Master Terraform’s count Meta‑Argument: Bulk, Conditional, and Blue‑Green Deployments

Introduction

In the journey of Infrastructure as Code (IaC), Terraform’s count meta‑argument acts as a “magician” that lets you create multiple identical resources efficiently instead of copying and pasting resource blocks.

How count Works

count

can be applied to any resource or module block. When you set count = N, Terraform treats the block as a list of N instances rather than a single object. If count is omitted, its default value is 1.

# Create 3 AWS S3 buckets
resource "aws_s3_bucket" "data_store" {
  count = 3

  # Ensure each bucket name is unique
  bucket = "my-app-data-store-${count.index}"
}

In the example above, aws_s3_bucket.data_store becomes a list of three bucket objects. To reference the first bucket’s ID you use aws_s3_bucket.data_store[0].id.

Scenario 1: Bulk Creation of Homogeneous Resources

Using count together with count.index (a zero‑based index) you can generate many similar resources, such as subnets across multiple Availability Zones.

variable "availability_zones" {
  description = "A list of Availability Zones to create subnets in."
  type        = list(string)
  default     = ["us-east-1a", "us-east-1b", "us-east-1c"]
}

resource "aws_subnet" "private" {
  count = length(var.availability_zones)

  vpc_id            = "vpc-12345678"
  availability_zone = var.availability_zones[count.index]
  cidr_block        = "10.0.${100 + count.index}.0/24"

  tags = {
    Name = "private-subnet-${count.index}"
  }
}

Here count is set to the length of the AZ list, and count.index drives the AZ selection, CIDR calculation, and unique naming.

Scenario 2: Conditional Resource Creation

Combine count with a ternary expression to create a resource only when a condition is true, e.g., only in production.

variable "is_production" {
  description = "A boolean flag to indicate if the environment is production."
  type        = bool
  default     = false
}

resource "aws_db_instance" "database" {
  # Create the DB only when is_production is true
  count = var.is_production ? 1 : 0

  allocated_storage = 20
  engine            = "mysql"
  engine_version    = "8.0"
  instance_class    = "db.t2.micro"
  db_name           = "mydb"
  username          = "admin"
  password          = "yourpassword"
  skip_final_snapshot = true
}

If var.is_production is false, the aws_db_instance.database resource becomes an empty list and any reference to it must be guarded accordingly.

Scenario 3: Blue‑Green Deployment Switch

By toggling count you can control which of two Auto Scaling Groups (blue or green) receives a load‑balancer listener, effectively switching traffic between environments.

Pitfalls of count

The main weakness of count is its reliance on numeric indexes. Removing an item from the middle of a list causes Terraform to shift subsequent indexes, leading to unnecessary destroy‑and‑create actions.

For example, a list of three servers indexed as server[0], server[1], server[2]. Deleting server[1] makes Terraform treat the original server[2] as server[1], destroying the old instance and creating a new one.

Modern Alternative: for_each

The for_each meta‑argument accepts a map or a set of strings, assigning a stable, meaningful key to each instance. Because the key does not depend on position, removing an element does not affect the others.

Guiding Principles

Use count when you need to create N nearly identical, numerically indexed resources.

Prefer for_each when each instance has a unique identifier such as a name, domain, or environment.

Conclusion

count

is a powerful Swiss‑army‑knife in Terraform, enabling bulk, conditional, and dynamic resource creation. Understanding its mechanics and the count.index expression is essential for any Terraform practitioner. However, due to its index‑based fragility, modern Terraform workflows favor for_each for managing collections, with count serving as a foundational stepping stone toward mastery.