Appearance

Orchestration ​

Orchestration allows the execution of any commands such as terraform apply or tofu plan in stacks using a graph-based approach. This comes especially handy whenever running Terramate in automation because the orchestration in Terramate comes with change detection that allows you to orchestrate commands in stacks that contain changes in a commit, branch or Pull Request.

INFO

Compared to other available tooling, Terramate CLI is not a wrapper around Terraform or OpenTofu. Instead, you can use Terramate to orchestrate any command in stacks.

Use Cases ​

  • Orchestrate commands: on a single or a graph of stacks using filters.
  • Define and use workflows: use Terramate Scripts to define custom commands to combine multiple commands into one executable unit of work.
  • Detect changes: to only orchestrate commands or workflows in stacks that contain changes in a commit, branch or Pull Requests.
  • Execute stacks in order: define the order of execution of related stacks to ensure dependency constraints.

Commands and Workflows ​

Currently, two ways of orchestrating commands exist:

  • Executing a single command such as terraform apply using the terramate run command, e.g., terramate run -- terraform apply.
  • Executing a sequence of commands combined as a workflow with Terramate scripts using the terramate script run command, e.g., terramate script run terraform deploy.

Change detection ​

Terramate comes with a change detection feature, which allows you to orchestrate stacks that contain changes only. The change detection comes with several integrations to cover IaC-specific use-cases, such as Terraform and OpenTofu module change detection or Terragrunt dependency change detection.

For details, please see the change detection documentation.

Sequential and parallel execution ​

By default, commands and workflows run sequentially in stacks but can also run in parallel.

Order of execution ​

Terramate can invoke commands and workflows on a single stack or a directed acyclic graph (DAG) of stacks.

  • Implicit order of execution: Allows to define the order of execution by nesting stacks using the file system hierarchy in a repository.
  • Explicit order of execution Allows the enforcement of a specific order of execution using stack configuration such as before and after by referencing related stacks by path or tags.
  • Forced execution: Allows enforcing the execution of specific stacks using stack configuration such as wants and wanted_by by referencing related stacks by path or tags.

Graph-based orchestration ​

Terramate can invoke commands and workflows on a single stack or a directed acyclic graph (DAG) of stacks. The graph depicts the dependency relations between the stacks. The run order follows a topological ordering of the nodes, which can be filtered using change detection or tag filters. Furthermore, the execution order can also be reversed, which is beneficial when running commands like terraform destroy on a graph of dependent stacks.

For details, please see the graph-based orchestration documentation.