villebro commented on PR #43689:
URL: https://github.com/apache/superset/pull/43689#issuecomment-5469900548
On the two failing tests:
- **`test_reap_cancels_orphaned_query_when_handle_present`** — real fixture
bug, fixed in a371d2103d. The helper wrote the engine cancel handle at
top-level `properties`, but the #43678 task-state hierarchy moved
`cancel_query_id`/`cancel_database_id` into `private["task"]` (where the
executor writes and the reaper reads them). Fixed the helper to use
`update_task_private`.
- **`test_related_subscribers_scoped_to_visible_tasks`** — I could not
reproduce this in isolation and believe it is a full-run test-isolation flake,
not a regression from this PR. The code path is correct: `TaskDAO.create_task`
auto-subscribes the creator via the passed `user_id` for all scopes (incl.
PRIVATE), so the fixture makes admin a subscriber of its 5 tasks; admin (sees
all via `TaskFilter`) should therefore appear in `/related/subscribers`.
Nothing in this PR touches `related_subscribers`, `visible_task_ids_query`,
`get_subscriber_principals`, or `create_task`. The failure (`{1,2} <= {2}`:
admin missing, gamma present) is consistent with shared-session/cross-test
pollution in the full integration run rather than a deterministic logic error.
I can dig into the ordering/pollution source if it reproduces in a re-run, but
I did not want to blind-edit a passing-in-isolation test and risk masking a
real issue — happy to if you would prefer I harden the fixture.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
