pedeveaux opened a new issue, #51906:
URL: https://github.com/apache/airflow/issues/51906
### Apache Airflow Provider(s)
standard
### Versions of Apache Airflow Providers
apache-airflow-providers-standard==1.2.0
### Apache Airflow version
3.0.2
### Operating System
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)" NAME="Debian GNU/Linux"
VERSION_ID="12" VERSION="12 (bookworm)" VERSION_CODENAME=bookworm
### Deployment
Docker-Compose
### Deployment details
Using the LocalExecutor in a docker-compose deployment.
There is an airflow-triggerer container running in the deployment.
### What happened
When running a DAG that triggers a dependent DAG via the
`TriggerDagRunOperator` with `deferrable=True`, the triggering task never
returns even though the dependent DAG has completed
### What you think should happen instead
The triggering task should complete and the downstream tasks in the parent
DAG should execute.
### How to reproduce
1. Create these two DAGs:
Parent DAG:
```python
import pendulum
from airflow.providers.standard.operators.trigger_dagrun import
TriggerDagRunOperator
from airflow.sdk import dag, task
@dag(
schedule=None,
start_date=pendulum.datetime(2025, 1, 1),
catchup=False,
tags=["test"],
)
def trigger_dagrun_dag():
@task
def start_task(task_type):
return f"The {task_type} task has completed."
@task
def end_task(task_type):
return f"The {task_type} task has completed."
trigger_dependent_dag = TriggerDagRunOperator(
task_id="trigger_dependent_dag",
trigger_dag_id="dependent_dag",
wait_for_completion=True,
deferrable=True,
)
start_task("starting") >> trigger_dependent_dag >> end_task("ending")
trigger_dagrun_dag()
```
Dependent DAG:
```python
from airflow.sdk import dag, task
import pendulum
@dag(
dag_id="dependent_dag",
schedule=None,
start_date=pendulum.datetime(2025, 6, 1),
catchup=False,
tags=["test"],
)
def dependent_dag():
@task
def dependent_task():
return "This is a dependent task."
dependent_task()
dependent_dag()
```
2. Run the Parent DAG and see that the `trigger_dependent_dag` task never
completes even though the dependent DAG has completed.
### Anything else
This problem can be mitigated by setting `deferrable=False` in the
`TriggerDagRunOperator`. However this then consumes an execution slot for the
duration of the triggered DAG. The triggerer service is supposed to allow tasks
to be deferred and not consume an execution slot.
### Are you willing to submit PR?
- [ ] Yes I am willing to submit a PR!
### Code of Conduct
- [x] I agree to follow this project's [Code of
Conduct](https://github.com/apache/airflow/blob/main/CODE_OF_CONDUCT.md)
--
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]
