Terraform deployment for Ansible in AWS

##Overview

It containts Terraform code to provision a new VPC and an Ansible server with either Dynamic or Static inventory.

Requirements

1. Terraform (version 0.8 or higher)
2. AWS account with allowed permissions to deploy services

Configuration

1. Cloned the repository using git clone https://github.com/lazam/terraform-ansible
2. Create a new non-version file named terraform.tfvars with the following details:

access_key      = "yourAWSAccessKey"
secret_key      = "yourAWSSecretKey"
aws_profile     = "default"
key_name        = "yourKeyPair"
key_path        = "pathtoKeyPair"
allowed_ip         = ["0.0.0.0/0"]   # your workstation IP, 0.0.0.0/0 if anywhere

For other configuration. Please see variables.tf for a list of variables and their default values.

Deployment Scenarios

Ansible with Dynamic Inventory

Ansible will use a Dynamic Invetory for this setup. Please see configuration below for using dynamic inventory:

variable "ansible_inventory" {
  description = "1 if dynamic inventory is true"

  type = "map"

  default = {
    dynamic_inventory = 1
    static_inventory  = 0
  }
}

When using dynamic inventory, you still need to manually export two environment variables:

export AWS_ACCESS_KEY_ID='AK123'
export AWS_SECRET_ACCESS_KEY='abc123'

Ansible with Static Inventory

Ansible will use the normal static inventory so each hosts need to be declared in /etc/ansible/host. Please see configuration below for using static inventory:

variable "ansible_inventory" {
  description = "1 if static inventory is true"

  type = "map"

  default = {
    dynamic_inventory = 1
    static_inventory  = 0
  }
}

For hosts declaration, I sugges to follow the best practice of layout here