The Snowflake CDC Snap Pack lets SnapLogic pipelines identify and capture changes made to data in a Snowflake database rather than reprocessing entire datasets. It works by reading from Snowflake Streams — Snowflake’s native change-tracking mechanism — to surface row-level Insert, Update, and Delete operations on a specified table. Changes can be retrieved within a defined date/time range, giving pipeline builders precise control over which delta of data to process. This enables efficient data replication, synchronization, and downstream integration without the overhead of full-table scans.
What you can do
With this Snap Pack, you can build pipelines that continuously or periodically pull only the rows that changed in a Snowflake table since the last check, using a configurable polling interval and batch output. This is well suited to keeping downstream systems, data lakes, or analytics platforms in sync with Snowflake without reloading entire tables.
Why it matters
Full-dataset reprocessing is slow and resource-intensive, especially at enterprise data volumes. By capturing only what changed, teams reduce compute cost in Snowflake, cut pipeline run times, and keep dependent systems closer to real time.
This Snap Pack includes the following key features
- Retrieves all data changes within a specific time range for precise filtering
- Captures Insert, Update, and Delete statements on specific tables
- Reads directly from Snowflake Streams (Snowflake’s native change-tracking object)
- Configurable polling interval and polling unit (seconds/minutes/hours)
- Supports batch output for bulk-processing changed records
- Compatible with Ultra Tasks for near-real-time pipelines
This Snap Pack contains the following Snaps
- Snowflake CDC — Reads from a Snowflake Stream and retrieves Insert, Update, and Delete changes on a specified table within a given time range.
- Snowflake CDC Account — Provides the Snowflake connection (account identifier, warehouse, database, schema, role, credentials) used by the Snowflake CDC Snap.
To learn more
To learn more, please check out the documentation page.
