Swalloow opened a new issue, #24103:
URL: https://github.com/apache/airflow/issues/24103

   ### Description
   
   > When managing the scale of a group of replicas using the 
HorizontalPodAutoscaler, it is possible that the number of replicas keeps 
fluctuating frequently due to the dynamic nature of the metrics evaluated. This 
is sometimes referred to as thrashing, or flapping.
   
https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/#flapping
   
   Sometimes clusters need to restrict the flapping of Airflow worker replicas.
   KEDA supports 
[`advanced.horizontalPodAutoscalerConfig`](https://keda.sh/docs/1.4/concepts/scaling-deployments/).
   It would be great if the users would have the option in the helm chart to 
configure scale down behavior.
   
   ### Use case/motivation
   
   KEDA currently cannot set advanced options.
   We want to set advanced options like `scaleDown.stabilizationWindowSeconds`, 
`scaleDown.policies`.
   
   ### Related issues
   
   _No response_
   
   ### Are you willing to submit a 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]

Reply via email to