Targets and load methods

The fastest load is rarely reading rows and inserting them. It’s letting the warehouse bulk-load files: straight from your bucket or folder when the data is already in files, or from a stage that SQLLoad fills with Parquet. SQLLoad picks that route wherever it loads the data exactly, and plan names the route for every table.

How data gets in

Data	Into DuckDB	Into Snowflake	Into Databricks
Parquet files in S3 or a folder	DuckDB reads them with read_parquet	COPY from an external stage over the bucket	Staged by SQLLoad in the volume, then COPY INTO
CSV, JSON, NDJSON files	Read by SQLLoad, appended as Arrow	COPY from an external stage over the bucket	Read by SQLLoad, staged as Parquet in the volume, then COPY INTO
Postgres or Python rows	Appended as Arrow	Written as Parquet to the load_method’s stage, then COPY (or streamed or inserted, if the load_method says so)	Written as Parquet to the volume, then COPY INTO

There are no silent fallbacks. If the route a table needs can’t be used, plan fails for that table and says how to fix it.

Targets

A target is where loaded tables go. It names a connection and where tables are placed in it. A project can have several targets, for example one per environment; each source loads into its destination, or the project’s default_target.

Connection kind	Target writes	Places tables with
snowflake	Snowflake tables	loader_schema, and optionally database
databricks	Unity Catalog tables	loader_schema, in the connection’s catalog
duckdb	DuckDB tables	loader_schema
s3, filesystem	Parquet files, one folder per table	path

Warehouse targets

A Snowflake, Databricks or DuckDB target places tables in a schema:

[connections.warehouse]
kind = "duckdb"
database = "shop.duckdb"

[targets.local]
connection = "warehouse"
loader_schema = "raw"

Key	Meaning
connection	A snowflake or duckdb connection.
loader_schema	The schema for tables whose source sets no schema. Required.
database	Overrides the connection’s database.

A source’s schema overrides loader_schema for its tables. A DuckDB connection’s database is a file path, or :memory:. A Snowflake connection also needs a load_method; see Snowflake.

Each kind has its own page: Snowflake, Databricks, DuckDB and Parquet files.