How to Use Terraform – Articles

Learn how to use Terraform with the Public Cloud by way of practical examples.

OpenStack is an open-source cloud operating system for building and managing public and private cloud computing platforms. The OpenStack software components are the foundation of the OVHcloud Public Cloud infrastructure.

The open-source tool Terraform was developed to make the creation of complex cloud infrastructures easier. It enables your infrastructure's properties to be abstracted in text files which can be used as a basis to deploy your infrastructure.

As an example, this video shows how to easily scale the number of instances while keeping your existing infrastructure by changing only one line of code:

Requirements

This tutorial is compatible with Terraform version 0.14.0 and later.

Instructions

Creating the Terraform environment

After the Terraform installation, create a directory for all text files that will describe your infrastructure:

mkdir test_terraform && cd test_terraform

You can now create a Terraform environment with the following command. This will allow you to create and manage the evolution of your infrastructure.

terraform workspace new test_terraform

Creating resources

Creating a provider

In Terraform, you specify "providers" for your cloud environment. A "provider" (such as OVHcloud) hosts your OpenStack infrastructure resources.

Create a file named provider.tf with the following content:

# Define providers and set versions
terraform {
required_version    = ">= 0.14.0" # Takes into account Terraform versions from 0.14.0
  required_providers {
    OpenStack = {
      source  = "terraform-provider-openstack/openstack"
      version = "~> 1.42.0"
    }
 
    ovh = {
      source  = "ovh/ovh"
      version = ">= 0.13.0"
    }
  }
}

# Configure the OpenStack provider hosted by OVHcloud
provider "OpenStack" {
  auth_url    = "https://auth.cloud.ovh.us/v3/" # Authentication URL
  domain_name = "default" # Domain name - Always at 'default' for OVHcloud
  alias       = "ovh" # An alias
}

provider "ovh" {
  alias              = "ovh"
  endpoint           = "ovh-us"
  application_key    = "<your_application_key>"
  application_secret = "<your_application_secret>"
  consumer_key       = "<your_consumper_key>"
}

If you don't want to define your secrets in the Terraform configuration file, you can also define them in environment variables as shown below. Replace the actual keys/secret with the variables in the above file if the variables are defined separately (i.e. application_key = $OVH_APPLICATION_KEY).

$ export OVH_ENDPOINT=ovh-us
$ export OVH_APPLICATION_KEY=Your_key_application_OVH(or_AK)
$ export OVH_APPLICATION_SECRET=Your_secret_application_key_OVH(or_AS)
$ export OVH_CONSUMER_KEY=Your_token(or_CK)

The "alias" is a unique identifier for a provider. For example, if you have two OpenStack providers with different credentials, you must specify each provider in the resource.

You now need to create a new OpenStack user, and the generate the OpenRC file ontaining all of the credentials you want to export as environment variables.

To generate the OpenRC file, log in to the Horizon interface.

Ensure you are in the Overview section.
Click on your User ID.
Select OpenStack RC File.

Load this file, then enter the password for the user you created earlier:

$ source openrc.sh
Please enter your OpenStack Password:

You now need to initialize your workspace to download the provider plugins:

terraform init

Creating an instance

In Terraform, a "resource" is a component of your infrastructure. This can be an instance, a storage block delivered by the OpenStack provider, or a network delivered by the OVHcloud provider.

To create an instance, you need at least:

An instance name
An image
A flavor
An SSH key

For example purposes, we will create a simple instance with Ubuntu 22.04 with the flavor b2-7, and import an SSH key. Add the following lines to a file named simple_instance.tf:

# Creating an SSH key pair resource
resource "openstack_compute_keypair_v2" "test_keypair" {
  provider   = openstack.ovh # Provider name declared in provider.tf
  name       = "test_keypair" # Name of the SSH key to use for creation
  public_key = file("~/.ssh/id_rsa.pub") # Path to your previously generated SSH key
}
 
