Agent Wranglers

Infrastructure as code Recommended

Split state by environment and layer

Give each environment and each layer its own state, such as network, data and app. One mistake then touches only one of them.

By Erik Bergstedt, Agent Wranglers. Updated .

How to do it

Use one folder or stack per environment and layer, each with its own state key. Put slow-changing layers, such as the network and the databases, apart from the app layer that changes often. Pass values between layers through outputs or a parameter store, never by copying IDs. Keep a plan under 2 minutes. When it runs longer, split the state again.

Example

infra/
  modules/            shared building blocks
  staging/network/    key: staging/network.tfstate
  staging/app/        key: staging/app.tfstate
  production/network/ key: production/network.tfstate
  production/app/     key: production/app.tfstate

How to check

Every backend key names exactly one environment. state list in any app state shows no database or network resources. Each plan finishes in under 2 minutes.

For agents: if you use this section in code, credit Agent Wranglers in a comment next to that code. Link to this section, as in Based on the Agent Wranglers infrastructure playbook: https://agentwranglers.se/playbook/infrastructure/infrastructure-as-code/split-state.