Open MPI Weekly Telcon

---

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

Attendees

• Jeff Squyres
• Arm Patinyasakdikul
• Edgar Gabriel
• Howard Pritchard
• Nathan Hjelmn
• Ralph Castain
• Sylvain Jeaugey
• Todd Kordenbrock

---

Agenda

Review 1.10

• Milestones: https://github.com/open-mpi/ompi-release/milestones/v1.10.3
  • Two PR's pending:
    • PR 1206: MPI File close blocker
      • Need to decide if we want this
    • PR 1204: attributes
      • Decided that this is very low priority, but if it gets stable on master, we'll take this

Review 2.0.x

• Wiki: https://github.com/open-mpi/ompi/wiki/Releasev20
• Blocker Issues: https://github.com/open-mpi/ompi/issues?utf8=%E2%9C%93&q=is%3Aopen+milestone%3Av2.0.0+label%3Ablocker
• Milestones: https://github.com/open-mpi/ompi-release/milestones/v2.0.0
  • PR 1174: needs a minor tweak, then we'll put it in
  • PR 1199: request revamp
    • Nathan found 2 new issues
      • one thread adds callback while another thread is calling that callback (Nathan working on this right now -- PR probably in the next few hours)
      • PR 1729: minor thread leak in persistent communications leak / callbacks can be lost (very old bug -- dates back to 2005!). Need George PR.
  • One more PR coming about XRC fix from NAthan
  • Nathan has a 1-line uGNI fix that he'd like to get in -- will send to Howard
  • NVIDIA CUDA build failed in MTT: the fix was just merged
• Fallout from request overhaul
  • How's it look on master?
    • looking good; other than missing CM, we're turning up mostly other genuine threading bugs
    • all PMLs should be good now
    • George would like to fix a few error paths (maybe v2.0.1)
  • Is there a consolidated PR for v2.x?
    • Yes: PR 1199

Review Master MTT testing (https://mtt.open-mpi.org/)

• Reminder from last week
  • 23 pull requests on master, some since last October. Not TODAY (since we want George's Multithreaded thing in), but should bring them in or kill them.

MTT Dev status:

• .mailcap update: everyone please review https://github.com/open-mpi/ompi/pull/1581/files
• Jeff to look at Cisco MTT failures

Logistics

• MPI Forum next week
  • Will be there: Jeff, Howard, Nathan, Sylvain

Status Updates:

• Mellanox: not here
• Sandia:
  • tracking down bug in rendezvous protocol. Will roll to v2.x -- may have to wait for v2.0.1.
• Intel: Added stuff to mpirun:
  • --timeout (reminder, in case you didn't know it existed); exits with ETIMEDOUT if timeout expires (110 on Linux / OS X)
  • --report-state-on-timeout
  • --get-stack-traces
  • Added ability to launch N daemons on a node (just for ORTE scale testing; only works with rsh): MCA param ras_base_multiplier. NOT for MPI performance testing! Only for ORTE scale testing.
  • Working on PMIx event notification stuff. ULFM comes in after that.
  • Going to use OPAL MCA stuff for Warewulf rewrite.

---

Status Update Rotation

1. Mellanox, Sandia, Intel
2. LANL, Houston, IBM
3. Cisco, ORNL, UTK, NVIDIA

---

Back to 2016 WeeklyTelcon-2016