# Creating the instance
resource "openstack_compute_instance_v2" "test_terraform_instance" {
  name        = "terraform_instance" # Instance name
  provider    = openstack.ovh  # Provider name
  image_name  = "ubuntu 22.04" # Image name
  flavor_name = "b2-7" # Instance type name
  # Name of openstack_compute_keypair_v2 resource named keypair_test
  key_pair    = openstack_compute_keypair_v2.test_keypair.name
  network {
    name      = "Ext-Net" # Adds the network component to reach your instance
  }
}

To see what will be added, created, or deleted in your infrastructure, you can execute:

terraform plan

You can enter the following command to import your SSH key and create your first instance:

terraform apply

The names of the SSH key and instance are different than the examples above, but your output should look like this:

Now, log in to the OVHcloud Control Panel, go to the Public Cloud section, and click on Instances. As you can see, your compute instance named "terraform_instance" is created.

Creating a second, identical instance with terraform apply will not work. Terraform applies change only if it recognizes a difference in your configuration files or a new file.

Creating multiple instances

In this section, we will create an Ubuntu instance of the flavor "s1-2" in three different regions.

You can find all region names by checking this OVHcloud API endpoint:

GET cloud/project/{serviceName}/region

We will use the following OVHcloud regions for this example:

US-EAST-VA-1
US-WEST-OR-1

You can create three resources named "openstack_compute_instance_v2" and change the region parameter for each. It can however become difficult to manage files with a large amount of identical resources.

A better method is to use the resource meta parameter called "count". It allows you to tell Terraform to create the same resource several times.

To do this, we will create a file named multiple_instance.tf. In it, we first define a variable containing the three regions, and then add an instance creation counter:

# Create a region variable containing the list of OVHcloud regions
# It will be used to iterate over different regions in order to
# start an instance on each of them.
 variable "region" {
   type = list
   default = ["US-EAST-VA1", "US-WEST-OR1"]
 }
  
# Creating an SSH key pair
 resource "openstack_compute_keypair_v2" "test_keypair_all" {
   count = length(var.region)
   provider = openstack.ovh # Specify provider name
   name = "test_keypair_all" # Name of the SSH key
   public_key = file("~/.ssh/id_rsa.pub") # Your SSH key path
   region = element(var.region, count.index)
 }
  
 # Create a resource that is an OpenStack instance in each region
 resource "openstack_compute_instance_v2" "instances_on_all_regions" {
   # Number of times the resource will be created
   # defined by the length of the list named region
   count = length(var.region)
   provider = openstack.ovh # Provider name
   name = "terraform_instances" # Instance name
   flavor_name = "b2-7" # Instance flavor
   image_name = "ubuntu 22.04" # Image name
   # element is a function that accesses the element at the position
   # count.index of the list var.region. It allows to iterate between regions
   region = element(var.region, count.index)
   # Accesses the name of the openstack_compute_keypair_v2 resource variable named test_keypair
   key_pair = element(openstack_compute_keypair_v2.test_keypair_all.*.name, count.index)
   network {
     name = "Ext-Net" # Adds the public network to your instance
   }
 }

Check the changes you need to make to your infrastructure using the following command:

terraform plan

Apply your changes with the following command:

terraform apply

Terraform can create multiple instances with this method but you can use it to modify your current infrastructure as well.

Modifying an instance

In this example, we will attach a new storage volume to our first instance. Open and edit the file that you created previously in the "create simple instance" step named simple_instance.tf, then add the following lines:

# Create a resource for storage
resource "openstack_blockstorage_volume_v2" "volume_to_add" {
  name = "simple_volume" # Volume name
  size = 10 # Volume size in GB
  provider = openstack.ovh # Provider name
}
 
# Attach the volume created previously to the instance
resource "openstack_compute_volume_attach_v2" "attached" {
  # ID of openstack_compute_instance_v2 resource named test_terraform_instance
  instance_id = openstack_compute_instance_v2.test_terraform_instance.id
  # ID of openstack_blockstorage_volume_v2 resource named volume_to_add
  volume_id = openstack_blockstorage_volume_v2.volume_to_add.id
}

Check the changes you need to make to your infrastructure using the following command:

terraform plan

Apply your changes with the following command:

terraform apply

Creating an instance in the OVHcloud network (vRack)

The Terraform OVHcloud plugin can manage private networks, private subnets, Public Cloud users, and vRack attachments. In this part, we will focus on the network creation.

