uranusjr commented on code in PR #27106:
URL: https://github.com/apache/airflow/pull/27106#discussion_r997965953
##########
airflow/cli/commands/task_command.py:
##########
@@ -126,10 +126,12 @@ def _get_dag_run(
dag_run = DagRun(dag.dag_id, run_id=exec_date_or_run_id,
execution_date=dag_run_execution_date)
return dag_run, True
elif create_if_necessary == "db":
+ data_interval =
dag.infer_automated_data_interval(logical_date=dag_run_execution_date)
Review Comment:
This would not always work since infer_automated_data_interval` is only
designed to provide backward compatibility for DAG runs that are created before
the data interval concept was introduced, and a DAG using a custom timetable
would not work correctly with this method. I think `infer_manual_data_interval`
is probably a better choice; it is not technically correct (the call below
created a scheduled run), but the exact value of the data interval does not
actually matter much here anyway since the run is only supposed to be a stub.
##########
airflow/cli/commands/task_command.py:
##########
@@ -126,10 +126,12 @@ def _get_dag_run(
dag_run = DagRun(dag.dag_id, run_id=exec_date_or_run_id,
execution_date=dag_run_execution_date)
return dag_run, True
elif create_if_necessary == "db":
+ data_interval =
dag.infer_automated_data_interval(logical_date=dag_run_execution_date)
Review Comment:
This would not always work since `infer_automated_data_interval` is only
designed to provide backward compatibility for DAG runs that are created before
the data interval concept was introduced, and a DAG using a custom timetable
would not work correctly with this method. I think `infer_manual_data_interval`
is probably a better choice; it is not technically correct (the call below
created a scheduled run), but the exact value of the data interval does not
actually matter much here anyway since the run is only supposed to be a stub.
--
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]
