Question

A growing platform team has three engineers who each run Terraform from their own laptops against the same production AWS environment. They keep experiencing conflicting changes, resources being unexpectedly destroyed, and a plan that doesn't match reality. Which characteristic of Terraform state is the root cause they must address to fix this?

Accepted Answer

State is stored locally by default, so each engineer has a separate, unshared copy of the resource mapping

Answer

State only records resource IDs and cannot track attribute values, so drift is inevitable across machines

Answer

State is regenerated from scratch on every apply, so past runs are never remembered

Answer

State is provider-specific and cannot represent multiple resource types in one file