Infrastructure as code with Terraform

What is Terraform?

There are two sides of IAC
1: Configuration Management
2: Orchestration

Ansible - Configuration Management Terraform - Orchestration Tool Kubernetes - Orchestration is used in containerisation - i.e Docker, Crio, Rocket

Why should you integrate Terraform?

It can be ran on premise which is free of cost, we spend no money until we are ready to push it into production
An orchestration tool that allows for the creation of multiple cloud compute services all through one script

Initisialising Terraform

terraform init

Terraform commands

  terraform init - initisialises terraform
  terraform plan - checks the steps inside the code and lists the success or failures
  terraform apply - will implement the code, deploying the infrastructure
  terraform destroy - will destroy all the cloud services that have been created
  terraform output - This will output any of the variables that have been specified

Note it is good practice to run terraform plan before running apply so we can see any faults before we make those changes

if you run terraform apply after making changes, it will then reboot the instance with the changes in place e.g. any subnets or security groups we have added

Creating our cloud services

Our orchestration has been split into three files

main.tf will have the name of the cloud service provider we are using as well as the EC2 instances
the network.tf file will have the configuration for all the network related things created
the security.tf file contains all of the security related items for our VPC

Creating an image using AWS

When we run our instances we must set an AMI id in which they can build from, remember AMI's are the blueprint which allow us to create instances
We can create an image of one of our own EC2 images and use this ami-id to run additional EC2's or we can use the default Ubuntu linux ami

Terraform Output

Terraform output allows us to see the value of a variable that we can specify in our terraform script

resource "aws_instance" "app_instance" {
          ami          = "ami-0f6c07744b6402c0d"
          key_name = "DevOpsStudents"

# What tpye of ec2 instance we want to create?
          instance_type = "t2.micro"

# Assign public ip?
          associate_public_ip_address = true


# Linking security group to instance
          vpc_security_group_ids = [aws_security_group.allow_tls.id]
          subnet_id = aws_subnet.pub.id

          tags = {
              Name = var.name
          }
}


output "public_ip_app" {
  value = aws_instance.app_instance.public_ip
}

Here we are asking for the public ip of our app instance because whenever it is created we get a random ip
When we run terraform output, our public ip will be shown in the terminal

Using Ansible with Terraform

We can use Terraform to create our servers and then Ansible to provision them, this ensures that our development environment remains consistent and thus limiting the bugs we would come across
This can be done by creating a script that will run terraform apply followed by the ansible playbook that will provision both our app and db with the required dependencies to run our app, in this case the node application
We can download both terraform and ansible through an on premise Virtual machine and do all the work to our cloud computers from there

aosbornee/Terraform-With-AWS