Most Terraform discussions focus on a single problem: large state files.
The usual recommendation is to split infrastructure into smaller deployments. Networking receives its own state. Applications receive their own state. Security receives its own state.
The graph becomes smaller. Lock contention decreases. Planning becomes easier.
A new problem appears.
Applications depend on networking outputs. Shared services must deploy before dependent workloads. Teams begin tracking deployment order through pipeline logic, documentation, and tribal knowledge.
The original state problem becomes a coordination problem.
This lesson follows that progression step by step and explains why Terraform Stacks were introduced. The goal is not to learn stack syntax. The goal is to understand the architectural pressure that made stacks necessary.