What is Microsoft Fabric API for GraphQL?

As a data engineer, I constantly seek efficient ways to query data and improve accessibility, making it more valuable for data-driven businesses. If you’re interested in optimizing data workflows, this blog is for you.

First, let’s explore what GraphQL is.

What is GraphQL?

GraphQL is an API query language that enables precise data retrieval, minimizing both over-fetching and under-fetching. Unlike REST, which requires multiple endpoints for different data sets, GraphQL allows clients to specify and fetch only the required data in a single request, optimizing performance and reducing network overhead.

In Microsoft Fabric, GraphQL facilitates direct access to structured data across workloads such as data engineering, analytics, and business intelligence. It enables developers to query datasets efficiently, retrieve related entities in a single operation, and streamline data interactions without unnecessary complexity.

By integrating GraphQL, Microsoft Fabric supports dynamic data access, improving responsiveness in applications, enhancing analytical workflows, and enabling real-time data retrieval without dependence on fixed API structures.

GraphQL Operations

Microsoft Fabric supports only two core operations (Query and Mutation) in the latest release (Nov 2024) but there are three as listed below:
1. Queries are used to retrieve specific data from Microsoft Fabric. Unlike traditional REST APIs, GraphQL queries enable clients to request only the required fields, minimizing data transfer and improving performance.
2. Mutations allow modifications to data, such as INSERT, UPDATE, or DELETE records. These operations ensure that changes are executed reliably while maintaining data consistency.
3. Subscriptions enable real-time data updates by establishing a connection between the client and server. This is particularly useful for tracking changes in datasets without continuously polling the API.

Supported Fabric Items

1. Data Warehouse
2. Lakehouse via SQL Analytics Endpoint
3. Mirrored Databases via SQL Analytics Endpoint
4. Fabric SQL Databases
5. Azure SQL Databases
6. Datamarts

How to create an API for GraphQL

Before using the GraphQL API, ensure you have necessary permissions to create fabric items and interact with data.

1. To create an API for GraphQL select New Item from your workspace and under Develop data select API for GraphQL. Give a name to the item and select Create

2. In the newly created API, choose Select data source. You can connect to already existing supported data sources. One API item can connect to multiple data sources simultaneously.

3. Choose the connectivity option, single sign-on (SSO) to access API through client credentials assuming users already have access to the data source or, saved credentials such as API keys or tokens.

To connect to data source, make sure the registered app in Microsoft Entra ID have GraphQL.Execute.All API permissions. To learn more about creating service principal/registering apps for Microsoft Fabric refer this article and click here to know more about permissions.
4. Next choose the data source you want to connect. Lets choose Fabric SQL Database and select Connect.

5. In the Get data, choose the objects you want to query in your GraphQL schema and click Load.

6. Once data is loaded, you can see the page is divided into four sections:
Schema Explorer that shows all the loaded tables, views and stored procedures
Query Editor to write the queries and mutations
Query Variable Editor to define dynamic variables for the queries