What is the default state of query pushdown in the Snowflake Spark Connector?

• A. Enabled
• B. Disabled
• C. Auto
• D. Not supported

Answer: (B)

Query pushdown means having the Snowflake engine handle as much of the work as possible—such as filters, projections, and some aggregations—so Spark transfers less data and Snowflake does the heavy lifting.

In the Snowflake Spark Connector, this capability exists but is off by default. Keeping pushdown disabled avoids potential issues when Spark expressions can’t be translated into valid Snowflake SQL, ensuring correctness across a wide range of queries. If you want to leverage Snowflake to do more work, you can enable pushdown with the appropriate configuration option, allowing the connector to translate supported operations into Snowflake SQL and push them down.

Not supported would imply there’s no pushdown capability at all, which isn’t the case, and Auto would suggest a per-query automatic decision, which isn’t how the default behavior is described.