As discussed in Chapter 9, Anti-Fragile Systems, disaster recovery implies ensuring business continuity and application availability in the event of a large failure like an entire data center going down. A failover deployment is kept in near sync with the main deployment via replication. Each component/service defines how important consistency in the failover site is. The tradeoff for consistency is the performance/efficiency drop in shipping data in real time to the failover side. The actual flip happens through DNS changes. The following diagram from Chapter 9, Anti-Fragile Systems, provides a quick recap: