A scrub detects latent corruption, a resilver reconstructs missing redundancy, a snapshot preserves an earlier dataset state, and replication puts that state somewhere else. Calling all four 'backup' hides the failures each one cannot solve.
A failed job can leave a fresh-looking directory that should never replace the last known complete recovery point.
Live files that are absent from Git are technical debt even when the service is currently healthy.
Separating secrets from source control creates a recovery dependency that must be documented and tested.
Production fixtures need a recovery path below the application firmware.
State maps resource addresses in configuration to real remote objects. Treating it as disposable cache data is how infrastructure gets orphaned or recreated.
Resource targeting narrows Terraform to a selected subset plus dependencies. It is valuable for exceptional recovery but can leave the rest of the configuration unapplied.
Losing the authenticator device creates a second-factor recovery problem that should not silently collapse to the password path.
state mv, state rm, state replace-provider, state pull and state push can change Terraform's ownership model without directly changing remote infrastructure.
A scheduler can enqueue work before crashing; recovery logic must discover incomplete deliveries after the process returns.
A device that loses power or Wi-Fi halfway through setup should not become permanently ambiguous.