How not to structure subdags (unless you want them to run on their own schedule). Module bad_dags

maindag.py

from airflow.models import DAG
from airflow.operators import PythonOperator, SubDagOperator
from bad_dags.subdag import hive_dag
from datetime import timedelta, datetime

main_dag = DAG(
    dag_id='main_dag',
    schedule_interval=timedelta(hours=1),
    start_date=datetime(2015, 9, 18, 21)
)

# Obviously, this doesn't make sense without some other tasks
transform_hive = SubDagOperator(
    subdag=hive_dag,
    task_id='hive_transform',
    dag=main_dag,
    trigger_rule=TriggerRule.ALL_DONE
)

subdag.py

from airflow.models import DAG
from airflow.operators import HiveOperator
from datetime import timedelta, datetime

# This will be run on its own schedule as well as via the subdag operator
hive_dag = DAG('main_dag.hive_transform',
          # note the repetition here
           schedule_interval=timedelta(hours=1),
           start_date=datetime(2015, 9, 18, 21))

hive_transform = HiveOperator(task_id='flatten_tables',
                              hql=send_charge_hql,
                              dag=dag)

do you refers to hive_dag in L13 of subdag.py ? or main_dag?