A recovery target does not prescribe one universal architecture
A fictional order service has a recovery-point objective of fifteen minutes and a recovery-time objective of four hours. A tested option creates consistent recovery points every ten minutes and restores the service in three hours after the declared recovery procedure begins. A second option replicates transactions continuously but its tested restoration procedure takes five hours. A third option uses hourly snapshots and restores in two hours. The exercise assumes the stated intervals and restoration durations are achieved for the failure scenario under discussion, with suitable authentication, recovery copies and dependency availability. No geographic-separation requirement is supplied. A manager proposes declaring the continuously replicated option best solely because it loses the least recent data. The team also knows that replication can propagate an accidental deletion if there is no protected recovery history.
1. Compare the three options with both stated targets.
Read the worked answer
The ten-minute/three-hour option meets both supplied targets. Continuous replication with a five-hour restoration misses RTO, while hourly snapshots with a two-hour restore miss RPO. These conclusions rely on the stated successful tests and failure scenario.
2. Explain why replication alone does not settle the protection decision.
Read the worked answer
A live replica may receive the same unwanted deletion or corruption. Protected recovery history, access controls, retention and restore testing must address that risk. The targets do not by themselves mandate continuous cross-region replication or a particular standby label.