Terragrunt

Platform teams deploying identical cloud foundations across development, staging, and production environments frequently watch repetitive provider blocks and backend configurations drift across dozens of directories. Terragrunt removes that boilerplate by wrapping OpenTofu and Terraform with hierarchical HCL configuration files and explicit dependency graphs. Infrastructure engineers define remote state backends and shared environment variables once, executing coordinated multi-module rollouts without duplicating root manifests.

Why Terragrunt instead of Terraform Cloud

Terraform Cloud centers its workflow on hosted workspaces, storing state remotely and running plan or apply jobs on managed runners through a web console. Terragrunt approaches infrastructure workflow from the code repository instead. Rather than configuring workspaces in a graphical UI, teams declare their multi-environment architecture in Git using nested directories. An include block inherits remote backend settings, provider generation, and common tags down to individual components. When components rely on one another, explicit dependency blocks pass outputs between units, building a DAG that replaces Terraform Cloud's cross-workspace lookups with CLI orchestration.

The operational tradeoffs reflect this architectural split. Terraform Cloud provides turnkey runners, role-based access control, visual plan diffs, and audit logging, but it binds teams to proprietary workspaces, per-resource billing tiers, and a remote execution model that complicates local debugging. Terragrunt requires no external control plane and runs identically on a developer workstation, in GitHub Actions, or within Atlantis. Teams forgo a hosted dashboard to maintain their own CI pipelines. In exchange, they gain total control over execution order, local execution parity, and the freedom to run either OpenTofu or Terraform without paying recurring SaaS fees as managed resource counts expand.

What it does well

• Remote state auto-provisioning: Provisions remote storage buckets and lock tables on AWS, Google Cloud, or Azure dynamically, removing manual backend setup.
• Explicit module dependencies: Resolves inter-module outputs via dependency blocks, ordering execution in a topological DAG across stacks.
• Dynamic code generation: Injects provider blocks and backend configurations into child modules using generate blocks, keeping module repositories clean.