Provider version 2.30.0 failed provider configuration on terraform plan

[jtgorny]

Attempted to upgrade from 2.29.0 to 2.30.0 and suddenly the provider is throwing errors on a terraform plan. Looking at the release notes here, I'm not seeing anything relevant to our current configuration that would produce this error. Simply downgrading back to 2.29.0 is sufficient as a workaround. Also worth noting, this only takes place during initial cluster creation. After a successful run on 2.29.0, I can upgrade to 2.30.0 and everything works fine.

│ Error: Provider configuration: cannot load Kubernetes client config
│ 
│   with provider["registry.terraform.io/hashicorp/kubernetes"],
│   on main.tf line 17, in provider "kubernetes":
│   17: provider "kubernetes" {
│ 
│ invalid configuration: default cluster has no server defined

Issue is only present on new builds, if I am running a terraform plan against an environment that already has a cluster, I get no error.

Terraform Version, Provider Version and Kubernetes Version

Terraform version: `1.8.4` (also tried `1.4.7`)
Kubernetes provider version: `2.30.0`
Kubernetes version: `1.29`

Affected Resource(s)

tf plan failing due to "invalid configuration" in provider

Terraform Configuration Files

• Provider definition

provider "kubernetes" {
  host                              = data.aws_eks_cluster.default.endpoint
  cluster_ca_certificate = base64decode(data.aws_eks_cluster.default.certificate_authority[0].data)

  exec {
    api_version = "client.authentication.k8s.io/v1"
    command    = "aws"
    # This requires the awscli to be installed locally where Terraform is executed
    args = ["eks", "get-token", "--cluster-name", data.aws_eks_cluster.default.name, "--region", var.region]
    env = {
      AWS_PROFILE = var.profile
    }
  }
}

• Kube Config Template file

apiVersion: v1
clusters:
- cluster:
    server: ${EKS_SERVICE_ENDPOINT}
    certificate-authority-data: ${EKS_CA_DATA}
  name: ${APPLICATION}-${ENVIRONMENT}

contexts:
- context:
    cluster: ${APPLICATION}-${ENVIRONMENT}
    user: ${APPLICATION}-${ENVIRONMENT}-admin
  name: ${APPLICATION}-${ENVIRONMENT}-system

current-context: ${APPLICATION}-${ENVIRONMENT}-system
kind: Config
preferences: {}

users:
- name: ${APPLICATION}-${ENVIRONMENT}-admin
  user:
    exec:
      apiVersion: client.authentication.k8s.io/v1
      interactiveMode: IfAvailable
      command: aws
      args:
        - "eks"
        - "get-token"
        - "--cluster-name"
        - "${K8_CLUSTER_NAME}"
      env:
        - name: AWS_PROFILE
          value: ${AWS_PROFILE}

• Cluster Data definition

data "aws_eks_cluster" "default" {
  depends_on = [ module.EKS.EKS-Server-Endpoint ]
  name             = var.EKSClusterName
}

• EKS Module definition

module "EKS" {
  source  = "terraform-aws-modules/eks/aws"
  version = "18.30.0"
  ...blah...
  ...blah...
  ...blah...
}

Expected Behavior

What should have happened? A successful plan to create an EKS cluster.

Actual Behavior

What actually happened?

│ Error: Provider configuration: cannot load Kubernetes client config
│ 
│   with provider["registry.terraform.io/hashicorp/kubernetes"],
│   on main.tf line 17, in provider "kubernetes":
│   17: provider "kubernetes" {
│ 
│ invalid configuration: default cluster has no server defined

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment