10.12 Release Notes (with behavior changes): Apr 03, 2026-Apr 08, 2026¶

Dynamic table refresh boundaries¶

You can now use DYNAMIC_TABLE_REFRESH_BOUNDARY() in a dynamic table definition to prevent an upstream dynamic table from being refreshed together with the downstream dynamic table. This lets you decouple dynamic table pipelines so that each pipeline refreshes independently. Cascading refreshes and snapshot isolation do not apply across the boundary.

For more information, see Dynamic table refresh boundary.

Access history improvements¶

Access history lets you monitor the SQL statements executed in Snowflake. It keeps track of the following types of statements:

• Data Manipulation Language (DML) statements. For example, statements used to insert data into a table.

• Data Query Language (DQL) statements. For example, statements that use a SELECT statement to project data.

• Data Definition Language (DDL) statements. For example, statements that create or alter a Snowflake object.

Snowflake is expanding which SQL statements are included in the access history. This release adds support for the following:

• CREATE STREAM statements.

• SHOW AGGREGATION POLICIES, SHOW AUTHENTICATION POLICIES, SHOW NETWORK POLICIES, and SHOW PASSWORD POLICIES.

• DESCRIBE JOIN POLICY, DESCRIBE NETWORK POLICY, DESCRIBE AUTHENTICATION POLICY, and DESCRIBE PASSWORD POLICY (DESCRIBE can be abbreviated to DESC).

• DDL related to MCP servers.

• DDL related to Postgres instances.

• DDL related to Cortex agents.

For a complete list of objects and commands that appear in your access history, see Supported Objects.