An Amazon Aurora MySQL cluster has automatic backups, point-in-time recovery, and backtracking enabled. You must revert the existing production cluster to a specific recovery point within the last 72 hours, performing the rollback in-place on the same DB cluster. What should you use?

Aurora backtracking allows you to rewind an Aurora DB cluster to a specific time within the configured backtracking window, which can be up to 72 hours. This operation is performed in-place on the existing cluster, making it ideal for reverting changes quickly without creating a new cluster. Creating an Aurora Replica and promoting it would create a new cluster, not rewind the existing one. Invoking a Lambda function to restore an automatic backup would also create a new cluster from the backup, not perform an in-place rewind. Point-in-time restore (PITR) creates a new cluster restored to a specific time, rather than rewinding the existing cluster.