Add bootstrap

[FrankYang0529]

issue: harvester/harvester#1797

Test plan

1. Create a new Harvester cluster and don't login.
2. Build terraform-harvester-provider.

CGO_ENABLED=0 go build -mod=vendor -o bin/terraform-provider-harvester .
mkdir -p ~/.terraform.d/plugins/registry.terraform.io/harvester/harvester/0.0.0-master/linux_amd64
cp bin/terraform-provider-harvester ~/.terraform.d/plugins/registry.terraform.io/harvester/harvester/0.0.0-master/linux_amd64/terraform-provider-harvester_v0.0.0-master

3. Create a temporary folder and put the following content in main.tf (remember to change api_url).

terraform {
  required_providers {
    harvester = {
      source = "harvester/harvester"
      version = "0.0.0-master"
    }
  }
}

provider "harvester" {
  alias = "bootstrap"

  api_url    = "https://192.168.3.131"
  bootstrap = true
}

resource "harvester_bootstrap" "admin" {
  provider = harvester.bootstrap

  initial_password = "admin"
  password = "passwdpasswd"
}

provider "harvester" {
  alias = "admin"

  api_url    = "https://192.168.3.131"
  kubeconfig = harvester_bootstrap.admin.kubeconfig
}

resource "harvester_volume" "test-disk" {
  provider = harvester.admin

  name      = "test-disk"
  namespace = "default"

  size = "1Gi"
}

resource "harvester_image" "ubuntu20" {
  provider = harvester.admin

  name      = "ubuntu20"
  namespace = "harvester-public"

  display_name = "ubuntu-20.04-server-cloudimg-amd64.img"
  source_type  = "download"
  url          = "http://cloud-images.ubuntu.com/releases/focal/release/ubuntu-20.04-server-cloudimg-amd64.img"
}

resource "harvester_virtualmachine" "ubuntu20" {
  provider = harvester.admin

  name                 = "ubuntu20"
  namespace            = "default"
  restart_after_update = true

  description = "test ubuntu20 raw image"
  tags = {
    ssh-user = "ubuntu"
  }

  cpu    = 2
  memory = "2Gi"

  efi         = true
  secure_boot = true

  run_strategy    = "RerunOnFailure"
  hostname        = "ubuntu20"
  reserved_memory = "100Mi"
  machine_type    = "q35"

  network_interface {
    name           = "nic-1"
    wait_for_lease = true
  }

  disk {
    name       = "rootdisk"
    type       = "disk"
    size       = "10Gi"
    bus        = "virtio"
    boot_order = 1

    image       = harvester_image.ubuntu20.id
    auto_delete = true
  }

  cloudinit {
    user_data    = <<-EOF
      #cloud-config
      password: 123456
      chpasswd:
        expire: false
      ssh_pwauth: true
      package_update: true
      packages:
        - qemu-guest-agent
      runcmd:
        - - systemctl
          - enable
          - '--now'
          - qemu-guest-agent
      EOF
    network_data = ""
  }
}

4. Run terraform init and terraform apply. Terraform should update password to passwdpasswd and create a new image and vm.

[futuretea]

Refer to harvester/harvester#3673, It seems like we won't disable MCM in 1.2.0. So users can use terraform-provider-rancher2 to do this.

harvester/harvester#1797 (comment)

[FrankYang0529]

Refer to harvester/harvester#3673, It seems like we won't disable MCM in 1.2.0. So users can use terraform-provider-rancher2 to do this.

harvester/harvester#1797 (comment)

Hi @futuretea, yes, users can use rancher2 directly if they enable the mcm feature. However, the mcm feature is an option for users. If they don't enable it, we still need to give users a choice to bootstrap with terraform harvester provider.

[belgaied2]

This PR is interesting because I am encountering issues where it is not possible to use the Harvester provider where the KUBECONFIG file does not yet exist. I get the following error during terraform plan phase:
Error: the server could not find the requested resource (get settings.harvesterhci.io server-version)

Is this solved somewhere? Can the plan phase be successful without an existing kubeconfig file ?

[FrankYang0529]

Hi @belgaied2, thanks for the testing. I updated the code. Now, it can handle terraform plan without an existing kubeconfig file. Thanks.