Skip to content

Provider Credentials

Atlantis runs Terraform by simply executing terraform plan and apply commands on the server Atlantis is hosted on. Just like when you run Terraform locally, Atlantis needs credentials for your specific provider.

It's up to you how you provide credentials for your specific provider to Atlantis:

  • The Atlantis Helm Chart and AWS Fargate Module have their own mechanisms for provider credentials. Read their docs.
  • If you're running Atlantis in a cloud then many clouds have ways to give cloud API access to applications running on them, ex:
  • Many users set environment variables, ex. AWS_ACCESS_KEY, where Atlantis is running.
  • Others create the necessary config files, ex. ~/.aws/credentials, where Atlantis is running.
  • Use the HashiCorp Vault Provider to obtain provider credentials.

TIP

As a general rule, if you can ssh or exec into the server where Atlantis is running and run terraform commands like you would locally, then Atlantis will work.

AWS Specific Info

Multiple AWS Accounts

Atlantis supports multiple AWS accounts through the use of Terraform's AWS Authentication (Search for "Authentication").

If you're using the Shared Credentials file (Search for "Shared Credentials file") you'll need to ensure the server that Atlantis is executing on has the corresponding credentials file.

If you're using Assume role (Search for "Assume role") you'll need to ensure that the credentials file has a default profile that is able to assume all required roles.

Using multiple Environment variables (Search for "Environment variables") won't work for multiple accounts since Atlantis wouldn't know which environment variables to execute Terraform with.

Assume Role Session Names

If you're using Terraform < 0.12, Atlantis injects 5 Terraform variables that can be used to dynamically name the assume role session name. Setting the session_name allows you to trace API calls made through Atlantis back to a specific user and repo via CloudWatch:

bash
provider "aws" {
  assume_role {
    role_arn     = "arn:aws:iam::ACCOUNT_ID:role/ROLE_NAME"
    session_name = "${var.atlantis_user}-${var.atlantis_repo_owner}-${var.atlantis_repo_name}-${var.atlantis_pull_num}"
  }
}

Atlantis runs terraform with the following variables:

-var ArgumentDescription
atlantis_user=lkysowThe VCS username of who is running the plan command.
atlantis_repo=runatlantis/atlantisThe full name of the repo the pull request is in. NOTE: This variable can't be used in the AWS session name because it contains a /.
atlantis_repo_owner=runatlantisThe name of the owner of the repo the pull request is in.
atlantis_repo_name=atlantisThe name of the repo the pull request is in.
atlantis_pull_num=200The pull request number.

If you want to use assume_role with Atlantis and you're also using the S3 Backend, make sure to add the role_arn option:

bash
terraform {
  backend "s3" {
    bucket   = "mybucket"
    key      = "path/to/my/key"
    region   = "us-east-1"
    role_arn = "arn:aws:iam::ACCOUNT_ID:role/ROLE_NAME"
    # can't use var.atlantis_user as the session name because
    # interpolations are not allowed in backend configuration
    # session_name = "${var.atlantis_user}" WON'T WORK
  }
}

Why does this not work in TF >= 0.12?

In Terraform >= 0.12, you're not allowed to set any -var flags if those variables aren't being used. Since we can't know if you're using these atlantis_* variables, we can't set the -var flag.

You can still set these variables yourself using the extra_args configuration.

Next Steps