem to tell me whihc docs we should be creating
    • Finishing and maintaining design documents (especially LDM-151): how worthwhile is effort to polish these and make them scientist-readable (and up-to-date)?
    • Discoverability of and all-team access to documentation; accessibility of mature drafts (related to but not identical to the above)

Topics of interest to T/CAMs:

  • Integration & test, Commissioning: Effort and schedule for DM verification and system verification
  • Data Challenges: What are we doing when?
  • Related: Future of PDAC?
  • Simulated data:
    • Some was generated for DESC, Euclid but we haven't seen it; is it useful for us?
    • What about list of simulated data desires sent to Sims group at the last JTM?
  • Binary distributions: answers to questions in Tim's e-mail.
  • We have risks associated with specific WBSs but currently have no interaction between T/CAMs and the people reporting on risks. Closer integration of per-WBS risks with system engineering is desired to more accurately reflect current status.
  • Pipeline QA functionality definition discussions.  Is it the precursor for commissioning support?
  • KPMs - this work is orphaned in the short term - are we still committed to doing release characterisations?


Day #1: Jan 10 - start 9:00AM (Room 27)

09:00   Introduction of Wil to the team - revisit the Good/Bad/Ugly talk from LSST Europe.

10:00   AGILE on Gaia talk (orginally for ESOC) to focus the discussion

10:45   Break

11:00   Individual/Team intros and how each do Agile

    • 15 minutes per team + 5 margin
    • what experiences you have had
    • what problems you have had
    • suggstions to improve the Agile process

11:00   Project Science Group Overview - Mario

11:20   Architecture Overview - KT/Tim

11:40   Data Release Production - John Swinbank / Jim Bosch

12:00   Alerts - Simon/Andy

12:30   LUNCH (off-site; on your own)

14:00   Data Access Services - Fritz

14:20   SUIT - Xiuqin Wu / David Ciardi

14:40   SQuaRE - Frossie Economou/ Michael Wood-Vasey

15:00   Production, Infrastructure and Ops - Margaret/Don

15:30   Break

15:45   What is our goal:  the DM mission statement ...

16:15   Moving to OPS from AGILE dev - wil

17:30   End (or earlier if we make it)

Day #2: Jan 11 (with Victor) (Room 27)

09:00   Documentation (and other technical topics)

              some impressions - WOM

              some history - FE

10:00   What the project manager expects (20min)- Victor

10:30   Overview of plan status (20min) - Mario

10:45   Break

11:05   Individual TCAM views on the plan specifically

    • 15-20 mins max
    • A single slide presenting the main milestones (10km view)  - many of which may be the dependencies of other teams
    • Overview of dependencies and timing ... Dependency spreadsheet giving all milestones between teams, synchronized with the plan shoudl be available refer to issued document
    • Rolled up overview of resources ... Detailed planning packages (in a variety of tools) with resource estimates for each team should be available refer to issued document
    • Narrative description of potential descopes involving each team with order-of-magnitude estimates for impact

11:05   Alerts - Simon Krughoff

11:30   Data Release Production - John Swinbank

11:55   SQuaRE - Frossie Economou

12:30   LUNCH (off-site; on your own)

13:45   Operations and pre-operations - Beth Willman

14:00   SUIT - Xiuqin Wu

14:25   Data Access Services - Fritz

14:50   Production, Infrastructure and Ops - Margaret

15:15   Break

15:30   Open discussion with Victor

16:00   Actions for finalizing the plan and scheduling Day #3 work (Victor/wil/Mario)

17:00   End

18:00   Gather for off-site Team Dinner (seating at 18:30): Cafe a la Cart; Directions

Day #3: Jan 12 (Main Conference Room)

09:00-12:00   Topics not covered by first two days, including (see above for full list):

09:00   Tim on some points that came up the last days 10 minutes?

09:20   Finishing LDM-151Jim Bosch  

10:00   Customers of binaries - Tim Jenness

10:45   Break

11:00   Future of DM risks and how we manage themTim Jenness

11:20   Clarification of usage of term QA (Quality Assessment vs Quality Assurance vs Quality Analysis)

12:00   Finish

13:00-17:00   Meetings with Wil and time for work determined to be urgently needed in discussions from Days #1 and #2

Short-Term Action Items

  • Prepare an overview of DMSR update: Tim Jenness
  • Prepare an overview of KPMs (what they are, where they come from, what they're used for): Tim Jenness
  • Create a stub LDM-XXX document for the QA Design/ConOps: Tim Jenness (done as LDM-522)
  • Interested parties for Thursday topics: prepare one slide for each issue
  • Write a guidance note about how to write up risks: Wil O'Mullane
  • Wil and Frossie and Mario to co-author verification and validation plan based on template: Wil O'Mullane, Frossie Economou, Mario Juric