Can not import a cluster with terraform-provider-rke V1.4.3.

[LefterisBanos]

Terraform: v1.5.0
RKE Terraform Provider: v1.4.3
RKE Cluster: v1.26.9-rancher1-1
RKE: 1.4.7

When importing a cluster using latest 1.4.3 version of the provider getting following error :
Error: │ ============= RKE outputs ============== │ time="2023-11-23T15:27:29+02:00" level=info msg="Creating RKE cluster..." │ │ kubernetes version v1.26.9-rancher1-1 requires enable_cri_dockerd to be set to true

Although is set: enable_cri_dockerd: true

How to reproduce:
Use a cluster.yaml file and a cluster.rkestate file for a running cluster created from RKE CLI.
At the cluster.yaml set the enable_cri_dockerd: true.

Import cluster using terraform-provider-rke:

terraform import rke_cluster.qa cluster.yml:cluster.rkestate.

Import will fail, even if enable_cri_dockerd is properly set to true

[jiaqiluo]

Waiting for RC

[jiaqiluo]

the fix can be validated at https://github.com/rancher/terraform-provider-rke/releases/tag/v1.4.5-rc1

[Josh-Diamond]

Ticket # 432 - Test Results - ✅

Reproduced with RKE v1.4.7 and TFP-RKE v1.4.3:

1. Using RKE v1.4.7 -- create a standalone RKE cluster using the following cluster.yml:

ssh_key_path: <path_to_ssh_key_here>
enable_cri_dockerd: true
kubernetes_version: v1.26.6-rancher1-1

nodes:
  - address: <REDACTED>
    internal_address: <REDACTED>
    user: ubuntu
    role: [etcd, controlplane, worker]

  - address: <REDACTED>
    internal_address: <REDACTED>
    user: ubuntu
    role: [etcd, controlplane, worker]

  - address: <REDACTED>
    internal_address: <REDACTED>
    user: ubuntu
    role: [etcd, controlplane, worker]

2. Once cluster is up and active, create a main.tf file with the following configurations, and run terraform init:

terraform {
  required_providers {
    rke = {
      source = "rancher/rke"
      version = "1.4.3"
    }
  }
}

provider "rke" {}

resource "rke_cluster" "rke_cluster" {}

3. Once terraform has been successfully initialized, import the standalone RKE cluster created in step 1, by running the following command:

terraform import rke_cluster.rke_cluster cluster.yml:cluster.rkestate

4. Reproduced - kubernetes version v1.26.6-rancher1-1 requires enable_cri_dockerd to be set to true

Screenshot:

---

Verified with RKE v1.4.7 and TFP-RKE v1.4.5-rc1:

1. Using RKE v1.4.7 -- create a standalone RKE cluster using the following cluster.yml:

ssh_key_path: <path_to_ssh_key_here>
enable_cri_dockerd: true
kubernetes_version: v1.26.6-rancher1-1

nodes:
  - address: <REDACTED>
    internal_address: <REDACTED>
    user: ubuntu
    role: [etcd, controlplane, worker]

  - address: <REDACTED>
    internal_address: <REDACTED>
    user: ubuntu
    role: [etcd, controlplane, worker]

  - address: <REDACTED>
    internal_address: <REDACTED>
    user: ubuntu
    role: [etcd, controlplane, worker]

2. Once cluster is up and active, create a main.tf file with the following configurations, and run terraform init:

terraform {
  required_providers {
    rke = {
      source = "terraform.local/local/rke"
      version = "1.4.5-rc1"
    }
  }
}

provider "rke" {}

resource "rke_cluster" "rke_cluster" {}

3. Once terraform has been successfully initialized, import the standalone RKE cluster created in step 1, by running the following command:

terraform import rke_cluster.rke_cluster cluster.yml:cluster.rkestate

4. Verified - cluster successfully imports; as expected

Screenshot: