Skip to content

Test loadscope recovery after worker failure - #1381

Open
GruffElixir wants to merge 1 commit into
pytest-dev:masterfrom
GruffElixir:fix/1378-completed-scopes
Open

GruffElixir wants to merge 1 commit into
pytest-dev:masterfrom
GruffElixir:fix/1378-completed-scopes

Conversation

@GruffElixir

Copy link
Copy Markdown

Fixes #1378

Add a regression test for loadscope scheduling after a worker crash. Tests: python -m pytest testing/test_dsession.py -q

Copilot AI lite review requested due to automatic review settings September 8, 2026 12:12

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@GruffElixir

Copy link
Copy Markdown
Author

The new LoadScope regression test passes in the matrix and locally. The only failures are the two pytest-main jobs, where existing LoadGroupScheduling acceptance tests fail against the current pytest development build; all pytestlatest jobs and the other matrix jobs pass. This appears to be an upstream compatibility issue rather than a failure in the added regression test. Could a maintainer confirm whether the pytest-main jobs should be rerun or treated separately?

@GruffElixir

Copy link
Copy Markdown
Author

Rebased onto master and dropped the obsolete grouped-node-ID workaround now that upstream passes group names separately. The regression test and changelog are unchanged. Validation: 49 scheduler/worker tests and both loadgroup integration tests passed, along with all pre-commit hooks.

@GruffElixir
GruffElixir force-pushed the fix/1378-completed-scopes branch from af705e7 to 072447c Compare October 7, 2026 11:35

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

Projects

None yet

Development

Successfully merging this pull request may close these issues.

loadscope/loadgroup: a crashed worker's COMPLETED scopes are re-queued as empty work units and can deadlock the session at the end of the run

2 participants