Apache Airflow Provider(s)
microsoft-azure
Versions of Apache Airflow Providers
=10.4.0
Apache Airflow version
2.10.2
Operating System
Official Airflow image apache/airflow:2.10.2-python3.12 hosted on AKS
Deployment
Official Apache Airflow Helm Chart
Deployment details
Official Airflow image apache/airflow:2.10.2-python3.12 hosted on AKS
What happened
This issue resembles Unclosed connector / client session when using asyncio #13242
However, it occurs when triggering an Azure Data Factory pipeline via Airflow using the AzureDataFactoryRunPipelineOperator.
Regardless of whether the pipeline completes successfully or fails, the following error is consistently logged:
[2025-09-16, 15:08:11 CEST] {base_events.py:1820} ERROR - Unclosed client session
client_session: <aiohttp.client.ClientSession object at 0x7f80f1a34380>
[2025-09-16, 15:08:11 CEST] {base_events.py:1820} ERROR - Unclosed connector
connections: ['[(<aiohttp.client_proto.ResponseHandler object at 0x7f80f1bb1970>, 102727.274540543)]']
connector: <aiohttp.connector.TCPConnector object at 0x7f80f1a34b00>
What you think should happen instead
Displaying the error message is not intended in this context.
How to reproduce
- Use AzureDataFactoryRunPipelineOperator to trigger an ADF pipeline.
- Wait for the pipeline to either succeed or fail.
- Observe the logs — the error appears regardless of the pipeline outcome.
Anything else
This occurs consistently upon task completion, irrespective of the pipeline's success or failure.
Are you willing to submit PR?
Code of Conduct
Apache Airflow Provider(s)
microsoft-azure
Versions of Apache Airflow Providers
Apache Airflow version
2.10.2
Operating System
Official Airflow image apache/airflow:2.10.2-python3.12 hosted on AKS
Deployment
Official Apache Airflow Helm Chart
Deployment details
Official Airflow image apache/airflow:2.10.2-python3.12 hosted on AKS
What happened
This issue resembles Unclosed connector / client session when using asyncio #13242
However, it occurs when triggering an Azure Data Factory pipeline via Airflow using the AzureDataFactoryRunPipelineOperator.
Regardless of whether the pipeline completes successfully or fails, the following error is consistently logged:
What you think should happen instead
Displaying the error message is not intended in this context.
How to reproduce
Anything else
This occurs consistently upon task completion, irrespective of the pipeline's success or failure.
Are you willing to submit PR?
Code of Conduct