Skip to main content

The Terragrunt Grain

The Terragrunt grain is Torque's native support for Terragrunt modules. Terragrunt is a thin wrapper that provides extra tools for working with multiple Terraform modules and keeping your configurations DRY. Torque allows designers to use Terragrunt-specific features to easily orchestrate self-developed and community Terragrunt modules in a standard way and share them as building blocks.

Terragrunt Grain Specification​

grains:
  terragrunt-grain:
    kind: terragrunt
    tf-version: 1.5.5 # Optional, specify the Terraform version to use
    spec:
      source:
        path: terragrunt/s3-bucket/non-prod/us-east-1/qa/bucket
        store: my_repository
      agent:        
        name: '{{ .inputs.agent }}' # Reference an input to specify the agent
      inputs:
        - name: '{{ .inputs.bucket_name }}' # Reference an input as the input value
      outputs:
        - bucket_arn # Specify the outputs to capture from the Terragrunt module

In the above example, we define a Terragrunt grain named terragrunt-grain. The kind field is set to terragrunt to indicate that this is a Terragrunt grain.

Terraform Version​

The tf-version field is optional and allows you to specify the Terraform version to use for this Terragrunt module. If not specified, the latest compatible version will be used.

Source​

The source block specifies the location of the Terragrunt module. The path field should point to the root directory of the Terragrunt module, and the store field should reference the repository or storage location where the module is stored.

Agent​

The agent block specifies the agent to use for executing the Terragrunt module. In the example, we reference an input named agent to dynamically specify the agent.

Inputs​

The inputs block is a list of input values to pass to the Terragrunt module. In the example, we reference an input named bucket_name as the value for the input named name.

Outputs​

The outputs block is a list of output values to capture from the Terragrunt module. In the example, we specify bucket_arn as an output to capture.

Blueprint Usage Example​

Here's an example of how the Terragrunt grain can be used in a Torque blueprint:

spec_version: 2

inputs:
  agent:
    type: agent
  bucket_name:
    type: string

outputs:
  output1:
    value: 'http://my-env-{{ envId }}/some/api'
    kind: link
    quick: true
  output2:
    value: '{{ .grains.terragrunt-grain.outputs.bucket_arn }}'     

grains:
  terragrunt-grain:
    kind: terragrunt
    tf-version: 1.5.5
    spec:
      source:
        path: terragrunt/s3-bucket/non-prod/us-east-1/qa/bucket
        store: test-spec2-public
      agent:        
        name: '{{.inputs.agent}}'
      inputs:
        - name: '{{.inputs.bucket_name}}'     
      outputs:
        - bucket_arn

In this example, we define two inputs: agent and bucket_name. We also define four outputs, including a link to an external URL, references to the input values, and a reference to the bucket_arn output captured from the terragrunt-grain grain.

The terragrunt-grain grain is defined with the kind set to terragrunt, and we specify the Terraform version to use (1.5.5). The source block points to the location of the Terragrunt module, and the agent block references the agent input value. The inputs block includes the bucket_name input value, and the outputs block specifies bucket_arn as an output to capture.

When this blueprint is deployed, Torque will execute the Terragrunt module specified in the terragrunt-grain grain, passing in the bucket_name input value and capturing the bucket_arn output value. The captured output value can then be referenced in the blueprint's outputs or used in other parts of the blueprint.