nathadfield opened a new issue, #60049:
URL: https://github.com/apache/airflow/issues/60049
### Apache Airflow version
3.1.5
### If "Other Airflow 3 version" selected, which one?
_No response_
### What happened?
In Airflow 3.1.5 with `apache-airflow-providers-standard==1.10.0`, a
`TriggerDagRunOperator` configured for “fire-and-forget”
(`wait_for_completion=False`) enters the `DEFERRED` state (e.g. “queued
deferred”) and does not progress, even though the target DAG run completes
immediately or has no running tasks.
This occurs when `deferrable=True` is set on the operator (either explicitly
or via default_args). Setting deferrable=False causes the task to behave as
expected (trigger the target DAG and succeed).
### What you think should happen instead?
When `wait_for_completion=False`, `TriggerDagRunOperator` should never
defer. `deferrable=True` should be treated as a no-op (or overridden/validated)
unless `wait_for_completion=True`.
Expected behaviour:
- Trigger target DagRun (optionally reset it, if configured)
- Finish the task in success
- Do not enter deferred state
### How to reproduce
Create two DAG files:
**dags/target_dag.py**
```
from datetime import datetime
from airflow.sdk import DAG
from airflow.operators.empty import EmptyOperator
with DAG(
dag_id="repro_target_dag",
start_date=datetime(2024, 1, 1),
schedule=None,
catchup=False,
) as dag:
start = EmptyOperator(task_id="start")
end = EmptyOperator(task_id="end")
start >> end
```
**dags/trigger_dag.py**
```
from datetime import datetime
from airflow.sdk import DAG
from airflow.providers.standard.operators.trigger_dagrun import
TriggerDagRunOperator
with DAG(
dag_id="repro_trigger_dag",
start_date=datetime(2024, 1, 1),
schedule=None,
catchup=False,
) as dag:
TriggerDagRunOperator(
task_id="trigger_target_fire_and_forget",
trigger_dag_id="repro_target_dag",
wait_for_completion=False, # fire-and-forget
deferrable=True, # triggers the issue
reset_dag_run=True, # optional (issue observed with this set)
)
```
**Steps**
1. Start Airflow (ensure scheduler + triggerer are running).
2. Trigger repro_trigger_dag from the UI.
3. Observe repro_target_dag completes immediately.
4. Observe repro_trigger_dag.trigger_target_fire_and_forget enters DEFERRED
and remains stuck.
**Workaround / control**
* Change deferrable=True to deferrable=False and rerun. The trigger task
completes successfully.
### Operating System
[Breeze:3.12.12] root@eaed748ca5ce:/opt/airflow$ cat /etc/os-release
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)" NAME="Debian GNU/Linux"
VERSION_ID="12" VERSION="12 (bookworm)" VERSION_CODENAME=bookworm ID=debian
HOME_URL="https://www.debian.org/"; SUPPORT_URL="https://www.debian.org/support";
BUG_REPORT_URL="https://bugs.debian.org/";
### Versions of Apache Airflow Providers
_No response_
### Deployment
Official Apache Airflow Helm Chart
### Deployment details
_No response_
### Anything else?
This issue appears related to previous reports where `TriggerDagRunOperator`
becomes stuck in the `DEFERRED` state when running in deferrable mode,
particularly around edge cases in task retries, restarts, or DagRun handling.
While those issues often involve wait_for_completion=True or other specific
conditions, this report highlights a narrower case where deferral occurs even
when wait_for_completion=False.
Related issues (for context):
* TriggerDagRunOperator stuck in deferred state after restart/reset (#57756)
* Deferred trigger tasks stuck in Airflow 3.x (#52247)
These prior issues suggest that `TriggerDagRunOperator`’s deferrable
execution path may need additional guarding to ensure deferral is only used
when explicitly waiting for downstream DagRun completion.
### 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]
