CI Perf Lint

terraform-parallelism-unconfigured

Why it matters

Terraform defaults to parallelism=10, which is slow for large configurations. Tuning parallelism to match runner capacity is one of the highest-leverage Terraform CI optimizations.

Without explicit parallelism configuration, nobody on the team is thinking about it, and CI runs may be slower than necessary.

What it flags

Workflows that run terraform plan, apply, or destroy without configuring --parallelism or TF_CLI_ARGS.

Suggested action

Add --parallelism=N to terraform plan/apply/destroy commands:

- run: terraform plan -parallelism=30

Or set it at the workflow/job level:

env:
  TF_CLI_ARGS: -parallelism=30

Start with 30-50 on standard GitHub runners and adjust based on resource contention and API rate limits.

Verification

Compare plan/apply duration before and after changing parallelism. Also monitor API rate limiting (e.g., AWS, Azure) at higher values.