Automation with Terraform

This guide will allow you to deploy your first instances on the Trust Cloud in less than 5 minutes.

Prerequisites

1. Have subscribed to the Cloud Temple offer (subscription to the IaaS offer).
2. Have the permissions enabled for the 'IaaS' object driver

Deploy a Virtual Machine via Terraform

In this section, we will see how to deploy a virtual machine on the Trust Cloud via the Cloud Temple Terraform provider in a few minutes. If you have not yet used the Cloud Temple provider, follow the instructions at this address to install it and authenticate to your tenant.

We will start by creating a .tf file that describes the instance we want to deploy. The following script allows you to deploy a virtual machine from scratch.

data "cloudtemple_compute_virtual_datacenter" "dc" {
  name = "DC-EQX6"
}

data "cloudtemple_compute_host_cluster" "flo" {
  name = "clu002-ucs01_FLO"
}

data "cloudtemple_compute_datastore_cluster" "koukou" {
  name = "sdrs001-LIVE_KOUKOU"
}

resource "cloudtemple_compute_virtual_machine" "scratch" {
  name = "from-scratch"

  memory                 = 8 * 1024 * 1024 * 1024
  cpu                    = 2
  num_cores_per_socket   = 1
  cpu_hot_add_enabled    = true
  cpu_hot_remove_enabled = true
  memory_hot_add_enabled = true

  datacenter_id                = data.cloudtemple_compute_virtual_datacenter.dc.id
  host_cluster_id              = data.cloudtemple_compute_host_cluster.flo.id
  datastore_cluster_id         = data.cloudtemple_compute_datastore_cluster.koukou.id
  guest_operating_system_moref = "amazonlinux2_64Guest"

  tags = {
    created_by = "Terraform"
  }
}

The parameters used in this script are as follows:

• datacenter_id (required): datacenter in which the virtual machine is deployed
• host_cluster_id (required): cluster in which the virtual machine is deployed
• name (required): name of the machine
• memory: RAM allocated to the machine initially
• cpu: number of vCPU allocated to the machine initially
• num_cores_per_socket: number of cores per socket
• datastore_cluster_id: datastore to which the machine is attached
• guest_operating_system_moref: guest operating system

Other parameters can be applied to a virtual machine during its deployment. You can find all these parameters on the following page of the Terraform documentation

Once the .tf file is created and saved, run the following command to verify your code:

terraform validate

Then, plan the deployment and check that the plan matches what you want to achieve:

terraform plan

Finally, deploy the virtual machine by running the following command:

terraform apply

Use cloud-init to configure a virtual machine deployed from the Terraform provider

The 'cloud-init' tool allows you to customize a virtual machine, or a cloud instance, during its first boot. It is a widely used standard. For more information, refer to the documentation

Compatibility

In order to be able to configure a virtual machine deployed from the Cloud Temple Terraform provider via cloud-init, the OVF used to deploy it must be compatible with cloud-init.

To check the compatibility of your virtual machine with cloud-init, enter the following command:

systemctl status cloud-init.service

If cloud-init is correctly installed on the machine, you should see a response like this. (See screenshot below)

If needed, you can find cloud-init compatible images on the internet (for example Ubuntu Cloud Image) or install it yourself on your machine before converting it to an OVF.

Deployment

Now that you are sure the deployed OVF is compatible with cloud-init, here is an example of a Terraform file (.tf) that you can use to configure your virtual machine. NB: All examples shown here can be found in the examples folder of the Cloud Temple Terraform provider repository here: Examples

main.tf

resource "cloudtemple_compute_virtual_machine" "ubuntu-cloud-init" {
  name = "ubuntu-cloud-init"

  memory                 = 8 * 1024 * 1024 * 1024
  cpu                    = 2
  num_cores_per_socket   = 1
  cpu_hot_add_enabled    = true
  cpu_hot_remove_enabled = true
  memory_hot_add_enabled = true

  datacenter_id   = data.cloudtemple_compute_virtual_datacenter.TH3S.id
  host_cluster_id = data.cloudtemple_compute_host_cluster.CLU001.id
  datastore_id    = data.cloudtemple_compute_datastore.DS003.id

  content_library_id      = data.cloudtemple_compute_content_library.local.id
  content_library_item_id = data.cloudtemple_compute_content_library_item.ubuntu-cloudimg.id

  power_state = "on"

  backup_sla_policies = [
    data.cloudtemple_backup_sla_policy.sla001-daily-par7s.id,
    data.cloudtemple_backup_sla_policy.sla001-weekly-par7s.id,
  ]

  cloud_init = {
    network-config = filebase64("./cloud-init/network-config.yml")
    user-data      = filebase64("./cloud-init/user-data.yml")
  }
}

network-config.yml

#cloud-config
network:
  version: 2
  ethernets:
    eth0:
      dhcp4: false
      addresses:
        - 172.16.100.192/24
      gateway4: 172.16.100.1
      nameservers:
        addresses:
          - 172.16.11.4

user-data.yml

#cloud-config
users:
  - default
  - name: terraform
    primary_group: users
    shell: /bin/bash
    sudo: ['ALL=(ALL) NOPASSWD:ALL']
    groups: sudo
    lock_passwd: false
    plain_text_passwd: password

The interesting element here is the presence of the cloud-init property, you can see that it is composed of two sub-properties: network-config, and user-data.

These last ones are part of a set of 7 properties that you can use to configure your virtual machine with cloud-init.

• user-data: This value must be encoded in base64 and contains notably configuration information for the user accounts of the virtual machine. You can also add scripts to install or update packages.
• network-config: This value must be encoded in base64 and contains notably network configuration information for the virtual machine.
• public-keys: Indicates that the instance should fill the 'authorized_keys' of the default user with this value.
• instance-id: Allows to define a unique instance identifier with cloud-init.
• password: If defined, the default user's password will be set to this value to allow password-based connection. The password will only be valid for one connection. If the value is 'RANDOM', a random password will be generated and displayed on the console.
• hostname: Specifies a hostname for the deployed instance.
• seedfrom: Allows to define a URL where cloud-init will go to retrieve the configuration files it must use. For more information on how cloud-init works, please refer to the official documentation

Execution

To verify the correct execution of cloud-init, you should be able to connect with the user you configured in the user-data.yml file or the hostname should change to be set to the one you put in 'hostname'.

In case of issues, you can check the cloud-init logs using the following command:

sudo cat /var/log/cloud-init-output.log

You should see various information about the execution of cloud-init. In the following screenshot, it can be seen that the network configuration was successfully completed.