ashb commented on code in PR #51952:
URL: https://github.com/apache/airflow/pull/51952#discussion_r2162028676


##########
airflow-core/src/airflow/utils/db_cleanup.py:
##########
@@ -104,29 +106,53 @@ def readable_config(self):
 
 config_list: list[_TableConfig] = [
     _TableConfig(table_name="job", recency_column_name="latest_heartbeat"),
-    _TableConfig(table_name="dag", recency_column_name="last_parsed_time"),
+    _TableConfig(
+        table_name="dag",
+        recency_column_name="last_parsed_time",
+        dependent_tables=[
+            "task_reschedule",
+            "task_instance_history",
+            "xcom",
+            "task_instance",
+            "dag_run",
+            "dag_version",
+        ],
+    ),
     _TableConfig(
         table_name="dag_run",
         recency_column_name="start_date",
         extra_columns=["dag_id", "run_type"],
         keep_last=True,
         keep_last_filters=[column("run_type") != DagRunType.MANUAL],
         keep_last_group_by=["dag_id"],
+        dependent_tables=["xcom", "task_reschedule", "task_instance_history", 
"task_instance"],
     ),
     _TableConfig(table_name="asset_event", recency_column_name="timestamp"),
     _TableConfig(table_name="import_error", recency_column_name="timestamp"),
     _TableConfig(table_name="log", recency_column_name="dttm"),
     _TableConfig(table_name="sla_miss", recency_column_name="timestamp"),
-    _TableConfig(table_name="task_instance", recency_column_name="start_date"),
+    _TableConfig(
+        table_name="task_instance",
+        recency_column_name="start_date",
+        dependent_tables=["task_instance_history", "xcom", "task_reschedule"],
+    ),
     _TableConfig(table_name="task_instance_history", 
recency_column_name="start_date"),
     _TableConfig(table_name="task_reschedule", 
recency_column_name="start_date"),
     _TableConfig(table_name="xcom", recency_column_name="timestamp"),
     _TableConfig(table_name="_xcom_archive", recency_column_name="timestamp"),
     _TableConfig(table_name="callback_request", 
recency_column_name="created_at"),
     _TableConfig(table_name="celery_taskmeta", 
recency_column_name="date_done"),
     _TableConfig(table_name="celery_tasksetmeta", 
recency_column_name="date_done"),
-    _TableConfig(table_name="trigger", recency_column_name="created_date"),
-    _TableConfig(table_name="dag_version", recency_column_name="created_at"),
+    _TableConfig(
+        table_name="trigger",
+        recency_column_name="created_date",
+        dependent_tables=["task_reschedule", "xcom", "task_instance_history", 
"task_instance"],
+    ),
+    _TableConfig(
+        table_name="dag_version",
+        recency_column_name="created_at",
+        dependent_tables=["task_instance_history", "task_reschedule", "xcom", 
"task_instance", "dag_run"],

Review Comment:
   Ditto here -- maybe we should only record the direct dependencies and let 
the code case down the transitive deps?



-- 
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