Azure API Management - Fabric GraphQL Integration

This repository demonstrates how to integrate Fabric GraphQL API with Azure API Management (APIM).

Configure Fabric

1. Create a new Lakehouse, Load the data (fabriq-graphql/factory_iot_data.csv) and transform them into a table
2. Create a new API For GraphQL item, bind it to the table.
3. Copy the endpoint and set value in infra/main.parameters.json

   "fabricGraphQLEndpoint": {
      "value": "https://cb0442cc43ea4c819fea0bba9b62f870.zcb.graphql.fabric.microsoft.com/v1/workspaces/cb0442cc-43ea-4c81-9fea-0bba9b62f870/graphqlapis/64f58335-5d12-441d-b5e5-51778048a084/graphql"
    }

Configure The sample

1. Trigger azd up to update the Azure APIM Configuration
2. Get the name of the created managed identity apim-mi-xxxxxxxx
3. At the workspace level,using the Manage Acess menu; assign the managed identity apim-mi-xxxx as a contributor of the workspace.

Test

cd fabric-graphql
azd env get-values > .env
./test-api.sh


uv venv
source .venv/bin/activate
uv run fabric_graphql_apim.py 

The output should look like

Using FABRIC_GRAPHQL_API_URL: https://apim-tgebslojbs6y2.azure-api.net/fabric-graphql

query {
  factory_iot_datas(first: 10) {
     items {
        Timestamp
        BuildingID
        DeviceID

     }
  }
}

Making request to: https://apim-tgebslojbs6y2.azure-api.net/fabric-graphql
Headers: {'Content-Type': 'application/json', 'Ocp-Apim-Subscription-Key': '4a33de485f7c432f9192a9a19cd1a79b'}
Response status code: 200
Response headers: {'Content-Type': 'text/plain; charset=utf-8', 'Date': 'Mon, 17 Nov 2025 08:25:46 GMT', 'Access-Control-Expose-Headers': 'x-ms-latency,RequestId', 'Transfer-Encoding': 'chunked', 'Strict-Transport-Security': 'max-age=31536000; includeSubDomains', 'x-ms-latency': 'overhead=1729;queryEngine=10080', 'x-ms-routing-hint': 'host001_graphql-006', 'x-ms-workload-resource-moniker': '64f58335-5d12-441d-b5e5-51778048a084', 'x-ms-root-activity-id': '31f510dd-fe75-45c1-a077-855e4dca8f2d', 'x-ms-current-utc-date': '11/17/2025 8:25:35 AM', 'X-Frame-Options': 'deny', 'X-Content-Type-Options': 'nosniff', 'request-redirected': 'true', 'home-cluster-uri': 'https://wabi-west-us3-a-primary-redirect.analysis.windows.net/', 'RequestId': '31988eb5-4cbe-44c0-9e2f-e75a3678b9e9', 'Request-Context': 'appId=cid-v1:4e244c7e-6d87-44a6-9022-c1aa461fa0c9'}
{
    "data": {
        "factory_iot_datas": {
            "items": [
                {
                    "Timestamp": "2025-11-11T17:36:24.407Z",
                    "BuildingID": "BLD-MAR-003",
                    "DeviceID": "EM-05B-G1"
                },...

Documentation:

• https://learn.microsoft.com/en-us/fabric/data-engineering/get-started-api-graphql
• https://learn.microsoft.com/en-us/fabric/data-engineering/api-graphql-azure-api-management

Fabric Rest to GraphQL

The files in fabric-rest-2-graphql folder declare and implement two REST operations on sensors:

API Operations

• GET /sensors: Returns a list of available sensors. This endpoint retrieves sensor records from the underlying Fabric GraphQL API and is typically used to browse or list sensors and their metadata.
• GET /sensors/{deviceid}: Returns details for a single sensor identified by its deviceid. This endpoint looks up a specific sensor and returns its details (for example, timestamp, building identifier, and device identifier). Internally, the REST call is translated into a filtered GraphQL query against the Fabric API.

Both endpoints require an API key provided via the Ocp-Apim-Subscription-Key header or query string, as configured in Azure API Management.

Test

cd fabric-rest-2-graphql
azd env get-values > .env
./test-api.sh

Fabric Rest to GraphQL MCP Server

We will utilize the MCP Servers APIM feature to present the new Sensors Rest API as an MCP (Model Context Protocol) server, allowing an Agent to manage the sensors data.

• APIM > APIs > MCP Servers
• Create MCP Server > Expose an API as MCP Server
• API Rest to GraphQL Fabric API
• API Operations : All
• Display Name: sensors-mcp

The Sensors Agent

Once the sensors-mcpsensor available, it's possible to use it with Agents.

Demo: MCP and Github Copilot Agent Mode

• Open .vscode/mcp.json Click on Start
• Open Github Copilot Side Window and interact with the agent.

MCP.Sensors.mp4

Orders REST API

The files in orders-rest-api folder implement a complete REST API for managing orders with full CRUD (Create, Read, Update, Delete) operations.

Features

• List Orders: Get all orders with optional filtering by status and limit
• Get Order: Retrieve a specific order by ID
• Create Order: Create a new order
• Update Order: Update order status, shipping address, or notes
• Delete Order: Remove an order from the system
• Fake Data: Automatically generates 20 sample orders for testing
• FastAPI: Modern Python framework with automatic OpenAPI documentation
• Pydantic: Data validation and serialization

Local Development and Testing

Installation

cd orders-rest-api
python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
pip install fastapi uvicorn pydantic

Running the API

python main.py

The API will be available at http://localhost:8000

• Interactive API docs (Swagger UI): http://localhost:8000/docs
• Alternative API docs (ReDoc): http://localhost:8000/redoc
• OpenAPI schema: http://localhost:8000/openapi.json

Test

cd orders-rest-api
./test_api.sh

API Operations

List Orders

GET /orders?status_filter=pending&limit=10

Optional query parameters:

• status_filter: Filter by status (pending, processing, shipped, delivered, cancelled)
• limit: Limit the number of results

Get Order by ID

GET /orders/{order_id}

Example:

curl http://localhost:8000/orders/ORD-2024-001

Create Order

POST /orders
Content-Type: application/json

{
  "customer_id": "CUST-12345",
  "customer_name": "John Doe",
  "customer_email": "john.doe@example.com",
  "items": [
    {
      "product_id": "PROD-001",
      "product_name": "Laptop",
      "quantity": 1,
      "unit_price": 999.99,
      "total_price": 999.99
    }
  ],
  "shipping_address": "123 Main St, City, State 12345",
  "notes": "Please handle with care"
}

Update Order

PUT /orders/{order_id}
Content-Type: application/json

{
  "status": "processing",
  "notes": "Updated notes"
}

Delete Order

DELETE /orders/{order_id}

Order Status Values

• pending: Order has been placed but not yet processed
• processing: Order is being prepared
• shipped: Order has been shipped
• delivered: Order has been delivered
• cancelled: Order has been cancelled

Sample Data

The API includes a fake data generator that creates 20 realistic orders with:

• 10 different customers
• 10 different products (laptops, accessories, peripherals)
• Various order statuses based on order age
• Calculated totals (subtotal, tax at 8%, shipping)
• Free shipping for orders over $100

Deployment with Azure API Management

The Orders API is integrated with Azure API Management. After deploying the infrastructure:

1. Deploy your Orders API backend to Azure (e.g., Azure App Service, Container Apps)
2. Update the ordersApiBackendUrl parameter in infra/main.parameters.json
3. Run azd provision to update the APIM configuration

Access the API through APIM:

export ORDERS_API_URL=<from azd env>
export ORDERS_APIM_SUBSCRIPTION_KEY=<from azd env>

curl "${ORDERS_API_URL}/orders" \
  -H "Ocp-Apim-Subscription-Key: ${ORDERS_APIM_SUBSCRIPTION_KEY}"

📚 Additional Resources

• GitHub GraphQL API Documentation
• Azure API Management GraphQL Support
• GraphQL Best Practices

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.