bmanan7 opened a new issue, #58313:
URL: https://github.com/apache/airflow/issues/58313
### Apache Airflow version
3.1.2
### If "Other Airflow 2/3 version" selected, which one?
_No response_
### What happened?
When retrieving a connection from AWS Secrets Manager(programmatically) that
does not include conn_type, Airflow 3 raises:
```
AirflowNotFoundException: The conn_id `dev_sec` isn't defined
```
However, running:
```
airflow connections get dev_sec -o yaml
```
returns the expected connection details, showing that the secret is found
and accessible via the CLI - only the programmatic lookup fails.
Also, In Airflow 2 the same secret is retrieved
successfully(programmatically) but with type=None.
### What you think should happen instead?
Airflow 3 should maintain backward compatibility and allow retrieving
connections without explicitly defining `conn_type`, defaulting to None or
similar (consistent with Airflow 2 behavior).
### How to reproduce
1. Create an AWS secret named airflow/connections/dev_sec with the following
JSON:
```
{
"username": "tuser",
"password": "tpass"
}
```
(no conn_type field)
2. Configure Airflow to use SecretsManagerBackend with proper AWS
credentials. For example:
```
AIRFLOW__SECRETS__BACKEND_KWARGS='{"connections_prefix":
"airflow/connections","variables_prefix": "airflow/variables","region_name":
"us-east-2"}'
AIRFLOW__SECRETS__BACKEND=airflow.providers.amazon.aws.secrets.secrets_manager.SecretsManagerBackend
AWS_DEFAULT_REGION=us-east-2
AWS_ACCESS_KEY_ID=XXXXXX
AWS_SECRET_ACCESS_KEY=XXXX
```
3. In a DAG, try to access the connection:
```
import logging
from airflow import DAG
from airflow.operators.python import PythonOperator
from airflow.utils.timezone import datetime
from airflow.hooks.base import BaseHook
from airflow.models.connection import Connection
CONN_ID = "dev_sec"
def _print_conn():
conn: Connection = BaseHook.get_connection(CONN_ID)
logging.info(
"Got conn: type=%s host=%s port=%s schema=%s extras=%s",
conn.conn_type, conn.host, conn.port, conn.schema,
list(conn.extra_dejson.keys())
)
print(f"TYPE={conn.conn_type}")
print(f"HOST={conn.host}")
print(f"PORT={conn.port}")
return "printed"
with DAG(
dag_id="print_conn_via_basehook",
start_date=datetime(2024, 1, 1),
schedule=None,
catchup=False,
tags=["secrets", "aws", "basehook"],
) as dag:
PythonOperator(task_id="print_conn", python_callable=_print_conn)
```
4. Observe the error on Airflow 3:
```
AirflowNotFoundException: The conn_id `dev_sec` isn't defined
```
5. Run the same on Airflow 2 - it succeeds but sets type=None.
### Operating System
macOS 15.6.1
### Versions of Apache Airflow Providers
_No response_
### Deployment
Official Apache Airflow Helm Chart
### Deployment details
_No response_
### Anything else?
When `conn_type: aws` is added to the secret JSON, Airflow 3 succeeds:
{
"conn_type": "aws",
"username": "tuser",
"password": "tpass"
}
### 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]