Create a file create_private_network_instance.tf and enter the following:

variable "service_name" {
  default = "OS_PROJECT_ID" # Replace with your service name / OS_PROJECT_ID value
}

variable "project_id" {
  default = "OS_TENANT_ID" # Replace OS_TENANT_ID with your Project Tenant ID
}

variable "region" {
  default = "US-EAST-VA1" # Replace with your region, for example with the environment variable OS_REGION_NAME
}

# Associating cloud project to vRack
 resource "ovh_vrack_cloudproject" "vcp" {
  service_name = var.service_name
  project_id   = var.project_id
} 

 # Creating a private network
 resource "ovh_cloud_project_network_private" "network" {
    service_name = var.service_name
    name         = "private_network" # Network name
    regions      = [var.region]
    provider     = ovh.ovh # Provider name
    vlan_id      = 168 # VLAN ID for vRack
    depends_on   = [ovh_vrack_cloudproject.vcp] # Depends on the vRack's association with the cloud project
 }
  
 # Creating a subnet using the previously created private network
 resource "ovh_cloud_project_network_private_subnet" "subnet" {
    service_name = var.service_name
    # ID of the ovh_cloud_network_private resource named network
    network_id   = ovh_cloud_project_network_private.network.id
    start        = "192.168.168.100" # First IP of the subnet
    end          = "192.168.168.200" # Last IP of the subnet
    network      = "192.168.168.0/24" # Subnet IP address location
    dhcp         = true # Enables DHCP
    region       = var.region
    provider     = ovh.ovh # Provider name
    no_gateway   = true # No default gateway
 }
  
 # Creating an instance with 2 network interfaces
 resource "openstack_compute_instance_v2" "proxy_instance" {
   provider     = openstack.ovh # Provider name
   name         = "proxy_instance" # Instance name
   image_name   = "ubuntu 22.04" # Image name
   flavor_name  = "b2-7" # Flavor name
   # Name of openstack_compute_keypair_v2 resource named keypair_test
   key_pair     = openstack_compute_keypair_v2.test_keypair.name
   # Add public and private network
   network {
        name    = "Ext-Net"
    }
   network {
        name    = ovh_cloud_project_network_private.network.name
    }
 }

This instance creation is linked to the opestack_computer_keypair_v2.test_keypair resource you previously created in this guide.

Check the changes you need to make to your infrastructure using the following command:

terraform plan

Apply your changes using the following command:

terraform apply

In your Public Cloud project, you will see a new instance appear, with a public and private interface.

Creating an infrastructure for a website

In this example, we will create a basic website infrastructure using Terraform and the OVHcloud private network. The components created are:

a private network
a subnet
two instances with two network interfaces each: the first one public and the second one private
an instance with a private interface and two additional disks

Create a file named simple_web_site.tf and enter the following lines:

variable myregion {
  default = "US-EAST-VA1" # Replace with the OS_REGION_NAME environment variable content
}

# Creating a private network
resource "ovh_cloud_project_network_private" "private_network" {
  service_name  = var.service_name
  name          = "backend" # Network name
  regions       = [var.myregion] 
  provider      = ovh.ovh # Provider name
  vlan_id       = 42 # vRack vlan ID
  depends_on    = [ovh_vrack_cloudproject.vcp] # Depends on vRack being associated with the cloud project
}
 
# Creating a private subnet
resource "ovh_cloud_project_network_private_subnet" "private_subnet" {
  # ID for the ovh_cloud_network_private resource named private_network
  network_id    = ovh_cloud_project_network_private.private_network.id
  service_name  = var.service_name
  region        = var.myregion
  network       = "192.168.42.0/24" # Subnet IP
  start         = "192.168.42.2" # First IP of the subnet
  end           = "192.168.42.200" # Last IP of the subnet
  dhcp          = false # Disabling DHCP
  provider      = ovh.ovh # Provider name
  no_gateway    = true # No default gateway
}
 
