Prev Next

SW&IT Meeting Minutes 2021-11-15

Attendees: Brandon, David, Glenn, Matthieu, Paul, Stephen, Xianyu, Yang

Outlook

Labor Reporting

Report auto-generated every Monday at 6:30am, Do not edit this table, Update your entries in smartsheets.

20211101
LR20211101 - Bec, Matthieu 72.0
LR20211101 - Mechtley, Brandon 48.0
LR20211101 - Eychaner, Glenn 40.0
LR20211101 - Edmeades, David 40.0
LR20211101 - Grenz, Paul 40.0
LR20211101 - Hooper, Stephen 0.0
LR20211101 - Zhang, Yang 72.0
LR20211101 - Zhang, Xianyu 72.0

News from management

Previous Action Items

Owner Action Due Date Completion Date

Open topics

Team reports

MB

  • Last week
    • OC, Management
    • science archive meeting
    • learning some more about AO s/w, join discussions and scheming on #ao-sw
    • IssueTrak compat IMS
    • PB s/w discussions
  • This week
    • Performance Appraisals
    • git/github "policy" (= philosophy slides) for new staff
    • LBTplot conversion to DMS (ctd.)
    • PB s/w discussions (ctd)
    • logrotate mongodb

DE

  • MS Server media and license, finally!
  • Backup server ready to get Veeam installed
  • Synology gear ordered, dellivery to Steward
  • Re-replacement SSD ordered for current array, long shipping lead (part of the reason I got a different model in the first place)
  • Finally got some attention to the Google ticket with JP, she is in conversation with them on what she needs.
  • Conversation with IssueTrak re: new versions breaking the screen scraper; they offered their preview site.
This week:
  • Build Windows Server VM, verify keys, etc.
  • Veeam install and setup

YZ

  • Previous:
    • Argos swingarm and OSS test/debug with Paul and Gustavo (with some new problems).
    • Science archive meeting (discussed MODS and general utility to be run on archive machine, everyone seemed to agree on what we want to do).
    • Some TMS discussion with Andrew and Heejoo.
    • LBC checkup on obs5/6 with Olga (some problems with IRAF/lbcrangebal, and it would be nice to fix some TCSSendWavefront warnings).
    • Started to work on a wiki page about LBC bias level adjustment.
    • Safety trainings.
  • Next:
    • Continue to support Argos swingarm test/development.
    • Check AOS patch is activated.
    • Discuss with Olga again about IRAF/lbcrangebal on obs5/6, and MODS WCS status.
    • TMS meeting and some work.
    • Try to finish LBC bias level adjustment wiki page.

PSG

  • Last week
    • Previously determined that commands to white light source (WLS) controller were being interleaved with other commands.
    • Had discussion with Gustavo about WLS and how to fix problem.
    • Cannot selectively run one process or the other, so a proxy is called for.
    • Found Python proxy code on github which will get us 90% of the way there.
    • Modified code to accept multiple client connections and forward each to one WLS source controller conection.
  • This week
    • Testing proxy.
    • Debugging proxy.
    • Use rc.local or systemd to start on bootup?

GE

  • Last week
    • Digging through the 550,000 lines of code of the LUCI system.
    • Tested (in simulation) a test version of LUCI software that fixes the image-cube issue and updates Aladin to version 11. Found and fixed some issues.
    • An aside on the RTD and threading. Aladin provides apparently concurrency-safe interfaces (VOObserver and VOApp) for an external program to send commands to Aladin and listen for events. For various reasons, this was deemed insufficient, and the RTD was given access to the internal Aladin data display model (the stack of displayed image planes). Unfortunately, rather than containing this access to the AladinControl class, a set of accessors for the Aladin internal data model was created in AladinControl, and accessed from the individual GUI elements in other classes, without regard for which thread was making the access (sometimes the EventThread, sometimes a SwingWorker thread, sometimes a RMI thread). The internal Aladin data model is not concurrency-safe. When this caused various threading issues (stalled EventThread and/or deadlocks), Aladin's synchronization was altered to solve the problem, making it even less concurrency-safe. I am gradually unwinding this and confining access to a single Executor thread to provide concurrency safety, but it's not trivial.
  • This week
    • Digging through the 550,000 lines of code of the LUCI system.
    • Test the new LUCI software live during engineering.
    • Continue working on the RTD issues.
    • Examine the other LUCI GUIs to check for GUI modifications from threads other than the Event Dispatch Thread.

BM

  • Last week
    • Power backplane discussion, started looking through code
    • AO code deployment plan: set up paths, ready to change over when safe.
    • Housekeeper partial timeout PR, can discuss with Doug/Juan Carlos/others about a testing strategy when they're ready.
  • This week
    • PB upgrade:
      • Look over code this week after meeting this morning. Start planning out "Phase I."
    • Code deployment:
      • List of "necessary" ao-confcalib files before setting up auto-commit.
    • LBTI issues:
      • Issues with new Pause/Resume with safe skips--deemed high priority in SOUL commissioning meeting this morning.
      • Look into logs from night ... understand when diagnostic dumps were cleared with respect to stopping/starting adsec.
      • 2x PauseAo clears gains
      • ResumeAo after LoadGainTag overwrites manual gains

SH

  • LBT orders (quotes for Cisco?)
  • Workshop Wednesday/Thursday
  • Ignition (SCADA) with Dan R.
  • Ubuntu burn, Mark S.
  • obs5 and obs6 statification
  • domain questions with Jenny et al

next:

  • Cisco quotes!
  • PEPSI things
  • SSL certs
  • Workshop Wednesday, Thursday.
Topic revision: r14 - 15 Nov 2021, PaulGrenz
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback