This repo is about my learning from terraform.

Sample code

• Sample code is under the modules folder in this project. Its not extensive but basic stuff.

Terrform folder structure

- modules/project
  - main.tf
  - variables.tf
  - output.tf

Provider

• Represent any of the cloud provider
• Provider block allows you to define provider and the configuration for it.

provider "aws" {
  region = "eu-west-2"
}

Backend

• Terraform stores is state locally in the file called terraform.tfstate.
• Storing Terraform state locally is good only when doing local development.
• In real life and production environment, you will like to store your Terraform state in some remove repo so that you and your team can collaborate and your infrastructure is consistent.
• Backend is a concpet which is used to define a remote storage for your tfstate file.
• Terraform provides various option which can used as Backend. e.g local, artifactory, azure, etcd, gcs, http, inmem, s3
• Below is the syntax for your backend definition.

terraform {
  backend "s3" {
    bucket = "sj-tf-state-bucket-uk" # name of the s3 bucket (exisiting) which will be used to store the tfstate file
    region = "eu-west-2" # region of your bucket.
    key = "sj-test-tf.tfstate" # name of the terraform state file
  }
}

• Thing to pay attention is that which ever resource you use to store the tfstate file should already exist. In above example the bucket sj-tf-state-bucket-uk should already present.
• You should pre-create your backend resource either manually or through cli.

resource

• Different resource avaliable from your cloud provider.
• e.g aws has resource like ec2-instance, vpc, elb etc
• terrafrom has its own naming convention but all aws resource will have corresponding option in terrafrom

resource "aws_instance" "my_test_ec2" {
  ami = "ami-43242fsd"
  instance_type = "t2-micro"
}

data

• Every resource has some metadata about it. E.g every ec2 instance has ami, instance_type, key_pair etc.
• If you want to have know and use this information, data is the terrafrom way to do it.
• Using data to get information about a resource is two step process.
• First you have to tell which type of resource you are to get infromation about.
• Second then use filters to narrow down the to the specific resource, about which you need information.

data "aws_instance" "myec2data" {
  tags = {
    name = "welcome"
  }
}

• In example above we are trying to find metadata about a aws ec2 instance which is taggged as name = "welcome"
• using data is simple.

  resource "aws_instance" "myec2instance" {
  ami = "ami-43242fsd"
  instance_type = "t2-micro"
  key_name = data.aws_instance.myec2data.key_name
}

variable

• This is terraform way to make your script dynamic. Allows values to be passed on fly.
• You can define variables using variable tf resource and then use it in your other tf resrouces.
• variable can have default values or you can pass them as argument to your tf script.

variable "myec2tag" {
  default = "awesometag"
  description = "this is the awesome way to tag a ec2 instance"
  type =  string
}

• using variable in your main tf script

resource "aws_instance" "myec2instance" {
  ami = "ami-43242fsd"
  instance_type = "t2-micro"
  tags = {
    name = var.myec2tag
  }
}

• passing variable from command line

terrafrom plan -var="myec2tag=besttagever"

locals

• locals are like constant in programming language.
• They allow us to have central places to define common values which be used accross all other terrafrom resource.

locals {
  ec2-instnce-type = "t2.micro"
}

• using locals is simple

resource "aws_instance" "myec2instance" {
  ami = "ami-43242fsd"
  instance_type = local.ec2-instnce-type
}

tfvars

• tfvars are ways to provide values to your variables defined.
• Come handy when you have hunderds of variables as using `--var="var1=value1" wont be visuablly astetic nor easy to write.
• tfvars allow you to create a file and define values for your variables.
• you can pass this file to your tf script as below
• tfvars file is key value pair file, where key is your variable name and value si the value.
• tf file are stored with extention tfvars. Sample file dev.tfvars will look like this

myec2tag = dev-instance

• using tfvars file in your terrafrom plan/apply

terrafrom plan -var-file="dev.tfvars"

variable interpolation

• interpolation systax is "${something}"
• variable interpolation was primarly used in Terraform 0.11 and has a sysntax like below

"${var.myec2tag}"

• From Terraform 12 you can use below syntax for interpolation.

var.myec2tag

• can used in the scenario when you have to prefix or suffix someting to a variable or want to perform some conditional assignment.

Expressions

• Read this

Comman Parameters

• count to iterate/loop over a resurce
• for_each
• for

Command inbuilt function

• length to determine the lenght of any array, string, etc
• upper to change the case of a string to uppper
• toset to flatten a arrar
• map
• file allow us to refer to a file on local system as input value

Useful terrafrom commands

• terraform fmt - format our terrafrom files.

Modules

• Modules are primarly used for reuse of code.
• When using module your terrafrom project will have below structure

- Main Project
  - main.tf
  - variables.tf
  - outputs.tf
  - vpc_module
    - main.tf
    - variables.tf
    - outputs.tf
  - .
  - .
  - ModuleN
    - main.tf
    - variables.tf
    - outputs.tf

• Modules can be used in the project main.tf as below

module "vpc_module" {
  source              = "./module"
  private_subnet_cidr = var.private_subnet_cidr # variables defined in the vpc_module
  public_subnet_cidr  = var.public_subnet_cidr # variables defined in the vpc_module
  vpc_cidr_block      = var.vpc_cidr_block # variables defined in the vpc_module
}

• If you have defined variables in your modules and have default values assing to them then you have to provide those variables when using those modules in project main.tf