Resources and Data Sources

Resources and data sources are the building blocks of every Terraform configuration. Resources create and manage infrastructure, while data sources read existing information. This lesson covers how to define, reference, and connect them.

---

Resources

A resource block declares a piece of infrastructure that Terraform will manage:

resource "aws_instance" "web" {
  ami           = "ami-0c55b159cbfafe1f0"
  instance_type = "t3.micro"

  tags = {
    Name = "web-server"
  }
}

Resource Address

Every resource has a unique address: <type>.<name>

• aws_instance.web — the resource above
• google_compute_instance.api — a GCP compute instance named "api"

Resource Arguments vs Attributes

Term	When Set	Example
Argument	In your configuration (input)	instance_type = "t3.micro"
Attribute	After creation (output)	aws_instance.web.public_ip

You reference attributes of one resource in another to create implicit dependencies.

---

Meta-Arguments

Resources support special meta-arguments that change their behaviour:

count

Create multiple instances of a resource:

resource "aws_instance" "web" {
  count         = 3
  ami           = "ami-0c55b159cbfafe1f0"
  instance_type = "t3.micro"

  tags = {
    Name = "web-${count.index}"
  }
}

Access individual instances: aws_instance.web[0], aws_instance.web[1], etc.

for_each

Create instances from a map or set:

resource "aws_s3_bucket" "buckets" {
  for_each = toset(["logs", "data", "backups"])
  bucket   = "${var.prefix}-${each.key}"
}

Access individual instances: aws_s3_bucket.buckets["logs"].

depends_on

Explicitly declare dependencies when Terraform cannot infer them:

resource "aws_instance" "web" {
  ami           = "ami-0c55b159cbfafe1f0"
  instance_type = "t3.micro"

  depends_on = [aws_iam_role_policy.web_policy]
}

provider

Use a non-default provider configuration:

resource "aws_instance" "eu_web" {
  provider      = aws.eu
  ami           = "ami-0abc12345def67890"
  instance_type = "t3.micro"
}

---

count vs for_each

Feature	count	for_each
Input	Number	Map or Set
Index	count.index (integer)	each.key / each.value
Reordering	Shifting indices causes recreation	Stable — keyed by name
Best for	Identical resources	Resources with unique attributes

Tip: Prefer for_each over count when resources have distinct identities. Removing an item from a count list shifts all subsequent indices, causing unnecessary resource recreation.

---

Data Sources

A data source reads information from an existing resource that Terraform does not manage:

data "aws_ami" "ubuntu" {
  most_recent = true
  owners      = ["099720109477"]  # Canonical

  filter {
    name   = "name"
    values = ["ubuntu/images/hvm-ssd/ubuntu-jammy-22.04-amd64-server-*"]
  }
}