Snowpipe Streaming - Classic Architecture¶

The Snowpipe Streaming classic architecture offers a proven and efficient method for continuous, low-latency, row-based data ingestion directly into Snowflake tables. This implementation, referred to as Snowpipe Streaming Classic in the documentation, remains a reliable choice for diverse streaming workloads such as application event data, Internet of things (IoT) sensor readings, and low-latency Change Data Capture (CDC).

Snowpipe Streaming Classic uses the snowflake-ingest-java SDK and operates without the explicit PIPE object concept for managing data flow that is central to the Snowpipe Streaming high-performance architecture. Instead, in Snowpipe Streaming Classic, channels are configured more directly against tables, offering a familiar and established approach to streaming data into Snowflake.

To information about how to use Snowpipe Streaming Classic, see the following topics:

• Configuration for Snowpipe Streaming Classic

• Snowpipe Streaming Classic best practices

• Costs for Snowpipe Streaming Classic

• Using Snowflake Connector for Kafka with Snowpipe Streaming Classic

• Using Snowpipe Streaming Classic with Apache Iceberg™ tables

• Replication and Snowpipe Streaming