Skip to content
Artem Polyakov edited this page Mar 15, 2019 · 29 revisions

April 23-25, 2019

Attendees:

  • Jeff Squyres, Cisco
  • Ralph Castain, Intel
  • Brian Barrett, AWS
  • Brice Goglin, Inria
  • Geoff Paulsen, IBM
  • George Bosilca, UTK
  • Shinji Sumimoto, Fujitsu
  • Harumi Kuno, HPE
  • Hunter Easterday, LANL
  • Artem Polyakov, Mellanox

Attending Remotely:

  • Josh Hursey - IBM (Available from 8:30am-5pm Central) (Added a ☎️ icon next to the items I'd like to call in for, if possible)
  • Akshay Venkatesh - NVIDIA

Agenda Items:

  • ☎️ How much MTT do we want to be running?
    • If folks aren't going to be looking regularly, perhaps once a week would be more appropriate?
    • How do we better present failure data to make analyzing results less off-putting?
    • Any way we can help others learn how to setup and run MTT?
  • ☎️ How do we get devs to set up to find and fix new MTT errors?
  • Pick 5.0 RMs
  • Topic: move more Open MPI resources to AWS
    • Move HostGator web site to AWS: Because our pre-paid HG subscription is ending, so we might as well just move it over to AWS where we already have a PHP-enabled web site running.
    • Our DNS is hosted by [email protected] Google Domains: if we move it to AWS, SSL certificates become significantly easier. Any downsides to this?
    • Do we need to send any keys or something to SPI for "ownership" of our AWS account?
  • ☎️ Transition ORTE to PRRTE
    • Gilles has a prototype - needs more work, testing
    • How do we determine when to make the switch?
  • ☎️ PMIx-based tools
    • Ralph to provide status update
  • ☎️ Resolve the "multiple OPALs" problem
    • Brian to provide plan for git submodules
  • Topic: Open MPI and containers
    • low hanging fruit for improving interoperability with, e.g. Singulariy and Charliecloud?
Clone this wiki locally