An administrator, who operates a public cloud, has been tasked with implementing a secondary datacenter for failover purposes. Immediate replication has too much of an impact on the WAN link during…

Asynchronous replication transmits data updates to the secondary site with a delay, allowing the primary site to acknowledge writes locally before replication occurs. This approach minimizes WAN link impact during production hours by batching updates and often sending them during off-peak times or with less urgency, making it ideal when immediate replication is too taxing.

Storage deduplication reduces the amount of storage space required by eliminating redundant data, but it does not directly address the timing or impact of replication on WAN bandwidth for failover purposes.

A cold site failover with offsite archiving implies a slower recovery time objective (RTO) and recovery point objective (RPO) and typically involves restoring from backups, which is not suitable for a 'failover' implying a more active state with data synchronization.

Site mirroring, often synonymous with synchronous replication, requires immediate and consistent data writes across both sites, which would have a significant and immediate impact on the WAN link, precisely what the administrator wants to avoid.