Snowflake
Replication & Failover Across Regions
Maintain business continuity by replicating databases and account objects across regions or clouds.
Snowflake Replication allows databases (and, more broadly, entire accounts via Business Continuity & Disaster Recovery features) to be continuously or periodically copied to a secondary Snowflake account in a different region or even a different cloud provider, protecting against regional outages and enabling geographically distributed access to the same data.
It's like maintaining a mirror office in another city that's kept in sync with regular data deliveries — if the main office suddenly goes dark, the mirror office can be promoted to take over operations, ready-made with the same files, staff badges, and access rules already in place.
Key Concepts
1
At the database level, a primary database is linked to one or more secondary databases in target accounts; replication is triggered on a schedule (often orchestrated via a task calling ALTER DATABASE ... REFRESH) and copies only the changed data since the last refresh, similar in spirit to incremental backups. Secondary databases are read-only until a failover is initiated.
ALTER DATABASE ... REFRESHfailover
2
Failover groups extend this beyond just databases to include account-level objects — users, roles, warehouses, resource monitors, network policies — allowing a more complete account "promotion" of a secondary region to primary in a disaster scenario, rather than just having replicated data with none of the surrounding access/security configuration replicated alongside it.
Failover groups
3
Replication is priced based on the compute needed to detect and transfer changed data plus the storage of the replicated copy, and organizations typically weigh replication frequency (more frequent = lower RPO/data loss window, but higher ongoing cost) against their actual recovery point/time objectives (RPO/RTO) rather than replicating continuously by default.