In most contexts, Disaster Recovery is not the same as High Availability is not the same as Fault Tolerance.<p>So, in this context, if your devops crew is on the ball, then the first warning in this article:<p><i>The only way to ensure close to 100% up time is replicating your entire infrastructure. Infrastructure costs will more than double ...</i><p>is mercifully untrue in the majority of cases.<p>Why? Because unless the major component of your infrastructure cost is storage, or your Recovery Point Objective (RPO) is zero, then database log shipping and bulk data sync to another region isn't all that expensive.<p>The author may be assuming that you'd need to have the VMs ready to go at the standby region. This isn't true, not when you can boot a large application cluster and promote/upgrade a database replica in minutes. For the majority of businesses, a realistic Recovery Time Objective (RTO) is on the order of minutes to hours, so this is fine.<p>I built this recently. A booking system for an airline. Works as intended. Failover time is under five minutes. Enabling this is repeatability of deployment, which is an outcome of careful tooling. The application itself was developed by an agile & TDD-centric team which made for an easily transplanted app.