Supported Target Types

Applies To:	Pipeline Bundle
Configuration Scope:	Data Flow Spec
Databricks Docs:	https://docs.databricks.com/en/delta-live-tables/python-ref.html

The Lakeflow Framework supports multiple target types.

Target Types

Type	Description	Key Features
Delta Streaming Table	Creates a streaming Delta table that continuously processes and updates data as it arrives.	Streaming write optimizations Automatic schema evolution Quality enforcement Incremental processing
Materialized Views	A materialized view is a view that contains precomputed records based on the query that defines the materialized view. Materialized views are commonly used for transformations and Gold Layer tables.	Automatic updates based on pipeline schedule/triggers Guaranteed consistency with source data Incremental refresh optimization Ideal for transformations and aggregations Pre-computation of slow queries Optimization for frequently used computations Detailed configuration details: Materialized Views
Delta Sink	Stream records to a Delta tables.	Product documentation: https://docs.databricks.com/en/dlt/dlt-sinks Limitations: https://docs.databricks.com/en/dlt/dlt-sinks#limitations
Kafka Sink	Stream records to a Kafka topic.	Product documentation: https://docs.databricks.com/en/dlt/dlt-sinks Limitations: https://docs.databricks.com/en/dlt/dlt-sinks#limitations
Foreach Batch Sink	Enables processing each micro-batch with custom logic similar to the Spark Structured Streaming foreachBatch functionality. With the ForEachBatch sink, you can transform, merge, or write streaming data to one or more targets that do not natively support streaming writes	Product documentation: to be release (Private Preview) Limitations: Similar to https://docs.databricks.com/en/dlt/dlt-sinks#limitations

General Data Flow Spec Configuration

Set as an attribute when creating your Data Flow Spec, refer to the Data Flow Spec Reference documentation for more information:

Detailed Target Type Configuration Details

Materialized Views