Revert "Remove remaining Airflow 2.5 backcompat code from GCS Task Ha…

[potiuk]

…ndler (#36443)"

This reverts commit 75faf11.

---

^ Add meaningful description above
Read the Pull Request Guidelines for more information.
In case of fundamental code changes, an Airflow Improvement Proposal (AIP) is needed.
In case of a new dependency, check compliance with the ASF 3rd Party License Policy.
In case of backwards incompatible changes please leave a note in a newsfragment file, named {pr_number}.significant.rst or {issue_number}.significant.rst, in newsfragments.

[potiuk]

Seems that after merging that limited PR - the error started to appear again. Let's see what happens in the PR.

[hussein-awala]

There is nothing related to asyncio in this change 🤔

[potiuk]

I know :(

[potiuk]

So far It somehow looks liks this failure only happens in main (I have not seen it in any PR) - so maybe there is a difference that I will be able to track if I will get enough evidences ...

[Taragolis]

I think this all errors refers to the old issue with this test, previous attempt before we migrate everything to the pytest:

• Test only appropriate logger in Google provider #27105
• Filter logs for specific logger in TestGetGcpCredentialsAndProjectId.test_disable_logging #26973
• Fix flakey test test_disable_logging #26871

[potiuk]

Also reverting it and running few times main merge without seeing the error after seeing it in few other builds before merge, might give more clues. I plan to merge a few more PRs and see if main builds will have consistent failures before/after merging this one. It's a bit slow and "try/error" but for now I am a bit lost to what causes it and have to revert to the merge-> revert see what happens :D

[potiuk]

I think this all errors refers to the old issue with this test, previous attempt before we migrate everything to the pytest:

🤔

[Taragolis]

Maybe we could return to previous solution, when we manually filtered log records by a logger name instead of provide it into the caplog.at_level.

More interesting why this error appear again 🤔

[potiuk]

More interesting why this error appear again 🤔

Indeed .. that's very interesting.

[potiuk]

Let's see https://github.com/apache/airflow/actions/runs/7339685558

Update: Failed again with the same error on Postgres 14/Python 3.10

[potiuk]

Let's see the next one https://github.com/apache/airflow/actions/runs/7340134340

[potiuk]

Let's merge it to see if it helps. We had two subsequent main failures after merging PRs

[potiuk]

Merged the https://github.com/apache/airflow/actions/runs/7340642180 to see the third failure.

[potiuk]

Merged. It failed 3 times in a row now -let's see if that one will fix main ... Would be super strange but with this kind of side effect, it could be something very, very unrelated

[potiuk]

Aaaand surprise surprise ... Merging the revert fixed the problem ?!?! .... this is a very strange issue