ecodina commented on issue #49931:
URL: https://github.com/apache/airflow/issues/49931#issuecomment-2862089049
Some more information that may be useful. I've activated DEBUG logging, and
found this:
```
[2025-05-08T06:52:50.927+0000] {_trace.py:47} DEBUG - connect_tcp.started
host='localhost' port=8080 local_address=None timeout=5.0 socket_options=None
[2025-05-08T06:52:50.928+0000] {_trace.py:47} DEBUG - connect_tcp.failed
exception=ConnectError(ConnectionRefusedError(111, 'Connection refused'))
2025-05-08 06:52:50 [warning ] Starting call to
'airflow.sdk.api.client.Client.request', this is the 4th time calling it.
[airflow.sdk.api.client]
```
The scheduler is trying to connect to localhost:8080, but the apiserver is
in another container. In this case, it should be using host.docker.internal, or
the IP of the other container, or the domain.
I've put the apiserver behind traefik again
(https://dev-airflow.meteosim.me), and updated the configuration as follows:
```
[api]
base_url = https://dev-airflow.meteosim.me
```
However, the scheduler is still trying to access localhost:8080.
There are also a couple of interesting log lines:
```
[2025-05-08T07:37:11.299+0000] {scheduler_job_runner.py:925} ERROR - DAG
'dag.example4' for task instance <TaskInstance: dag.example4.get_data
manual__2025-05-08T07:36:58.191280+00:00 [queued]> not found in serialized_dag
table
[2025-05-08T07:37:11.300+0000] {taskinstance.py:2459} ERROR - Executor
LocalExecutor(parallelism=32) reported that the task instance <TaskInstance:
dag.example4.get_data manual__2025-05-08T07:36:58.191280+00:00 [queued]>
finished with state failed, but the task instance's state attribute is queued.
Learn more:
[https://airflow.apache.org/docs/apache-airflow/stable/troubleshooting.html#task-state-changed-externally](https://airflow.apache.org/docs/apache-airflow/stable/troubleshooting.html#task-state-changed-externally)
```
However, the dag processor shows that there are no errors importing
dag.example4
As soon as Airflow 3.1 is released, I'll test it again.
--
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]
