Scenario: An incident happens, all of our pipelines are destroyed. And I now need a way to recover everything with all notification channels and Triggers in place.

The workaround we are stuck with: Everything, including Notification Channels and Triggers are Terraformed, then attached to the appropriate pipelines where fit. This means we have to do a lot of additional work to identify which Triggers and which Notification Channels should be applied to which pipelines.

The solution we want: Notification Channels and Triggers are stored in pipeline.yaml files as a separate block, rather than an independent resource. This means we don't have to create new Terraform config entries for each Notification channel and Trigger, but simply add to each pipeline.yaml file which reduces the amount of total entities in our Terraform code.