Disaster recovery prepares a service to recover after a major outage or data loss. Recovery Point Objective (RPO) is the maximum tolerable amount of data loss measured in time; Recovery Time Objective (RTO) is the target time to restore service. Backups, replication, infrastructure-as-code, and runbooks help, but only restore and failover exercises validate the plan. Replication alone can copy accidental deletion or corruption.