shubhamraj-git commented on code in PR #73087:
URL: https://github.com/apache/airflow/pull/73087#discussion_r4084498696


##########
task-sdk/src/airflow/sdk/definitions/dag.py:
##########
@@ -1177,6 +1177,22 @@ def _check_adjacent_tasks(task_id, current_task):
                 else:
                     path_stack.append(child_to_check)
 
+        task_group_dict = self.task_group.get_task_group_dict()
+        for task_group in task_group_dict.values():
+            try:
+                task_group.topological_sort(group_dict=task_group_dict)

Review Comment:
   Non-blocking: I found a case where the full Dag passes this validation, but 
filtering it still causes Graph and Grid to return HTTP 500.
   
   ```python
   from airflow.providers.standard.operators.empty import EmptyOperator
   from airflow.sdk import DAG, TaskGroup
   
   with DAG("filter_cycle", schedule=None) as dag:
       with TaskGroup("g"):
           first = EmptyOperator(task_id="first")
           second = EmptyOperator(task_id="second")
           guard = EmptyOperator(task_id="guard")
           guard >> second
       bridge = EmptyOperator(task_id="bridge")
       first >> bridge >> second
   ```
   
   Tested in Breeze on `f0677f67705`, using SQLite and the actual FastAPI 
routes:
   
   - Parsing and `dag.check_cycle()` succeed; unfiltered Graph and Grid 
requests return 200.
   - Adding `root=g.first&include_downstream=true` makes both requests return 
500. With server exceptions enabled, both raise `ValueError("A cyclic 
dependency occurred in dag: ...")`.
   
   Filtering removes `g.guard`, so `g.second` becomes a group root and the 
projected dependencies become `g ↔ bridge`. I also executed the same dependency 
shape with four PythonOperator tasks: all succeeded, but the filtered requests 
still failed afterwards.
   
   A control using the base revision's `DAG.check_cycle()` reproduces the same 
failure, so this is pre-existing, not a regression introduced by this PR. Would 
you consider this filtered-view case in scope here, or leave it for a separate 
follow-up?
   
   ---
   
   Drafted-by: Codex (GPT-5); reviewed by @shubhamraj-git before posting



-- 
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]

Reply via email to