spec.driver.useNvidiaDriverCRD: field not declared in schema .status.conditions: field not declared in schema

[ProboticsX]

1. Quick Debug Information

• OS/Version(e.g. RHEL8.6, Ubuntu22.04): Ubuntu22.04
• Container Runtime Type/Version(e.g. Containerd, CRI-O, Docker): Containerd
• K8s Flavor/Version(e.g. K8s, OCP, Rancher, GKE, EKS): EKS
• GPU Operator Version: v23.9.1

2. Issue or feature description

I've been seeing the following error when deploying gpu-operator on my EKS cluster managed through argocd:
spec.driver.useNvidiaDriverCRD: field not declared in schema .status.conditions: field not declared in schema

Although the gpu-operator still works fine and the gpu-workload does get assigned to a gpu-node but I still see this problem with the configuration in argocd which is not letting the application sync:

3. Steps to reproduce the issue

I'm using the gpu-operator v23.9.1

I downloaded the gpu operator using the command below:

• helm fetch nvidia/gpu-operator --untar

And then installed the gpu operator using argocd

4. Information to attach (optional if deemed irrelevant)

• kubernetes pods status: kubectl get pods -n ops-gpu-operator

• kubernetes daemonset status: kubectl get ds -n ops-gpu-operator
  

[cdesiniotis]

@ProboticsX I suspect the version of the clusterpolicy CRD deployed in your cluster is stale. Please follow our recommendations for managing CRDs when upgrading the gpu-operator: https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/upgrade.html

[cdesiniotis]

It appears this is an open issue with ArgoCD concerning schema validation of the new CRD. See these related issues:
#602 (comment)
argoproj/argo-cd#16144

[ProboticsX]

Thanks @cdesiniotis you're right, it was an argo issue.