Purpose

This repo can be used to stand up various parts of your deployment on DigitalOcean using Terraform and configured using Ansible. The following components are included:

• HAProxy load balancers (active-passive configuration) with a floating IP
• Backend nodes
• MariaDB Galera cluster
• Minio object storage

Prerequisites

• You'll need to install Terraform which will be used to handle Droplet provisioning.
• In order to apply configuration changes to the newly provisioned Droplets, Ansible v2.2.x needs to be installed.
• Ansible's inventory will be read from Terraform's local state file, so you'll need terraform-inventory.
• We're going to need a DigitalOcean API key. The steps to generate a DigitalOcean API key can be found here.
• Use the included gen_auth_key script to generate an auth key for your load balancing cluster.

Configuring

Use terraform.tfvars to store values required such as API key, project name, and SSH data. The sample file terraform.tfvars.sample has been supplied, just remember to remove the appended .sample. Once you have your all of the variables set, Terraform should be able to authenticate in order to deploy your Droplets.

All blocks are uncommented in main.tf so all modules will be used. You'll be able to edit the module arguments for things like Droplet size and number of Droplets.

module "galera-cluster" {
  source           = "./modules/galera-cluster"
  project          = "${var.project}"
  region           = "${var.region}"
  keys             = "${var.keys}"
  private_key_path = "${var.private_key_path}"
  ssh_fingerprint  = "${var.ssh_fingerprint}"
  public_key       = "${var.public_key}"
}

This should allow you to spin up your Droplets and you'll be able to use terraform-inventory to create your dynamic inventory listing for use by Ansible.

Deploying

We'll start by using Terraform. Make sure you head back to the repository home. You'll need to execute terraform get to load the modules into the .terraform directory. You can then run a quick check and create an execution plan by running terraform plan.

Once you're ready, use terraform apply to begin deploying your Droplets. This should take about a minute or two depending on how many nodes you're spinning up. Once it finishes up, wait about 30 seconds for the cloud-config commands that were passed in to complete. You can run a quick check by running ansible all -i /usr/local/bin/terraform-inventory -m ping. If all pings get a return response of 'pong', you should be good to go.

When you're ready to begin configuring your Droplets, refer to the documentation for each Ansible role.

Follow-up steps

Keep in mind that this is a very basic setup, but the template files and tasks can easily be altered to suit your needs. If you already have Droplets provisioned, you can import them into Terraform, as well as create an image manually or by using a tool like Packer and use the image ID to spin up additional nodes. Any additional configuration can simple by done by creating a simple Ansible role, or modifying the existing ones.