Skip to content

Latest commit

 

History

History
147 lines (110 loc) · 4.58 KB

README.md

File metadata and controls

147 lines (110 loc) · 4.58 KB
Raw

terraform-eks

Overview

This Terraform project facilitates the deployment of an AWS Elastic Kubernetes Service (EKS) cluster. The infrastructure is defined as code, allowing for easy provisioning and management of the Kubernetes cluster on AWS.

Project structure

.
├── eks-cluster
│   ├── eks-cluster.tf
│   ├── main.tf
│   ├── outputs.tf
│   ├── terragrunt.hcl
│   ├── variables.tf
│   ├── vars_dev.tfvars
│   ├── vars_prod.tfvars
│   └── vpc.tf
├── eks-config
│   ├── config_mongodb.tf
│   ├── ecr.tf
│   ├── main.tf
│   ├── namespaces.tf
│   ├── terragrunt.hcl
│   └── variables.tf
├── modules
│   └── ecr
│       ├── main.tf
│       ├── outputs.tf
│       └── variables.tf
├── README.md
└── terragrunt.hcl

The terraform infrastructure was created with the following key features:

  • Terraform best practices: the terraform state is stored remotely on s3 and a dynamoDB table was created for state locking.

  • Multiple terraform backend environments: the infrastructure was created with dynamic terraform backend generation based on the environment variables

TF_VAR_backend_s3_bucket
TF_VAR_backend_s3_region

This is perfect for AWS landing zones with multiple AWS environments so you define different GitHub/Gitlab environments with AWS keys for each account.

  • Dynamic .tfvars for each environment: By setting the environment variable TF_VAR_env_name to different values such as dev, qa, or prod, Terragrunt will automatically select the corresponding .tfvars file based on the naming convention (vars_<env_name>.tfvars).

For example, setting TF_VAR_env_name=dev will make Terragrunt use the vars_dev.tfvars file if it exists. This simplifies the management and configurations of the environment-specific variable

this Terragrunt config is located here: ./terragrunt.hcl

  • Terraform and Terragrunt version management: for maintaining consistency and compatibility within the project, the files .terraform-version and .terragrunt-version were added to set the versions, using tools like tfenv and tgswitch We can change between multiple projects reducing the risk of version conflicts or inconsistencies.

Prerequisites

Install terraform and terragrunt CLI tools

-terraform managed with tfenv https://github.com/tfutils/tfenv

-terragrunt managed with tgswitch https://github.com/warrensbox/tgswitch

Install/use the terraform and terragrunt correct version with the next command

Execute the commands in the root project folder in order to read the config files .terraform-version and .terragrunt-version

terraform

tfenv use

terragrunt

tgswitch

Then you have to create an s3 bucket and DynamoDB table for state-locking

aws dynamodb create-table \
         --region ap-south-1 \
         --table-name terraform-lock \
         --attribute-definitions AttributeName=LockID,AttributeType=S \
         --key-schema AttributeName=LockID,KeyType=HASH \
         --provisioned-throughput ReadCapacityUnits=1,WriteCapacityUnits=1

APPLY STEPS

  1. Export the next env vars to set up the terraform s3 backend and AWS creds
export AWS_ACCESS_KEY_ID=<AWS_ACCESS_KEY_ID>
export AWS_SECRET_ACCESS_KEY=<AWS_SECRET_ACCESS_KEY>
export AWS_DEFAULT_REGION='us-east-1'
export TF_VAR_env_name='dev'
export TF_VAR_backend_s3_bucket=<states-bucket>
export TF_VAR_backend_s3_region=<states-bucket-region>
export TF_VAR_dynamodb_table_state_locking=<DynamoDB-table-name>
  1. Apply (First time) following EKS blueprints pattern https://aws-ia.github.io/terraform-aws-eks-blueprints/getting-started/#deploy
cd eks-cluster
terragrunt init
terragrunt apply -target="module.vpc" -auto-approve
terragrunt apply -target="module.eks" -auto-approve
terragrunt apply -auto-approve
cd ../eks-config
terragrunt apply -auto-approve

This is only the first time

Then you can execute normally in the root folder

terragrunt run-all apply

This will run both components eks cluster and eks-config at the same time https://terragrunt.gruntwork.io/docs/features/execute-terraform-commands-on-multiple-modules-at-once/#the-run-all-command

DESTROY

Steps to Destroy all the resources

helm list -A

remove example API releases first

Then remove resources

cd eks-config
terragrunt destroy -auto-approve
cd eks-cluster
terragrunt destroy -target=module.eks_blueprints_addons -auto-approve
terragrunt destroy -target=module.eks -auto-approve
terragrunt destroy -auto-approve