We are taking backups of the entire service and have established restoration procedures.
Additionally, the production environment adopts a fully serverless architecture with redundant configuration, so automatic recovery occurs in the event of a failure.