apache / apache/airflow

ExternalTaskSensor External link with incorrect URL

Open
#61,850 1 comment 1 reaction 0 assignees View on GitHub
area:core kind:bug needs-triage
Dominant language
Python
Stars
46.9k
Forks
17.8k
Avg merge
2d 9h
Merged PRs (30d)
472

Description

### Apache Airflow version

3.1.7

### If "Other Airflow 3 version" selected, which one?

_No response_

### What happened?

The "External DAG" that appears in the details page of an ExternalTaskSensor instance contains a wrong URL.

It assumes that the task is waiting for a run with the exact same logical date even if you specify another logical date.

### What you think should happen instead?

_No response_

### How to reproduce

1. Create a Dag with a Cron trigger of (example) `35 22 * * *` that contains an EmptyOperator
2. Create a Dag with a Cron trigger of (example) `40 22 * * *` that contains an ExternalTaskSensor that waits for Dag 1 and uses a function for `execution_date_fn` similar to :

```python
def get_last_execution_date_of_dag(logical_date: datetime.datetime, **kwargs):
import airflow_client.client

from airflow.configuration import conf
from myprovider.hooks.api_token import AirflowAPITokenHook # Won't work, you need to obtain the token somehow

hook = AirflowAPITokenHook(airflow_api_token_conn_id="airflow_api_token_default")
token = hook.get_token()

base_url = conf.get("api", "base_url")

dag_id = kwargs["task"].external_dag_id

configuration = airflow_client.client.Configuration(
host=base_url, access_token=token
)
with airflow_client.client.ApiClient(configuration) as api_client:
api_instance = airflow_client.client.DagRunApi(api_client)

api_response = api_instance.get_dag_runs(
dag_id, logical_date_lte=logical_date, limit=1, order_by=["-logical_date"]
)

if not api_response.dag_runs or len(api_response.dag_runs) == 0:
raise ValueError(
f"No previous dag_run found for DAG {dag_id} before {logical_date}. Please ensure the parent Dag has at least one run before this date."
)

last_dag_run = api_response.dag_runs[0]
parent_logical_date = last_dag_run.logical_date

print(
f"Found last dag_run for DAG {dag_id} before {logical_date}: {parent_logical_date}"
)

return parent_logical_date
```

When Dag 2 gets executed, you'll see this line in the log:

```
Found last dag_run for DAG dag1 before 2026-02-11 22:35:00+00:00: 2026-02-11 22:30:00+00:00
```

Which is correct.

However, the link that is generated is:

```
https://myairflow.example.com/dags/dag1/runs/scheduled__2026-02-11T22:35:00+00:00
```

Which contains the run ID of Dag2, not Dag1.

### Operating System

Debian GNU/Linux 12 (bookworm)

### Versions of Apache Airflow Providers

_No response_

### Deployment

Docker-Compose

### Deployment details

_No response_

### Anything else?

_No response_

### 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)

Contributor guide

Open the contributing guide

Research direction

Start at the ExternalTaskSensor external-link generation used by the details page, then reproduce the case with execution_date_fn returning a different logical date. Verify that the External DAG link identifies the target DAG run rather than using the waiting DAG's logical date or run ID.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
backend, data-engineering
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
38/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.