When to use trigger policies vs dependencies

With the introduction of the dependency system, you have two options for managing relationships between your infrastructure components: trigger policies and stack dependencies. Here's when to use each approach.

Stack Dependencies (Recommended)

For most use cases, we recommend using stack dependencies as they provide a more modern and streamlined approach to managing relationships between your stacks. Stack dependencies can completely replace most existing trigger policies.

When to Still Use Trigger Policies

There are specific scenarios where trigger policies are still the preferred option:

• Automatic module updates - When you need to automatically update Terraform modules from the registry

• Retriggering failed runs - For automatically retrying failed executions

• Triggering with fields from data input - For automatically triggering runs with fields available in this data input, i.e if using a creator stack, to trigger a new run on the created stack when something has been updated.

Terraform Registry Modules

If you're using Terraform modules from the registry and want automatic module updates, you'll still need to use trigger policies for this functionality.

In summary, while stack dependencies are the recommended approach for most dependency management, trigger policies still serve specific automation purposes that aren't covered by the dependency system.