Target Catalog and Schema

Applies To:	Pipeline Bundle
Configuration Scope:	Pipeline Data Flow Spec
Databricks Docs:	Set the target catalog and schema Configure pipelines LIVE schema (legacy)

Overview

Lakeflow Spark Declarative Pipelines (SDP) default publishing mode publishes materialized views and streaming tables to Unity Catalog using the target catalog and schema configured on the pipeline. Unqualified identifiers resolve to those defaults unless you override them with fully qualified names or USE CATALOG / USE SCHEMA in pipeline code.

Note

New Lakeflow Framework pipelines should set catalog and schema in the pipeline resource file. LFF sample bundles use default publishing mode.

Warning

Pipelines created before February 5, 2025 may still use legacy publishing mode and the LIVE virtual schema. Migrating an existing pipeline to default publishing mode requires destroying and redeploying with catalog and schema configured instead of the legacy target field. Pipeline tables are dropped and recreated, so the next run reprocesses all data.

The Framework supports:

• Publishing to multiple catalogs and schemas from a single pipeline

• Legacy publishing mode and the LIVE virtual schema for existing pipelines only

This is configured in two places:

1. In the Data Flow Spec, under the targetDetails section.

2. In the pipeline resource file (catalog and schema). See Configure pipelines.

Refer to the section Building a Pipeline Bundle for more information.