Terraform: state lock error (already locked by another process)

Step-by-step Solution

1. 1Read the error message carefully — it shows the lock ID, who acquired it, and how long ago. If the lock was acquired more than a few minutes ago and no one is actively running Terraform, it's stale (a crashed run left it behind).
2. 2Ask on the team channel: 'Is anyone running Terraform on <workspace>?' Wait 60 seconds before touching anything.
3. 3If confirmed stale, force-unlock: `terraform force-unlock <LOCK_ID>`. Terraform will confirm before removing.
4. 4Common causes: a previous run was killed with Ctrl-C too aggressively, a CI job timed out mid-apply, a laptop lost network mid-apply, or DynamoDB (the lock table for the S3 backend) had a hiccup.
5. 5Prevent recurrence: enable CI job timeouts *shorter* than the lock timeout; ensure only one CI pipeline touches a given workspace; use `terraform apply -lock-timeout=5m` so runs wait a bit instead of failing instantly.
6. 6Never edit the DynamoDB lock table by hand unless you *really* know what you're doing — you can corrupt state by unlocking mid-apply.