Describe database and account replication and failover and failback for business continuity and disaster recovery.

An organisation runs its production account in one region and wants a disaster recovery account in a second region that can take over the business if the primary region fails. The recovery must bring across not just table data but also the account-level objects such as users, roles, warehouses and resource monitors, and it must be possible to promote the secondary so it serves writes during an outage. Which Snowflake capability best meets this requirement?

• AA failover group that replicates the chosen databases together with account objects such as users, roles, warehouses and resource monitors to the second account, and can be failed over to promote the secondary as primarycheck_circle Correct
• BA secure data share of every production database to the second account, so the recovery account reads live production data and is promoted to accept writes when the primary region is unavailable
• CDatabase replication configured separately on each production database, which copies the table data to the second account and lets each replicated database be promoted independently to accept writes during the outage
• DA zero-copy clone of each production database created in the second account, which provides an instant independent copy that the recovery account can write to as soon as the primary region goes down

Why A is correct: A failover group replicates selected databases plus account-level objects as one consistent unit and supports failover, so the secondary account can be promoted to primary and serve writes during a regional outage, exactly as required.

Why B is wrong: Secure data sharing gives read-only access to live data and shares only databases, not account objects like users and roles, and a share cannot be promoted to serve writes, so it does not provide failover.

Why C is wrong: Database replication copies database objects but not account-level objects such as users, roles and warehouses, and promoting many databases one by one is not a coordinated account failover, so it falls short of the requirement.

Why D is wrong: Zero-copy cloning operates within a single account and region and creates a point-in-time copy, not an ongoing cross-region replica, so it cannot keep a remote recovery account current or fail over to it.