HashiCorp Terraform Associate (003) Exam Survival Guide

Window: Global | IaC fundamentals | ~60 Q / 60 min

Must-do topics

• Core: providers, resources, data sources, variables/locals, outputs, state
• CLI workflow: init → validate → plan → apply → destroy; workspaces
• State mgmt: backends (local/remote), locking, state drift, taint/rm/import
• Modules: structure (versions.tf, variables.tf, outputs.tf), registry usage, version pinning
• Provisioners & lifecycle: create_before_destroy, ignore_changes, depends_on
• Cloud & security: Terraform Cloud, remote runs, VCS integration, sensitive values

Top traps (avoid)

• Hardcoding values instead of variables/locals; no reuse
• Mixing data sources for read-only lookups vs resources for creation
• Forgetting to pin provider/module versions → breaking changes
• Editing .tfstate manually; fix via import/refresh instead
• Using provisioners for what a proper resource or user data should do

Time split

• ~1 min/Q; first pass on terminology/CLI; second pass on state/modules

Last-48h checklist

• Practice: init/plan/apply/destroy on a tiny module
• Backend + lock demo (S3+DynamoDB or Terraform Cloud)
• for_each vs count, maps vs objects; variable validation
• Import one real resource (terraform import) and output its attrs

Quick facts

• Desired state model; plan shows diff; apply makes it real
• for_each needs stable keys; count uses index
• Remote backend with locking prevents concurrent apply
• Sensitive outputs are masked in UI/logs but exist in state

Speed tactics

• Translate scenario → choose resource/data/module correctly
• Prefer modules + variables over duplicating blocks
• If two answers work, pick the one that reduces drift and enables locking

Day-of mini-plan

• 10-min warm-up: workflow commands + one for_each snippet
• Read stems for what’s being changed (state? module? backend?)
• Flag long module/version questions; return after quick wins