Under which category would you file this issue?
Airflow Core
Apache Airflow version
2.10.5
What happened and how to reproduce it?
Apache Airflow version: 2.10.5
Python: 3.11
Executor: LocalExecutor
Database backend: PostgreSQL 15
Deployment: Docker Compose on Windows (WSL2/Docker Desktop)
Browser: Microsoft Edge (Chromium)
The main DAGs list page (/home) does not update the "Last Run"
and "Next Run" columns after DAG runs complete. The stale values
persist even after:
- Auto-refresh enabled (toggled on)
- Manual browser hard refresh (Ctrl+F5)
- Running
airflow dags reserialize from CLI
- Waiting 10+ minutes
The "Recent Tasks" column (green circles) updates correctly.
Clicking into the individual DAG detail page shows the correct
and latest run timestamps.
What you think should happen instead?
The "Last Run" and "Next Run" columns on the main DAGs list page (/home) should update automatically when Auto-refresh is enabled, and should show the latest values after a hard browser refresh (Ctrl+F5) or after running airflow dags reserialize.
The frontend appears to cache the Last Run and Next Run timestamps and does not re-fetch them from the backend, even when Auto-refresh is enabled. The "Recent Tasks" column on the same page updates correctly, suggesting the polling mechanism works but the Last Run / Next Run columns are not included in the refresh cycle. Running airflow dags reserialize and hard-refreshing the browser does not resolve it. The DAG detail page always shows the correct timestamps, confirming the backend data is accurate.
Operating System
Windows 11 (VMware VM) running Docker Desktop with WSL2
Deployment
Docker-Compose
Apache Airflow Provider(s)
No response
Versions of Apache Airflow Providers
No response
Official Helm Chart version
Not Applicable
Kubernetes Version
No response
Helm Chart configuration
No response
Docker Image customizations
Custom Dockerfile based on apache/airflow:2.10.5-python3.11 with Oracle Instant Client 21.14 added for thick-mode database connectivity. No frontend or webserver modifications.
Anything else?
- PostgreSQL 15 as metadata database
- LocalExecutor
- 9 DAGs across 3 subfolders (marketing_bi, itom, pbi_admin)
- Issue persists across multiple browser sessions and after VM restart
- Tested on Microsoft Edge (Chromium) and Chrome
- Screenshots attached showing mismatch between main list and detail page
Are you willing to submit PR?
Code of Conduct
Under which category would you file this issue?
Airflow Core
Apache Airflow version
2.10.5
What happened and how to reproduce it?
Apache Airflow version: 2.10.5
Python: 3.11
Executor: LocalExecutor
Database backend: PostgreSQL 15
Deployment: Docker Compose on Windows (WSL2/Docker Desktop)
Browser: Microsoft Edge (Chromium)
The main DAGs list page (/home) does not update the "Last Run"
and "Next Run" columns after DAG runs complete. The stale values
persist even after:
airflow dags reserializefrom CLIThe "Recent Tasks" column (green circles) updates correctly.
Clicking into the individual DAG detail page shows the correct
and latest run timestamps.
What you think should happen instead?
The "Last Run" and "Next Run" columns on the main DAGs list page (/home) should update automatically when Auto-refresh is enabled, and should show the latest values after a hard browser refresh (Ctrl+F5) or after running
airflow dags reserialize.The frontend appears to cache the Last Run and Next Run timestamps and does not re-fetch them from the backend, even when Auto-refresh is enabled. The "Recent Tasks" column on the same page updates correctly, suggesting the polling mechanism works but the Last Run / Next Run columns are not included in the refresh cycle. Running
airflow dags reserializeand hard-refreshing the browser does not resolve it. The DAG detail page always shows the correct timestamps, confirming the backend data is accurate.Operating System
Windows 11 (VMware VM) running Docker Desktop with WSL2
Deployment
Docker-Compose
Apache Airflow Provider(s)
No response
Versions of Apache Airflow Providers
No response
Official Helm Chart version
Not Applicable
Kubernetes Version
No response
Helm Chart configuration
No response
Docker Image customizations
Custom Dockerfile based on apache/airflow:2.10.5-python3.11 with Oracle Instant Client 21.14 added for thick-mode database connectivity. No frontend or webserver modifications.
Anything else?
Are you willing to submit PR?
Code of Conduct