# Restore validation checks

Use with [Validate the restored application](/blog/aws-backup-automated-restore-validation/). Checked 19 October 2026. Plans, vaults, and cadence stay on the [backup strategy guide](/blog/aws-backup-strategies-automated-data-protection/).

| Check | Pass | Report via `PutRestoreValidationResult` |
| --- | --- | --- |
| Restore job status is `COMPLETED` | Event matches the testing plan ARN | Do not validate a manual restore with this rule |
| Restored endpoint accepts a connection | TCP or database port open | `FAILED` with the error string |
| Expected row is present | Query returns the fixture key | `FAILED` if the count is zero |
| Optional HTTP health | Status 200 from the test host only | `FAILED` on timeout |
| Finished inside the validation window | Call returns before AWS Backup closes the window | `TIMED_OUT` if you did not finish. The status cannot be changed once set |

`ValidationStatus` values documented for the API: `FAILED`, `SUCCESSFUL`, `TIMED_OUT`, `VALIDATING`.

## Stop

- Deleting the restored resource yourself before AWS Backup's cleanup, unless you have confirmed the ARN is the test copy.
- Treating `SUCCESSFUL` as proof of a failover runbook. This sheet is one resource, one query, one health URL.