Kubernetes Terraform

Quickstart

cd ubuntu/instance/test
terraform init
terraform apply

How to create a new deployment

• copy _template to a directory of your choice (i.e. ubuntu)
• remove all resources that you do not require (use syntax highlighting to fix whatever breaks afterwards)
  • deployment.tf / statefulset.tf / daemonset.tf (keep one)
  • ingress.tf
  • pvc.tf
  • configEnv.tf
  • configVolume.tf
  • secretEnv.tf
  • secretVolume.tf
• replace all occurrences of TODO
• create as many instances as you require (i.e. dev, test, staging, production)
• make non-generic adjustments to your resources (i.e. add command to deployment.tf)

General

It is not recommended to take this repository as is as a base for all of your terraform kubernetes code! This repository is merely to provide a general approach that can be used to define a structure fitting your use case.

Recommendations

• check the inputs and outputs of _data and adjust them to fit your use case (i.e. replace linkerd annotations with istio)
• check the code in _template and adjust it to fit your use case (i.e. make resource requests/limits variable)
• keep each directory as a separate repository and either use the repository (including release tag) as module reference, or write a gitfile instead
• whenever you make a change to your _template run a diff over all repositories and apply the structure change in all of your repositories

Naming Conventions

The code in here makes use of 3 different conventions to make it easier to differentiate between the meaning of that particular part. Main reason for that is, that terraform normally recommends snake_case, but kubernetes resources do not support snake_case.

snake_case:

• terraform internal structures (i.e. resource identifiers)

kebab-case:

• kubernetes resource names (i.e. namespace name)

camelCase:

• everything that's not kubernetes or terraform relevant (i.e. terraform resource names)