Businesses typically employ disaster recovery solutions that replicate data to a secondary location and enable failover during outages. BigQuery offers Managed Disaster Recovery for this purpose, but data loss during testing remains a concern. Traditional hard failover forces a choice between immediate failover with potential data loss or delayed recovery. BigQuery's new soft failover addresses this by promoting the secondary region's compute and datasets only after replication is confirmed. This provides control and minimizes data loss risk during planned failovers. Soft failover is suitable for failover testing when both primary and secondary regions are available. In contrast, hard failover is for unplanned outages. Soft failover ensures replication completion before promoting the secondary, minimizing potential data loss. Hard failover initiates immediately, risking data loss if replication is incomplete. BigQuery's soft failover feature is accessible through the UI, DDL, and CLI. This allows for confident disaster recovery simulations and compliance without data loss during testing. It helps maintain uptime and prevent data loss during safe scenario testing.