What is DBT?

DBT stands for “Data Build Tool.” It is an open-source tool that analytics engineers to transform data in their warehouse easily and in a scalable manner. It is a declarative ELT framework that allows transformations to be expressed as SELECT statements and corresponding configurations in YAML files.

Microsoft Fabric has a dbt adapter, ‘dbt-fabric’ and discussing how to use it will be the focus of this blog.

Data Warehouse and T-SQL Functionality

Fabric (Synapse) Data Warehouse is a lake (OneLake) centric SaaS experience. It supports cross-database queries that can be created using tools like Visual Query Editor, SQL Query Editor, and SQL Server Management Studio (SSMS). Users from any experience level can easily grasp the basic ideas of using a Fabric warehouse due to its very intuitive nature.

Fabric Warehouse supports both read and write operations unlike the SQL Analytics endpoint of a Fabric Lakehouse. Thus, the warehouse supports more T-SQL functionality like COPY INTO, CREATE TABLE AS SELECT (CTAS), INSERT..SELECT, and SELECT INTO are available in a warehouse.

Connecting DBT to a Fabric Data Warehouse

Creating a DBT project and using it with a Fabric data warehouse is simple process. To accomplish this, first, we need to install the following tools in our system.

• Python version 3.7 (or higher)
• The Microsoft ODBC Driver for SQL Server
• The dbt-fabric adapter. It can be installed from the PyPI repository using <pip install dbt-fabric> command

After installing these tools, we can start creating a DBT project using the <dbt init> command.

Let’s create a project named <dbt_fabric_demo>. We are assuming that we already have a data warehouse in Fabric.

We need to configure the <profiles.yml> file accordingly. Following is the <profiles.yml> file that we are using in our case.