potiuk opened a new pull request, #35746:
URL: https://github.com/apache/airflow/pull/35746
When running test with pytest-xdist, tests can be executed in random order
in multiple workers. This means that if a test leaves a side-effect, it might
affect another test.
The `recalculate_p
potiuk commented on PR #35746:
URL: https://github.com/apache/airflow/pull/35746#issuecomment-1817999114
Very insteresting finding :). It took me a while to figure out the exact
scenario where https://github.com/apache/airflow/issues/35699 could have happen
-but I believe this one perfectly
potiuk commented on PR #35746:
URL: https://github.com/apache/airflow/pull/35746#issuecomment-1818002123
Cliché, but yet another proof that `Cache invalidation is the most .`
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHu
amoghrajesh commented on code in PR #35746:
URL: https://github.com/apache/airflow/pull/35746#discussion_r1398670287
##
tests/conftest.py:
##
@@ -1075,7 +1075,14 @@ def clear_lru_cache():
from airflow.utils.entry_points import _get_grouped_entry_points
ExecutorLoader
bolkedebruin commented on PR #35746:
URL: https://github.com/apache/airflow/pull/35746#issuecomment-1818439747
Ah indeed, I was wondering how the pollution could happen at all. I wasn't
aware of pytest-xdist. It feels to me like a bug in pytest-xdist, but this is a
good fix.
--
This is a
bolkedebruin merged PR #35746:
URL: https://github.com/apache/airflow/pull/35746
--
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: commits-unsubscr...@air
potiuk commented on PR #35746:
URL: https://github.com/apache/airflow/pull/35746#issuecomment-1818721261
> Ah indeed, I was wondering how the pollution could happen at all. I wasn't
aware of pytest-xdist. It feels to me like a bug in pytest-xdist, but this is a
good fix.
Just to expl