devops4me/terraform repository overview

⁠How to Run Terraform in Docker

⁠terraform docker use cases

A number of use cases lend themselves well to running terraform inside a docker container. These include

• running projects in Terraform 0.12 when your laptop has Terraform 0.11 installed
• multiple infrastructures with state managed in different docker volumes
• dockerizing terraform so that Jenkins can run it without polluting it with Terraform binaries
• delivering to a colleague (a terraform codebase) has a higher chance of working than guessing what is in their environment
• Terraform local_exec may demand python, curl etc on the host environment - docker is perfect for this
• using docker volumes (and maybe volume drivers) to manage Terraform state if you don't want to use S3 or Dynamo DB

⁠Terraform Docker Example | Creating VPCs in AWS

Our example will use a simple terraform module in github that creates VPCs⁠.

⁠Step 1 | git clone into docker volume

First we create a docker volume (called vol.tfstate) and add the terraform module code to it by way of an alpine git container.

docker volume create vol.tfstate
docker run --interactive \
           --tty         \
	   --rm          \
	   --volume vol.tfstate:/terraform-work \
	   alpine/git \
	   clone https://github.com/devops4me/terraform-aws-vpc-network /terraform-work
sudo ls -lah /var/lib/docker/volumes/vol.tfstate/_data

When you list the files in the container you will see the terraform module's contents there.

⁠Step 2 | terraform init via docker

As our volume contains the terraform module code from git we are now ready to perform a terraform init. We use the devops4me/terraform container⁠ container which adds a VOLUME mapping to the hashicorp/terraform⁠ container at the /terraform-work location.

example - there is a working example directory⁠ in the git terraform module that demonstrates module use and is used by continuous integration actors.

docker run --interactive \
           --tty \
	   --rm \
	   --name vm.terraform \
	   --volume vol.tfstate:/terraform-work \
	   devops4me/terraform init example
sudo ls -lah /var/lib/docker/volumes/vol.tfstate/_data

The directory listing verifies that our volume now contains a .terraform directory.

⁠Step 3 | terraform apply via docker

At last we can run the terraform apply. Provide a role arn only if your organization works with roles alongside the other 3 AWS authentication keys.

docker run --interactive \
           --tty \
	   --rm \
	   --name vm.terraform \
	   --env AWS_DEFAULT_REGION=<<aws-region-key>> \
	   --env AWS_ACCESS_KEY_ID=<<aws-access-key>> \
	   --env AWS_SECRET_ACCESS_KEY=<<aws-secret-key>> \
	   --env TF_VAR_in_role_arn=<<aws-role-arn>> \
	   --volume vol.tfstate:/terraform-work \
	   devops4me/terraform apply -auto-approve example
sudo ls -lah /var/lib/docker/volumes/vol.tfstate/_data

Examining the docker volume should reveal a tfstate file which documents the state of your infrastructure just after the terraform apply execution.

⁠Step 4 | terraform destroy via docker

After running plan and apply either once or multiple times you may feel the need to terraform destroy the infrastructure.

docker run --interactive \
           --tty \
	   --rm \
	   --name vm.terraform \
	   --env AWS_DEFAULT_REGION=<<aws-region-key>> \
	   --env AWS_ACCESS_KEY_ID=<<aws-access-key>> \
	   --env AWS_SECRET_ACCESS_KEY=<<aws-secret-key>> \
	   --env TF_VAR_in_role_arn=<<aws-role-arn>> \
	   --volume vol.tfstate:/terraform-work \
	   devops4me/terraform destroy -auto-approve example
sudo ls -lah /var/lib/docker/volumes/vol.tfstate/_data

Verify the destroy from your AWS console and also check that your volume now has a tfstate backup file created by terraform.