Managing environment-specific configurations is a fundamental part of building scalable and maintainable data pipelines. In Databricks, this is typically handled using cluster environment variables or secret scopes. Until recently, Microsoft Fabric lacked a native solution for this, making it difficult to develop flexible notebooks and deploy them seamlessly across multiple environments. The good news is that Fabric now introduces Variable Library, currently in public preview—a feature that allows you to define and manage reusable configuration variables centrally, making it easier to standardize and automate deployments across notebooks, pipelines, and environments.
What Is Variable Library?
The Variable Library in Fabric is a centralized store of key-value pairs (variables) that you can define once and reference across your workspace. Notebooks integration with variable libraries are available for preview since June 2025.
When to use Variable Libarary?
The Variable Library is useful in scenarios where notebooks or pipelines rely on environment-dependent parameters such as lakehouse names, file paths, or configuration flags. Traditionally, these values are hardcoded within notebooks, requiring manual updates during environment promotion.
By centralizing configuration management, the Variable Library enables automated substitution of environment-specific values during deployment. This is especially valuable when promoting assets from development to higher environments such as UAT or production. When integrated into Fabric’s deployment pipeline, the Variable Library ensures that the appropriate variables are injected at runtime, significantly reducing complexity in orchestration logic and enhancing both maintainability and scalability of the solution.
Setting up Variable Library
As the Variable Library is still under preview, you may need to ask your tenant admin to enable this feature in the Admin portal. Detail steps may refer to Get started with Variable libraries – Microsoft Fabric | Microsoft Learn.
- In your Fabric workspace, click on the “New Item” and select “Variable library” under the Develop data category.
- Enter the Variable Library name as “Variables_Environment_Config” and also choose a location to save. You may also assign to a task.
- In your variable library, click “New variable” to create new variables. Then, create “Alternative value sets” for different environments, e.g. dev and prod. In this case, I am working on the dev environment, so I marked the dev value sets as “Active”.
- Create a new notebook in your workspace, then test your variable library with the following codes.
notebookutils.variableLibrary.getLibraryis a method in Microsoft Fabric notebooks to access and retrieve a specific Variable Library. Once the Variable Library object is retrieved, you can then access individual variables within that library using dot notation
# Usage
# variable_library_object = notebookutils.variableLibrary.getLibrary("YourLibraryName")
# variable_value = variable_library_object.YourVariableName
vs_config = notebookutils.variableLibrary.getLibrary("Variables_Environment_Config")
print(f"You are working on {vs_config.environment} environment.")
print(f"The threshold is set to {vs_config.threshold}")
if vs_config.flag:
print("Hello World!")
print(f"The timestamp was set to {vs_config.date}")
# Output
# You are working on dev environment.
# The threshold is set to 100
# Hello World!
# The timestamp was set to 2025-01-14T16:15:20.123Z- After testing the notebook, you may run your deployment pipeline to deploy the Variable Library and the Notebook to higher environment.
- After deployment, goto the Variable Library in your Production workspace. You will see that the Active variable set is the “Default value set”. In the Alternative value set, set the Production value set as Active. You need to choose the active value set manually as this variable library in newly created.
- Run the notebook under your Production workspace. You should see the following output as the notebook is retriving the variable value set for Production environment.
Best Practices and Limitations
- Modular Design: You may create multiple variable libraries within a Fabric workspace. It is recommended to group related variables into a dedicated library aligned to a specific domain or component (e.g., ingestion, transformation, reporting).
- Naming Conventions: Use consistent prefixes to categorize variables by purpose or domain—for example,
sales_,etl_, oranalytics_. This improves discoverability and avoids naming collisions in large projects. - Secure Handling: Avoid storing secrets such as passwords, API keys, or tokens in the Variable Library. Since variable values are committed to source control when using Git integration, any sensitive information should instead be managed via Azure Key Vault or Fabric Connections (a secure credential mechanisms).
- Across Workspace: Each Variable Library is scoped to a single workspace. You can only access a variable library from notebooks and pipelines within the same workspace where the library is defined.