karpenter icon indicating copy to clipboard operation
karpenter copied to clipboard

Published 20 hours ago • verified publisher icon kubernetes-sigs

Add Validation for NodeClassReference

Open jonathan-innis opened this issue 9 months ago • 1 comments

Description

What problem are you trying to solve?

Right now, there's no CEL or webhook validation for what's passed through the NodeClassReference. This means that it can be easy to forget the version portion or not properly capitalize kinds, etc. when specifying the reference.

It probably also makes sense for the cloudprovider to inject its own CEL validation when it pulls the NodePool CRD into its helm chart (similar to what AWS does here in scripts: https://github.com/aws/karpenter-provider-aws/tree/main/hack/validation)

  • 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

jonathan-innis avatar Apr 25 '24 07:04 jonathan-innis

How about using a ValidatingAdmissionPolicy with a custom, singleton params kind, and then having controllers write to that custom resource based on a watch for their node class (or whatever).

We'd make the extra CRD part of the CRDs chart, leaving the controller to actually add make an instance of that CR and / or add itself in.

It's more to implement, but as a pattern it leaves room for multiple providers in one cluster, and it avoids [should avoid] the risk of different implementations clashing over CRD writes.

This is an outline, let me know if folks want details clarified.

sftim avatar Apr 29 '24 16:04 sftim

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

  • After 90d of inactivity, lifecycle/stale is applied
  • After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
  • After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

  • Mark this issue as fresh with /remove-lifecycle stale
  • Close this issue with /close
  • Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

k8s-triage-robot avatar Jul 28 '24 16:07 k8s-triage-robot

This should be fixed in v1. We're doing validation for Group, but aren't doing validation for the others (name, kind)

njtran avatar Aug 12 '24 22:08 njtran