Last Updated on November 17, 2023 by Arnav Sharma

Using for_each in Terraform is a powerful way to deploy multiple resources based on a set of inputs. This method is particularly useful when you want to create multiple instances of a resource with only slight variations between them. Here’s how you can use for_each to deploy multiple resources:

Step 1: Define Your Input Set

First, you need to define a set of inputs that for_each will iterate over. This can be a list, set, or a map. Maps are often used because they allow you to easily associate each resource with a unique key.

For example, suppose you want to create multiple storage accounts in Azure. You might start with a map defining each account:

variable "storage_accounts" {
  description = "A map of storage accounts to create"
  type        = map(object({
    location = string
    account_tier = string
    account_replication_type = string
  }))

  default = {
    "account1" = {
      location = "eastus"
      account_tier = "Standard"
      account_replication_type = "LRS"
    },
    "account2" = {
      location = "westus"
      account_tier = "Standard"
      account_replication_type = "GRS"
    }
    # Add more accounts as needed
  }
}

Step 2: Use for_each in Your Resource Configuration

Next, use for_each in your resource block to iterate over each element in your input set. For each element, Terraform will create a new instance of the resource.

resource "azurerm_storage_account" "example" {
  for_each = var.storage_accounts

  name                     = each.key
  location                 = each.value.location
  account_tier             = each.value.account_tier
  account_replication_type = each.value.account_replication_type

  # Other required fields...
}

In this example, each.key is the key from the storage_accounts map (e.g., “account1”, “account2”), and each.value is the corresponding value, which is an object with properties like location, account_tier, and account_replication_type.

Step 3: Initialize and Apply Your Terraform Configuration

• Run terraform init to initialize the working directory.
• Run terraform plan to see what resources Terraform will create or modify.
• Run terraform apply to create the resources.

Additional Tips

• Dynamic Blocks: For more complex scenarios, you might use dynamic blocks within resources to create nested structures based on a list or map.
• Dependency Management: Be mindful of dependencies between resources. Terraform automatically handles dependencies, but in some cases, you may need to explicitly define them.
• State Management: Each resource created with for_each will have a unique identifier in the Terraform state file based on its key in the map. This helps in managing each resource independently.
• Modifying the Set: If you modify the input set (e.g., add or remove an item), Terraform will add, update, or destroy resources to match the new set.

Using for_each can significantly simplify your Terraform configurations, especially when dealing with multiple similar resources. It ensures that your configurations are more dynamic and easier to maintain.

FAQ – Terraform for_each

Q: What is Terraform for_each and how is it used in resource management?

A: Terraform’s for_each is a meta-argument used in Terraform to provision multiple resources of the same type. It allows you to use a set or map of strings to create multiple resources, making it easier to manage similar infrastructure components. Each item in a map or set is used to create one resource, with the each.key and each.value providing access to the map’s keys and values.

Q: How does Terraform help in managing infrastructure on cloud platforms like AWS and Azure?

A: Terraform provides a powerful infrastructure as code tool that can manage infrastructure resources across various cloud platforms, including AWS and Azure. It allows users to define infrastructure components in configuration files (like main.tf and variables.tf), enabling consistent and repeatable deployment of resources such as AWS EC2 instances, Azure resource groups, and more.

Q: Can you explain how to create multiple Azure resource groups using Terraform?

A: To create multiple Azure resource groups using Terraform, you can use the for_each meta-argument. Define a map with each desired resource group’s properties, and then use for_each in the azurerm_resource_group resource block. Terraform will iterate over the map and create a resource group for each item, using properties like location and group name from the map.

Q: What are modules in Terraform, and how do they aid in resource management?

A: Modules in Terraform are self-contained packages of Terraform configurations that are used to group together and encapsulate a set of resources and configurations. They promote reusability and maintainability, allowing you to manage resources more efficiently. Modules can be shared and reused across multiple projects, and can be sourced from the Terraform Registry or local module references.

Q: How does Terraform handle the creation of an AWS EC2 instance using a security group?

A: Terraform allows you to create an AWS EC2 instance and associate it with a security group by defining both resources in your Terraform configuration. You can specify the security group within the EC2 instance resource block, either by referencing a security group defined elsewhere in your Terraform code or by directly defining the security group rules. Terraform will ensure that the EC2 instance is created with the specified security group settings.

Q: What is the role of a workspace in Terraform, and how does it support managing multiple environments?

A: A workspace in Terraform is a logical environment for managing and isolating different states within the same Terraform configuration. It allows you to use the same configuration to manage multiple environments (like development, staging, and production) by switching between different workspaces, each maintaining its own state file. This feature supports managing multiple projects or environments more efficiently.

Q: How does Terraform’s count differ from the for_each meta-argument in creating multiple resources?

A: While both count and for_each are used to create multiple resources in Terraform, they differ in their approach. count is used to create a specified number of similar resources, using an index number to differentiate them. for_each, on the other hand, iterates over a set or map, creating resources based on each item’s unique properties. for_each is more flexible when dealing with resources that require specific attributes or when managing a dynamic set of resources.

terraform for_each hashicorp terraform used vpc terraform module azurerm terraform will use using the count set of strings gcp load balancer multiple ec2  instances terraform supports resource_groups syntax infrastructure object  refactor provider configuration append features like remote state provisioner each.value.name