dbt-fabric

May 17, 2026 · View on GitHub

dbt adapter for Microsoft Fabric Synapse Data Warehouse.

The adapter supports dbt-core 1.4 or newer and follows the same versioning scheme. E.g. version 1.1.x of the adapter will be compatible with dbt-core 1.1.x.

Documentation

We've bundled all documentation on the dbt docs site

Installation

This adapter requires the Microsoft ODBC driver to be installed: Windows | macOS | Linux

Debian/Ubuntu

Make sure to install the ODBC headers as well as the driver linked above:

sudo apt-get install -y unixodbc-dev

macOS (Apple Silicon)

Install unixODBC and the ODBC driver via Homebrew:

brew install unixodbc msodbcsql18

Modern Homebrew (post-Sonoma) ships libodbc.3.dylib, but pyodbc wheels are often compiled against libodbc.2.dylib. This mismatch causes the following error at runtime:

Library not loaded: /opt/homebrew/opt/unixodbc/lib/libodbc.2.dylib

Fix — recompile pyodbc against the installed unixODBC:

export LDFLAGS="-L/opt/homebrew/opt/unixodbc/lib"
export CPPFLAGS="-I/opt/homebrew/opt/unixodbc/include"
pip install --force-reinstall --no-binary :all: pyodbc

Alternative — create a compatibility symlink:

ln -s /opt/homebrew/opt/unixodbc/lib/libodbc.3.dylib \
      /opt/homebrew/opt/unixodbc/lib/libodbc.2.dylib

Latest version:

pip install -U dbt-fabric

Performance guidance for large projects

Fabric Warehouse DDL operations (e.g. CREATE TABLE, sp_rename) hold catalog locks that can block sys.tables/sys.views reads from concurrent dbt sessions. With many models and high thread counts this causes list_<schema> steps to stall for minutes.

Strongly recommended for projects with 500+ models or concurrent dbt runs:

# profiles.yml
my_fabric_project:
  target: dev
  outputs:
    dev:
      type: fabric
      # ... connection settings ...
      threads: 4          # keep low (4–8) to reduce catalog lock pressure
      query_timeout: 30   # fail fast on blocked catalog reads (seconds)

In your dbt_project.yml:

flags:
  cache_selected_only: true  # only list schemas for models in the current run

Or pass --no-populate-cache on the CLI for a single run. This prevents dbt from listing every schema in the warehouse upfront, significantly reducing catalog read pressure during concurrent runs.

dbt-fabric

Documentation

Installation

Performance guidance for large projects

Changelog

Contributing

License

Code of Conduct