You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Which component are you using?:
Cluster Autoscaler on AWS
Is your feature request designed to solve a problem? If so describe the problem this feature should solve.:
Especially for smaller node types DaemonSet CPU/memory requests contribute a lot to node CPU/memory requests and therefore sometimes blocking node termination for scale-down. --ignore-daemonsets-utilization is already available as a global CAS option to solve this.
k8s.io/cluster-autoscaler/node-template/autoscaling-options/ignoredaemonsetsutilization: true (overrides ignore-daemonsets-utilization value for that specific ASG)
to override global setting per ASG using this label
The text was updated successfully, but these errors were encountered:
youwalther65
changed the title
Override --ignoredaemonsets-utilization for Cluster Auroscaler ona per AWS ASG basis using ASG label
Override --ignoredaemonsets-utilization for Cluster Auroscaler on a per AWS ASG basis using ASG label
Feb 14, 2023
Set IgnoreDaemonSetsUtilization to true globally so that it applies to all the nodegroups/ASGs. One problem with this (I think) is scale down wouldn't consider node resource utilization accurately because it would always ignore daemonset resource utilization.
youwalther65
changed the title
Override --ignoredaemonsets-utilization for Cluster Auroscaler on a per AWS ASG basis using ASG label
Override --ignore-daemonsets-utilization for Cluster Auroscaler on a per AWS ASG basis using ASG label
May 23, 2023
youwalther65
changed the title
Override --ignore-daemonsets-utilization for Cluster Auroscaler on a per AWS ASG basis using ASG label
Override --ignore-daemonsets-utilization for Cluster Autoscaler on a per AWS ASG basis using ASG label
May 23, 2023
Which component are you using?:
Cluster Autoscaler on AWS
Is your feature request designed to solve a problem? If so describe the problem this feature should solve.:
Especially for smaller node types DaemonSet CPU/memory requests contribute a lot to node CPU/memory requests and therefore sometimes blocking node termination for scale-down. --ignore-daemonsets-utilization is already available as a global CAS option to solve this.
Describe the solution you'd like.:
Similar to the already available approach of AWS ASG labels which override the global cluster-autoscaler settings for the labeled ASGs , see https://github.com/kubernetes/autoscaler/blob/master/cluster-autoscaler/cloudprovider/aws/README.md. I 'd like to introduce:
k8s.io/cluster-autoscaler/node-template/autoscaling-options/ignoredaemonsetsutilization: true (overrides ignore-daemonsets-utilization value for that specific ASG)
to override global setting per ASG using this label
The text was updated successfully, but these errors were encountered: