Installation¶

Prerequisites¶

Python¶

Make sure you have Python 3.11 or higher installed. You can check your Python version by running:

python --version

No other system-level dependencies are required. Unlike Microsoft's upstream dbt-fabric adapter (which depends on pyODBC and requires installing ODBC drivers like msodbcsql18 and unixODBC), this adapter uses mssql-python -- Microsoft's official pure Python driver for SQL Server and Fabric. It handles TDS protocol communication natively, so there is nothing else to install on your system.

Install dbt-fabric-samdebruyn¶

Data Warehouse (T-SQL)Lakehouse (Spark SQL)

Install dbt-fabric-samdebruyn for use with Fabric Data Warehouse:

pip install dbt-fabric-samdebruyn dbt-core

This is a drop-in replacement for the original dbt-fabric adapter. If you are migrating, run pip uninstall dbt-fabric first.

Install dbt-fabric-samdebruyn with the Spark extra for use with Fabric Lakehouse:

pip install dbt-fabric-samdebruyn[spark] dbt-core

The [spark] extra installs dbt-spark as a dependency, which provides the base Spark SQL adapter that the FabricSpark adapter builds on.

The [spark] extra is only needed for Lakehouse

If you only use Fabric Data Warehouse, you do not need the [spark] extra. The base pip install dbt-fabric-samdebruyn dbt-core is sufficient.

See the Lakehouse guide for configuration and usage details.

That's it. No ODBC driver setup, no platform-specific steps. The adapter works the same on Linux, macOS, and Windows.