Support for daemonset-eviction-for-occupied-nodes

[nemobis]

Is there an existing issue for this?

• I have searched the existing issues

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment and review the contribution guide to help.

Description

The cluster autoscaler in AKS is nearly useless by default for the default pool, which always contains kube-system pods and usually some daemonset pods.

Currently the provider supports skip-nodes-with-local-storage and skip-nodes-with-system-pods, but additionally it would be good to have support for more cluster-autoscaler-profile-settings

• daemonset-eviction-for-empty-nodes: "Whether DaemonSet pods will be gracefully terminated from empty nodes."
• daemonset-eviction-for-occupied-nodes: "Whether DaemonSet pods will be gracefully terminated from non-empty nodes."

Both are currently marked as "in preview".

New or Affected Resource(s)/Data Source(s)

azurerm_kubernetes_cluster (auto_scaler_profile)

Potential Terraform Configuration

No response

References

https://learn.microsoft.com/en-us/azure/aks/cluster-autoscaler?tabs=azure-cli#cluster-autoscaler-profile-settings
https://registry.terraform.io/providers/hashicorp/azurerm/latest/docs/resources/kubernetes_cluster#skip_nodes_with_local_storage

[rcskosir]

Thanks for taking the time to open this issue. It looks like the behavior you requested is in Preview by the underlying AKS API so I am going to close it for now, please read here about the AKS API change in 4.x. Once the feature has become generally available and is accessible in a Stable API version, we can reopen this request or you can create a new one.

[github-actions]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.
If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.