Dynamic tables continue incrementally refreshing after failover

Dynamic tables configured for incremental or adaptive refresh can now maintain their incremental behavior after a failover group promotion, avoiding unnecessary full reloads.### Use casesImagine you have a critical sales data pipeline replicated to a secondary account for disaster recovery and a failover occurs. With this update, you can:- Avoid the computational cost of a full refresh after promoting the secondary account.- Maintain incremental update continuity on eligible tables.- Ensure that schedules and warehouse assignments are preserved.### How to useTo use replication with dynamic tables, group the table and its base objects in a failover group:sqlCREATE FAILOVER GROUP myfg OBJECT_TYPES = DATABASES ALLOWED_DATABASES = analytics_db ALLOWED_ACCOUNTS = myorg.secondary_account REPLICATION_SCHEDULE = '10 MINUTE';What happens in this code:1. Creates a failover group named myfg.2. Sets the object type to databases.3. Specifies the allowed database and target secondary account.To monitor the scheduling state on the secondary:sqlSELECT name, scheduling_state FROM TABLE(INFORMATION_SCHEMA.DYNAMIC_TABLE_GRAPH_HISTORY());What happens in this code:1. Queries the DYNAMIC_TABLE_GRAPH_HISTORY system function.2. Returns the dynamic table name and current scheduling state to identify replication suspension reasons.To check refresh history for reinitialization events:sqlSELECT name, refresh_action, refresh_trigger, reinit_reason FROM TABLE(INFORMATION_SCHEMA.DYNAMIC_TABLE_REFRESH_HISTORY()) WHERE refresh_action = 'REINITIALIZE' ORDER BY data_timestamp DESC;What happens in this code:1. Queries the dynamic table refresh history.2. Filters by REINITIALIZE actions to check if any reinitialization occurred post-failover.### Limitations and requirements- Available in General Availability (GA).- Requires failover groups (simple replication groups do not support failover).- Dynamic Iceberg tables and dynamic tables with Apache Iceberg base tables do not yet support incremental refresh after failover.- All base objects must belong to the same failover group and remain unchanged since the last successful refresh in the primary account.