Terraforming mac1.metal at AWS

Updated on the 23rd of October, 2021: Terraform AWS provider now supports Dedicated Hosts natively

In November 2021, AWS announced the support for Mac mini instances.

I believe this is huge, even despite the number of constraints this solution has. This offering opens the door to seamless macOS CI/CD integration into existing AWS infrastructure.

So here is a quick-start example of creating the dedicated host and the instance altogether using Terraform.

I intentionally used some hardcoded values for the sake of simplicity in the example.

resource "aws_ec2_host" "example_host" {
  instance_type     = "mac1.metal"
  availability_zone = "us-east-1a"
}

resource "aws_instance" "example_instance" {
  ami           = data.aws_ami.mac1metal.id
  host_id       = aws_ec2_host.example_host.id
  instance_type = "mac1.metal"
  subnet_id     = data.aws_subnet.example_subnet.id
}

data "aws_subnet" "example_subnet" {
  availability_zone = "us-east-1a"
  filter {
    name   = "tag:Tier" # you should omit this filter if you don't distinguish your subnets on private and public 
    values = ["private"]
  }
}

data "aws_ami" "mac1metal" {
  owners      = ["amazon"]
  most_recent = true
  filter {
    name   = "name"
    values = ["amzn-ec2-macos-11*"] # get latest BigSur AMI
  }
}

Simple as that, yes. Now, you can integrate it into your CI system and have the Mac instance with the underlying host in a bundle.

💡 Pro tip: you can leverage the aws_ec2_instance_type_offerings Data Source and use its output with aws_subnet source to avoid availability zone hardcoding.

To make the code more uniform and reusable, you can wrap it into a Terraform module that accepts specific parameters (such as instance_type or availability_zone) as input variables.

If you liked the article, please support me by sharing it on social networks 🙏

Serhii VasylenkoFollow

I am an engineer from Ukraine. I like astronomy and everything related to DevOps. I thrive on developing great product offerings, great people, and great teams.