Skip to content

[JUnit] Preserve reports on job submission failure - #1091

Draft
podkidyshev wants to merge 5 commits into
ipod/fix-final-expfrom
ipod/junit-submit-fail
Draft

podkidyshev wants to merge 5 commits into
ipod/fix-final-expfrom
ipod/junit-submit-fail

Conversation

@podkidyshev

@podkidyshev podkidyshev commented Oct 7, 2026 •

Copy link
Copy Markdown
Contributor

Summary

Generate enabled reports before aborting on job submission failure. Persist the exception type and message in an optional execution_error field in test-run.toml, so JUnit distinguishes CloudAI errors from workload failures and retains that distinction when regenerating reports. Unattempted iterations and DSE steps remain absent.

Test Plan

Verified on macOS with Python 3.14.3 using mocked submissions and offline report generation; no benchmark or remote jobs were run.

  • uv run --locked --extra dev pytest tests/test_base_runner.py tests/test_cloudaigym.py tests/test_handlers.py tests/test_reporter.py tests/test_cli.py tests/test_api.py tests/test_test_scenario.py — 265 passed.
  • uv run --locked --extra dev pytest tests/test_output.py tests/test_single_sbatch_runner.py tests/systems/slurm/test_runner.py tests/systems/standalone/test_runner.py — 47 passed.
  • uv run --locked --extra dev pre-commit run --files <changed-files> — all applicable checks passed, including pyright, Ruff, and import-linter.
  • Validated the uploaded incident evidence offline: after adding an execution-error record reconstructed from the log, JUnit reports 11 successes, 1 error, and 0 skipped cases, including JobIdRetrievalError and the scheduler error message. Regeneration produced identical results. Original run metadata was preserved and the other evidence files were unchanged.

Additional Notes

Depends on #1065. Targets ipod/fix-final-exp, which owns output finalization and report generation; this PR adds submission-error propagation and JUnit classification. Retarget to main after #1065 merges.

@coderabbitai

coderabbitai Bot commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

Important

Draft PR not reviewed

Draft PRs are not automatically reviewed by default.

  • Trigger a manual review

To automatically review draft PRs, update your CodeRabbit configuration:

reviews:
  auto_review:
    drafts: true
  • Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Comment @coderabbitai help to get the list of available commands.

@podkidyshev
podkidyshev force-pushed the ipod/junit-submit-fail branch from f15155f to 3b86492 Compare October 7, 2026 16:51
Signed-off-by: Ivan Podkidyshev <ipodkidyshev@nvidia.com>
@podkidyshev
podkidyshev force-pushed the ipod/junit-submit-fail branch from 3b86492 to d23cf51 Compare October 7, 2026 18:23
@podkidyshev
podkidyshev changed the base branch from main to ipod/fix-final-exp October 7, 2026 18:23
Signed-off-by: Ivan Podkidyshev <ipodkidyshev@nvidia.com>
Signed-off-by: Ivan Podkidyshev <ipodkidyshev@nvidia.com>
Signed-off-by: Ivan Podkidyshev <ipodkidyshev@nvidia.com>
Signed-off-by: Ivan Podkidyshev <ipodkidyshev@nvidia.com>

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant