This was the smallest change I could come up with, but I believe the result backends could be updated to take care of this on the call to forget as well. In that case, I can create a new PR for that. Though I'm fine with this approach.

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 78.61%. Comparing base (6dcecbe) to head (a761b28).
⚠️ Report is 1 commits behind head on main.

@@            Coverage Diff             @@
##             main    #9806      +/-   ##
==========================================
- Coverage   78.62%   78.61%   -0.01%     
==========================================
  Files         153      153              
  Lines       19199    19200       +1     
  Branches     2547     2547              
==========================================
  Hits        15095    15095              
+ Misses       3811     3809       -2     
- Partials      293      296       +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

should we also add unit and integration tests? or they are already covered?

I think so. A sanity check seems appropriate for this issue.

lets do it

I added this unit test, but it feels a bit redundant.

self = <pytest_celery.api.worker.CeleryTestWorker object at 0x7f42339a6b10>
log = 'Starting long running task', message = '', timeout = 60

  def assert_log_exists(self, log: str, message: str = "", timeout: int = RESULT_TIMEOUT) -> None:
      """Assert that a log exists in the container.
  
      Args:
          log (str): Log to assert.
          message (str, optional): Message to display while waiting. Defaults to "".
          timeout (int, optional): Timeout in seconds. Defaults to RESULT_TIMEOUT.
      """
      try:
          self.wait_for_log(log, message, timeout)
      except pytest_docker_tools.exceptions.TimeoutError:

      assert False, f"Worker container '{self.name()}' did not log -> {log} within {timeout} seconds"

E AssertionError: Worker container 'peaceful_kowalevski' did not log -> Starting long running task within 60 seconds
E assert False

t/integration/test_canvas.py::test_stamping_mechanism::test_all_tasks_of_canvas_are_stamped [Assert all tasks are stamped] SUBPASS::debug::Code: null
Error: Final attempt failed. Timeout of 3600000ms hit

test failures are seems to be unrelated

so how should we move forward with this? #9797 (comment)

so how should we move forward with this? #9797 (comment)

I don't see how it could be related at the moment., and I had the same issues with multiple previous releases. I'm hoping it's a an issue with my local setup. We can go ahead.