Skip to content

WeeklyTelcon_20230523

Jump to bottom
Geoffrey Paulsen edited this page Jul 25, 2023 · 3 revisions

Open MPI Weekly Telecon ---

  • Dialup Info: (Do not post to public mailing list or public wiki)

Attendees (on Web-ex)

  • Jeff Squires (CISCO)
  • Geoff Paulsen (IBM)
  • Howard Pritchard (LANL)
  • Luke Robison (Amazon)
  • Thomas Naughton (ORNL)
  • Tommy Janjusic (nVidia)
  • Amir (ORNL)

New Issues:

v4.1

  • No new updates
  • PRs coming in as needed.
  • A release to roll out end of summer, unless someone needs something sooner.

v5.0

  • PMIX v4.2 async modex issue: https://github.com/openpmix/openpmix/issues/3077

    • Work around: -x PMIX_MCA_gds=hash
    • Need to evaluate performance impact.
    • Tommy will investigate when the issue was introduced and hopefully speed up a potential fix.

  • MAC Params issues are biggest issues now - no new updates.

  • Need to cherry-pick NIC selection (distances PR fixes) to v5.0.x

    • Several PRs will go into main, including coverity fixes.
    • Amir to open up a v5.0.x PR to track all main commits and cherry-pick to v5.0.x when finished.

  • UCX and enable mca dso do not mix issue: https://github.com/open-mpi/ompi/issues/11632

    • Fix: https://github.com/open-mpi/ompi/pull/11640
    • Fix is good as is; however, we want to investigate @hjelmn suggestion 'deregistering group as is done in mca_base_component_unload'.
    • If this is requires a lot of work, then we'll merge this PR (since it's like how other places in the code base solve the problem), and defer the real fix to #11680, which doesn't need to block v5.0.0.
    • If what @hjelmn is suggesting is easy, then let's go ahead and fold it into this PR, merge it, and cherry pick to v5.0.x in time for v5.0.0.
Clone this wiki locally