Child pages
  • 2014-10-01 Meeting notes
Skip to end of metadata
Go to start of metadata

Date

 at 10:30am MST/PDT

Tucson Location:  LSST Conference Room

Videocom: IP 140.252.24.8 (Direct Dial)

Call 1-866-330-1200, Participant Number 860-9352#

Attendees

  • George, Steve, Cathy, Kem, Peter, Simon, Francsico, Andy, Lynne, Srini

Agenda

New Meeting Schedule / Codes

Scheduling Workshop Update (Kem)

  • ACTION Srini has two suggestions from UA that have worked with LSST and will send to the email list

Developer Hire Update

Development planning

  • JIRA Cathy will work with Kem to update the plan spreadsheet (which will define epics in JIRA) to take into account the latest developments

Design Review (week of Oct 1)

MySQL v SQLite - two weeks Oct 15

  • JIRA Srini will create a confluence page with a proposal for how to move forward, capturing these issues, especially ones Simon brought up

AAS Posters (Abstracts due today)

JIRA - Updates, Blockers, and new Sprint

  • Login Time-out

Notes

New Meeting Schedule / Codes -

  • New meeting time for Simulations will be discussed at the usual Friday meeting (tentatively 10am Wed); then a time for OpSim can be set.  
  • We will likely use the SE code (518-9352) for OpSim for consistency. 
  • La Serena is 4 hours ahead of MST/PDT

Scheduling Workshop Update (Kem)

  • ACTION Srini has two suggestions from UA that have worked with LSST and will send to the email list 
  • November times may not work due to conflicting travel schedules; Kem will work on invitees (maybe P. Pinto) and when they might be available.
  • define objectives; desired outcome

 

-external review of greedy algorithm - they can say how to do it better

-overview of scheduling used in other areas

 

Developer Hire Update

  • Still in progress; will likely start after Nov 1

Development planning

  • JIRA Cathy will work with Kem to update the plan spreadsheet (which will define epics in JIRA) to take into account the latest developments
  • put in JIRA but will be assigned to a future sprint

Design Review

  • Scheduled for Thursday, Oct 2, 10am MST/PDT
  • Francisco will put materials on Confluence ahead of the meeting time.
  • Francisco gave an overview of the areas that will need to be addressed, including  
    • scheduler/simulator components, 
    • scheduler/simulator interfaces; 
    • organization of hardware, 
    • organiziaton of software repositories (common codes, etc); 
    • the need to synchronize with OCS development so we can structure/coordinate the work
    • How the scheduler is going to talk to Engineering Database - does that impact how we run OpSim?

MySQL v SQLite - two weeks Oct 15

  • Srini will create a confluence page with a proposal for how to move forward.
  • We will set 

AAS Posters (Abstracts due today)

  • Author list will be same as SPIE papers - let lead author know if you want to opt out.

MAF development for OpSim

  • While there is an understanding of the desire to keep MAF flexible and general, OpSim is concerned that this may limit its usefulness in the near term.  Cathy will review the latest version and write up what it will take to finalize a workable replacement for SSTAR.
  • The point was made that Simulations can allocate its resources in a way to address any issue. OpSim is concerned about the consistent availability of resources needed to extend the sstarDriver to include planned functionality such as computational comparisons of multiple runs and create enhancements in a timely manner as the need arises.  Cathy will collect and write up the list of desired enhancements so that resources can be allocated in a timely way.
  • the request for a view in MAF of metrics that evaluate whether a simulation meets the SRD will be put in JIRA and addressed on a 6 month time frame.

JIRA - Updates, Blockers, and new Sprint

  • Login Time-out - no one knows
  • slalib to palpy -
    • Srini reports it is in the quality review stage with Francisco as reviewer
    • Need to compare new branch with old branch as a test - benchmark comparison
    • create a text comparing slaib and palpy branches in accuracy and speed
    • merge branch and release by next Wed
  • machines
    • could use a third machine for simulations when MAF production is slow
    • ACTION: Cathy talk to Iain about setting up a VM on the DMZ (maf1) to prototype a distribution point as well as run MAF
    • ACTION Kem will identify runs from Run Log needed to transition to showMaf.py
  • All will close out JIRA issues by EOB today and a new sprint will be started.