SAA-C03 Practice Question 54: Warm Standby DR Strategy

Two constraints do the work here. "Serving traffic within five minutes, at reduced performance if necessary" means the recovery Region must already be able to answer requests at the moment of failover; "no second Region at full production capacity" rules out the most expensive option. Warm standby is the strategy defined by exactly that trade-off: a functional but under-sized copy of the stack is always on, so it takes traffic immediately and is then scaled out by raising the Auto Scaling group's desired capacity. An Aurora global database replicates with typical latency under a second and can be promoted in about a minute, which comfortably meets a one-minute recovery point.

Option A is the slowest strategy by design. Restoring Aurora snapshots and deploying the whole stack from templates takes far longer than five minutes, and a periodic snapshot schedule gives a recovery point measured in hours, not one minute.

Option B is the closest wrong answer, and the distinction is the point of the question. Pilot light keeps data replicating but cannot process requests without additional action first; servers have to be deployed or switched on and then scaled. That extra step is what pushes it past a five-minute recovery time, whereas warm standby only has to scale.

Option C would meet the targets, and then some, but it is the most costly and most complex approach, and running both Regions at full production capacity is precisely what finance rejected. Choosing it means solving a requirement nobody asked for at a price that was ruled out.

Reference: Disaster recovery options in the cloud