-
Notifications
You must be signed in to change notification settings - Fork 22
ASC Q4 2024 Meeting
- Date: October 15 & 17, 2024
- Time: 10 am - 1 pm US Central Daylight Time each day
- Location: Virtual Meeting. Zoom (link in the Active notes Google doc below)
- Active Notes Link: Google Doc - Please add your name and affiliation.
- Governance Document [latest]
This meeting has a floating agenda with specific synchronization points to keep us on track. Rough time estimates are provided per agenda item, and the co-chairs plan to cover the topics in the order seen below. However, since some agenda items will take longer/shorter than anticipated, an exact start/end timing is not guaranteed, and some items may float to the second day. If you cannot attend the full meeting and are presenting, please let the co-chairs know, and we can plan accordingly.
Start | End | Topic |
---|---|---|
10:00 am | 10:05 am | Gathering (--) |
10:05 am | 10:10 am | Roll Call (We will start roll call promptly at this time) |
10:10 am | 11:30 am | Discussion of agenda items |
11:30 am | 11:45 am | Voting and Break Voting Link |
11:45 am | 1:00 pm | Discussion of agenda items |
Start | End | Topic |
---|---|---|
10:00 am | 10:05 am | Gathering (--) |
10:05 am | 11:30 am | Discussion of agenda items |
11:30 am | 11:50 am | Voting and Break Voting Link |
11:50 am | 12:45 am | Administrative and Working Group agenda items |
12:45 am | 1:00 pm | Closing discussion and wrap up |
-
Administrative announcement
- Plans for meeting at SC24
- Calendar FY25
- ASC 25Q1: Jan 21, 23
- ASC 25Q2: May 6, 8
- ASC 25Q3: July 15, 17
- ASC 25Q4: Oct 14, 16
- Open PMIx ASC position:
- co-secretary: Open
- co-chair: A. Bouteiller reaching end-of-term; elections to be held in 25Q1
-
Working Groups Updates:
- IAWG (Dave Solt)
- Dynamic (Isaias Compres)
- Call for new WGs?
-
Release planning
- v5.1
-
Governance PRs up for a Second Vote:
- None
-
Governance PRs up for a Reading and First Vote:
- None
-
PMIx Standard PRs up for a Second Vote (stable):
-
PMIx Standard PRs up for a Reading and First Vote (stable):
- None
-
PMIx Standard PRs up for a Reading (Provisional):
- None
-
PMIx Standard PRs up for a Reading (Errata):
- None
-
Plenary discussion items
- Use-cases examples (David Sold, Ken Raffenetti)
- When is PMIx Fence required https://github.com/pmix/pmix-standard/issues/511
- Removal of reference implementation for some standard features https://github.com/openpmix/openpmix/issues/3409
- Other/Open discussion
Person | Institution | Day 1 |
---|---|---|
Aurelien Bouteiller | UTK | X |
Rajat Bhattarai | Tennessee Tech | X |
David Solt | IBM | X |
Ken Raffenetti | ANL | X |
Thomas Naughton | ORNL | X |
Isaias Compres | TUM | X |
Howard Pritchard | LANL | X |
- Administrative
- Meeting at SC 24
- Meet at exhibition floor (can host at UTK or DOE booth)
- Maybe coordinate with CASS software suite effort at DOE
- Need to narrow down time slots (will avoid MPI-Forum, MPICH, OMPI and Dynamic Workflows BoFs), Need to have before Thur. If there are other conflicts, please let us know.
- Meeting at SC 24
- Calendar FY25
- ASC 25Q1: Jan 21, 23
- ASC 25Q2: May 6, 8
- ASC 25Q3: July 15, 17
- ASC 25Q4: Oct 14, 16
- Note: Raise any known conflicts for these dates.
- Open PMIx ASC position:
- co-secretary: Open
- co-chair: A. Bouteiller reaching end-of-term; elections to be held in 25Q1
- Note: Please nominate/volunteer
- PMIx 5.1 Release planning
- https://github.com/pmix/pmix-standard/milestone/10
- Some value in pursuing a v5.1 to eliminate ambiguity between v4.2/v5.0 (see PR#514 changelog items for errors corrected in v5.1)
- Need to review other items and drift items (note openpmix may deprecate some things, which could eliminate some items from this “drift” list). Would be good to dedicate time to triage the drift list and identify the what should remain and what would fit into v5.1 or future.
- Note: Item for next monthly agenda
- Working Group updates
- Client Separation / Implementation Agnostic Document (Dave Solt)
- Mothballed chapter review activities (some PRs in backburner left)
- Focus on increasing engagement and easier to understand: use cases examples (more talk about it later)
- Meetings on monday 1530 EST
- Tools & Dynamic Workflows (Isaias Compres)
- Since April: group has reformed, work on PMIx has been scattered trying to form a collaborative effort
- Implementation of ambitious features for dynamic deployment (with Slurm in particular).
- Slurm impl. has proven challenging, due to some assumption with fixed allocation size, dynamicity supported only with additional allocs
- New approach is to refocus on research software development for resilience, dynamic workloads, MPI integration, scheduler. Ralph also has a scheduler (DynaSched) that is external to Open PMIx/prte and can be used for quicker turnaround.
- Allocation cancellation (no PR yet, an entry in the WG wiki) is shelved for now because of fear it may have race conditions
- First wednesday of the month 5PM German (Berlin) time
- Call for new WGs
- Client Separation / Implementation Agnostic Document (Dave Solt)
- PMIx Standard PRs up for a Reading (Errata):
- None
- PMIx Standard PRs up for a Reading and First Vote (stable):
- None
- Governance PRs up for a Second Vote
- None
- Governance PRs up for a Reading and First Vote
- None
- PMIx Standard PRs up for a Second Vote (stable)
- Discussion
- Use-cases examples (David Sold, Ken Raffenetti)
- When is PMIx Fence required https://github.com/pmix/pmix-standard/issues/511
- Need to review the tracking issues on open pmix, make a tracking issue on pmix-standard and verify that existing text does what we expect it to do, some contradictory text has been seen.
- Removal of reference implementation for some standard features https://github.com/openpmix/openpmix/issues/3409
- Other/Open discussion
- Discussion: Use-case examples
- Simple PMIx examples for tutorial or quickstart to getting code for use to help facilitate usage of large amounts of functionality in the standard. Presenting two examples today (starters). Questions: what’s right format, ways to get ideas across and any feedback/suggestions for improvement.
- Example1: Business Card Exchange
- Stand-alone example (
ipv4_bc_use_case.c
) - Working IPv4 example in ~100-200 lines
- https://gist.github.com/raffenet/9b41d212b610c18b61e6fb4dd3973acc
- Stand-alone example (
- Example2:
- Stand-alone example of a small server that can receive log from a node-local tool
- The example has been tested and works
- Only 2 functions of the server module are implemented
- There is some complication with thread locking, that could be avoided if the C++ queues were thread-safe
- https://gist.github.com/dsolt/3654060204b8e9251c2e4700bece97f5
- Any other examples we can think about that would be interesting
- Modex exchange
- Debugger tool
- LICENSE files with examples to enable outside usage
- Hosting the examples: we are thinking about readthedoc or similar, from tutorial.pmix.org, docs.pmix.org, having ci hooks publishing from source github
- Meeting cancelled - full agenda covered on Day 1
- Voting block open until end of Day 2 window, then results gathered/posted