dirrao opened a new issue, #35676: URL: https://github.com/apache/airflow/issues/35676
### Apache Airflow version main (development) ### What happened During scheduler restart, when the pods are already launched by the previous scheduler and in a pending state (task still in queued state), the new scheduler will move those queued tasks from queued state to reschedule. It leads to lunching the task pods again. I wouldn't expect to relaunch the task pods when they are already launched. ### What you think should happen instead It should detect the existing pods for the queued tasks and not move those tasks from the queued state to the reschedule state ### How to reproduce Launch the task worker pods with any of the existing DAGs and restart the scheduler process/pod and the new scheduler will launch the same task pods again when the existing task pod is still in a queued state ### Operating System CentOS 6 ### Versions of Apache Airflow Providers apache-airflow-providers-cncf-kubernetes=7.9.0 ### Deployment Other Docker-based deployment ### Deployment details Terraform ### Anything else _No response_ ### Are you willing to submit PR? - [X] 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]
