A customer with a single data center wants to provide resiliency in the form of an additional hosting location. The customer's engineering team requested asynchronous replication for the primary syste

A secondary data center only 2 km away is often too close to provide adequate protection against regional disasters (e.g., widespread power outages or natural events), thus not offering maximum business continuity.

A distance of 1000 km, while providing excellent geographic separation, would introduce significant latency, making even asynchronous replication less efficient in minimizing data loss and complicating Layer 2 spanning for high-performance applications.

Asynchronous replication tolerates some latency and is suitable for disaster recovery over longer distances, aiming to minimize data loss (RPO) while providing resilience against regional disasters. A 50 km distance with 5 ms RTT is far enough to protect against common localized failures, yet close enough to keep replication latency minimal, thereby reducing data loss. This setup aligns with the preference for minimal latency and a 4-hour Mean Time To Recovery (MTTR) tolerance, offering robust business continuity.

A 2 km distance, even with Layer 2 spanning and low RTT, is insufficient for a robust disaster recovery site against regional disasters, as it lacks the necessary geographic separation for maximum business continuity.