[GitHub] feng-tao commented on a change in pull request #3688: [AIRFLOW-2843] ExternalTaskSensor-check if external task exists

[GitBox <gi...@apache.org>]

feng-tao commented on a change in pull request #3688: [AIRFLOW-2843] ExternalTaskSensor-check if external task exists
URL: https://github.com/apache/incubator-airflow/pull/3688#discussion_r207604878
 
 

 ##########
 File path: airflow/sensors/external_task_sensor.py
 ##########
 @@ -70,9 +76,24 @@ def __init__(self,
         self.execution_date_fn = execution_date_fn
         self.external_dag_id = external_dag_id
         self.external_task_id = external_task_id
+        self.check_existence = check_existence
 
     @provide_session
     def poke(self, context, session=None):
+        TI = TaskInstance
+
+        if self.check_existence:
+            existence = session.query(TI).filter(
+                TI.dag_id == self.external_dag_id,
+                TI.task_id == self.external_task_id,
+            ).count()
+            session.commit()
+            if existence == 0:
+                raise AirflowException('The external task "' +
 
 Review comment:
   why stop waiting if the external task not exist? Shouldn't the right behavior to continue waiting for the task until task exist and finish(or timeout?). Sometimes the external task may not start right on time(e.g scheduler reason etc) and we still want downstream task wait for external task to finish first to make sure result align.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services