ORNL / ORNL/ExaGO

Saving outputs when running sopflow with HiOp PriDec

Open
#21 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
15
Forks
9
Avg merge
4h 34m
Merged PRs (30d)
1

Description

Issue type

Bug

Issue applies to:
  • PFLOW
  • OPFLOW
  • TCOPFLOW
  • SCOPFLOW
  • SOPFLOW
  • CMake and build system
  • Spack
  • Visualization
  • Documentation
  • Other
Summary

When running sopflow at scale on Summit with HiOp PriDec on GPUs, the code would crash when attempting to write the output files for each contingency (when this was requested). It looked to me like ExaGo was trying to run multiple instances of opflow again to convert to PS format, and it was running out of GPU memory. Note that HiOp PriDec had already converged when this happened. In the end, I turned off the final output generation to collect the scaling data.

Description how to reproduce the issue

See discussion here.

ExaGO version

develop

System and environment details

See Frontier build config.

Additional information

No response

Proposed solution

No response

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the linked discussion and the Frontier build configuration to understand the SOPFLOW HiOp PriDec execution and requested per-contingency output path. Reproduce the output-generation failure on the stated environment and verify that converged runs write the requested files without exhausting GPU memory.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
hpc
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.