iac-terraform-workshop

The following repo contains the source code for a brief workshop to learn how to provision and manage your infrastructure with terraform. If you find it useful feel free to use it.

Prerequisites

In order to be able to run these examples please be aware that you need a proper aws account and have terraform installed (v. 0.11 or higher).

Usage

Each commit in the repo is a different step towards having an overall basic knowledge of terraform features. The instructions here will detailed what you need at each commit or step.

Adding a VPC

In this step we just set up a basic vpc in order to run a basic plan and apply it. In order to be able to apply to run this step either:

• Create TF-VAR-access_key and TF-VAR-secret_key environmental variables and populate them with your aws credentials
• Add the variables in the terraform command, e.g.:

$> terraform apply -var 'access_key = <aws access key>' -var 'secret_key = <aws secret key>'

• Or create a file called terraform.tfvars and populate with the following code:

access_key = <aws access key>
secret_key = <aws secret key>

Also if you want to modify the default values just check out the file test.tfvars and modify it.

Adding a public subnet

In this step we are going to simply add a public subnet. Besides that there are a couple of new variables that can be modified. Have a look at the test.tfvars and input.tf.

Adding the service

In this step we are going to deploy the service that's going to be accessible. To do so, we also are going to create an auto scaling group and run all traffic through an ELB. As in the previous step we are adding a couple of variables to configure these new resources. Have a look at the test.tfvars and input.tf.

Refactoring into modules

In this step we are going to refactor and introduce a modules. Have a look at the different modules inside of the modules directory. Also you will be required to do a terraform init so that terraform can initialize the modules dependencies.

Workspaces and backing up the terraform state

In this step a good a idea would be to destroy your previous provisioned infrastructure and start from scratch.

After that, change directories and initialize the workspace or directory under bucket by running terraform init, and as in the initial stage, add a terraform.tfvars with your aws secret and access keys, and provision this infrastructure by applying it.

This will create a s3 bucket that will be used to hold the terraform state of the infrastructure we built before. Also pay attention to the definition of that s3 bucket. In a real situation you should set the prevent_destroy to true, because you don't want this bucket to be destroyed. This bucket will be used for the next part to store the terraform state, allowing multiple users or even your pipeline to have access to the latest terraform state.

Afterwards, if you look at the repository there are two new tfvars files. One for a test environment and other one for a prod environment. This is a simple way to manage different environment configurations. But remember terraform by default creates a default workspace, and each workspace manages its own state. So if you want to manage multiple environment without destroying or affecting them, you should create a workspace, and in each workspace you will be able to manage the state for that environment. For doing so, before applying any infrastructure change, please execute the following commands:

$> terraform workspace create test
$> terraform workspace create prod

After this, you should be able to switch between workspace by executing the following commands:

$> terraform workspace select <name of the workspace>

Now if you want to provision the test environment you could do the following

$> terraform workspace select test 
$> terraform apply -var-file=test.tfvars

Remember to switch to prod environment before apply it and use the prod.tfvars as the input var-file