10.14 Release Notes: Apr 20, 2026-Apr 23, 2026¶

Session policies: maximum session lifespan¶

Session policies now support a maximum session lifespan that ends sessions after a set duration, regardless of user activity. Previously, session policies could only enforce idle timeouts. The new properties let administrators require re-authentication after a fixed period, even for continuously active sessions.

Two new session policy properties are available:

• SESSION_MAX_LIFESPAN_MINS: for programmatic and Snowflake clients.

• SESSION_UI_MAX_LIFESPAN_MINS: for Snowsight.

The configurable range for both is 0 to 43200 minutes (30 days). A value of 0 (the default) means no maximum lifespan is enforced.

For more information, see Snowflake sessions and session policies.

Schema-level data metric functions (General availability)¶

You can now associate data metric functions (DMFs) at the schema level, making it easier to monitor data quality across all objects in a schema. This feature is now generally available and includes the following capabilities:

• Associate the ROW_COUNT or FRESHNESS system DMFs with all table-like objects in a schema using a single SQL statement.

• Enable anomaly detection for all table-like objects in the schema to automatically identify unusual patterns in data volume or freshness.

• Exclude specific table-like object types from the DMF association using the EXCLUDE_TABLE_TYPES parameter.

• Override schema-level settings for individual tables or views when needed.

When you associate a DMF at the schema level, Snowflake automatically creates object-level associations for all supported table-like objects within the schema (unless excluded), simplifying data quality monitoring for large numbers of tables and views.

New columns in DATA_METRIC_FUNCTION_REFERENCES¶

The INFORMATION_SCHEMA.DATA_METRIC_FUNCTION_REFERENCES function and the ACCOUNT_USAGE.DATA_METRIC_FUNCTION_REFERENCES view now include the following new columns to support schema-level DMF associations:

• level — Indicates whether the DMF was associated at the TABLE level (directly) or SCHEMA level (via schema-level association).

• exclude_table_types — Shows which object types are excluded when a DMF is added at the schema level.

These columns help you identify which DMF associations were created through schema-level configuration versus direct table-level associations.

For more information, see the following topics:

• Monitor the data quality of a schema

• Detecting anomalies in data quality

• DATA_METRIC_FUNCTION_REFERENCES

• DATA_METRIC_FUNCTION_REFERENCES view

Dynamic tables: expanded outer join support¶

Dynamic tables in incremental refresh mode now support the following outer join patterns:

• Outer joins where both sides are the same table.

• Outer joins where both sides are a subquery with GROUP BY clauses.

For more information, see Supported queries for dynamic tables.

ACCUMULATE aggregate function¶

The ACCUMULATE aggregate function is now available. It returns a custom aggregate value computed with four user-defined SQL lambda functions (initialize, accumulate, combine, and terminate) using a map-reduce model. It works with GROUP BY, HAVING, and subqueries the same way built-in aggregate functions do.

For more information, see ACCUMULATE.