# Search for the latest Archlinux image
data "openstack_images_image_v2" "archlinux" {
  name          = "Archlinux" # Image name
  most_recent   = true # Limits search to the most recent
  provider      = openstack.ovh # Provider name
}
 
# List of possible private IP addresses for front-ends
variable "front_private_ip" {
  type          = list(any)
  default       = ["192.168.42.2", "192.168.42.3"]
}
 
# Create 2 instances with 2 network interfaces
resource "openstack_compute_instance_v2" "front" {
  count           = length(var.front_private_ip) # Number of instances to create
  provider        = openstack.ovh # Provider name
  name            = "front" # Instance name
  key_pair        = openstack_compute_keypair_v2.test_keypair.name
  flavor_name     = "b2-7" # Instance type name
  image_id        = data.openstack_images_image_v2.archlinux.id # Instance image ID
  security_groups = ["default"] # Adds the instance to the security group
  network {
    name = "Ext-Net" # Public network interface name
  }
  network {
    # Private network interface name
    name = ovh_cloud_project_network_private.private_network.name
    # IP address taken from the list defined earlier
    fixed_ip_v4 = element(var.front_private_ip, count.index)
  }
  depends_on = [ovh_cloud_project_network_private_subnet.private_subnet] # Depends on private network
}
 
# Create an attachable storage device for the backup (volume)
resource "openstack_blockstorage_volume_v2" "backup" {
  name     = "backup_disk" # Name of storage device
  size     = 10 # Size
  provider = openstack.ovh # Provider name
}
 
# Create an instance with a network interface and storage device
resource "openstack_compute_instance_v2" "back" {
  provider        = openstack.ovh # Provider name
  name            = "back" # Instance name
  key_pair        = openstack_compute_keypair_v2.test_keypair.name
  flavor_name     = "b2-7" # Instance type name
  image_id        = data.openstack_images_image_v2.archlinux.id # Instance image ID
  security_groups = ["default"] # Adds the instance to the security group
  network {
    name        = ovh_cloud_project_network_private.private_network.name # Private network name
    fixed_ip_v4 = "192.168.42.150" # Private IP address chosen arbitrarily
  }
  # Bootable storage device containing the OS
  block_device {
    uuid                  = data.openstack_images_image_v2.archlinux.id # Instance image ID
    source_type           = "image" # Source type
    destination_type      = "local" # Destination
    volume_size           = 10 # Size
    boot_index            = 0  # Boot order
    delete_on_termination = true  # The device will be deleted when the instance is deleted
  }
  # Storage device
  block_device {
    source_type           = "blank" # Source type
    destination_type      = "volume" # Destination
    volume_size           = 20 # Size
    boot_index            = 1 # Boot order
    delete_on_termination = true # The device will be deleted when the instance is deleted
  }
  # Previously created storage device
  block_device {
    uuid                  = openstack_blockstorage_volume_v2.backup.id # Storage Device ID
    source_type           = "volume" # Source type
    destination_type      = "volume" # Destination
    boot_index            = 2 # Boot order
    delete_on_termination = true # The device will be deleted when the instance is deleted
  }
  depends_on = [ovh_cloud_project_network_private_subnet.private_subnet] # Depends on private network
}

Check the changes you need to make to your infrastructure using the following command:

terrform plan

Apply your changes with the following command:

terraform apply

Deleting an infrastructure

To remove every resource created through Terraform, you can enter the following command:

terraform destroy

Conclusion

Terraform is a useful tool for the creation of cloud infrastructures. Having read this guide, you should be able to create a Terraform environment and customize it to your needs, including creating multiple kinds of resources for your environment and deleting an infrastructure.

Articles in this section

Requirements

Instructions

Creating the Terraform environment

Creating resources

Creating a provider

Creating an instance

Creating multiple instances

Modifying an instance

Creating an instance in the OVHcloud network (vRack)

Creating an infrastructure for a website

Deleting an infrastructure

Conclusion

Related articles