Open MatthewHou opened 1 year ago
/cc @aws-controllers-k8s/applicationautoscaling-maintainer
cc @aws-controllers-k8s/ec2-maintainer as well
@MatthewHou please share any timelines on when you would like to use this feature. Thank you
@surajkota At the moment, we are using terraform managed scheduled actions as a workaround. We'd like to utilize this feature in ACK once it's available to manage Sagemaker models / endpoints in one place.
Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale
.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close
.
Provide feedback via https://github.com/aws-controllers-k8s/community.
/lifecycle stale
/lifecycle frozen
Is your feature request related to a problem? We are using Sagemaker real-time endpoint to serve traffic that is highly predictable and spikes quickly. In order to minimize cost during off peak hours, we would like to use scheduled action to change the min size before and after the peak hours. We also want to use target tracking scaling to scale out based on CPU usage. At the moment, Application Auto Scaling Controller only supports target tracking scaling and step scaling.
Describe the solution you'd like Add support to configure scheduled actions
Describe alternatives you've considered Terraform offers "aws_appautoscaling_scheduled_action" resource that appears to have the same desired effect. However, since we are using Sagermaker controller, we'd like to keep all relevant configs in the same place.