Synapse to Fabric: Your ADX Migration Guide 2025

The clock is ticking for Azure Synapse Data Explorer (ADX). With its retirement announced, a strategic Synapse to Fabric migration is now a critical task for data teams. This move to Microsoft Fabric’s Real-Time Analytics and its Eventhouse database unlocks a unified, AI-powered experience, and this guide will show you how.

This guide will walk you through the entire process, from planning to execution, complete with practical examples and KQL code snippets to ensure a smooth transition.

Why This is Happening: The Drive Behind the Synapse to Fabric Migration

Microsoft’s vision is clear: a single, integrated platform for all data and analytics workloads. This Synapse to Fabric migration is a direct result of that vision. While powerful, Azure Synapse Analytics was built from a collection of distinct services. Microsoft Fabric breaks down these silos, offering a unified SaaS experience where data engineering, data science, and business intelligence coexist seamlessly.

Eventhouse is the next evolution of the Kusto engine that powered ADX, now deeply integrated within the Fabric ecosystem. It’s built for high-performance querying on streaming, semi-structured data—making it the natural successor for your ADX workloads.

Key Benefits of Migrating to Fabric Eventhouse:

• OneLake Integration: Your data lives in OneLake, a single, tenant-wide data lake, eliminating data duplication and movement.
• Unified Experience: Switch from data ingestion to query to Power BI reporting within a single UI.
• Enhanced T-SQL Support: Query your Eventhouse data using both KQL and a more robust T-SQL surface area.
• AI-Powered Future: Tap into the power of Copilot and other AI capabilities inherent to the Fabric platform.

Phase 1: Assess and Plan Your Migration

Before you move a single byte of data, you need a clear inventory of your current ADX environment.

1. Document Your Clusters: List all your ADX clusters, databases, and tables.
2. Analyze Ingestion Pipelines: Identify all data sources. Are you using Event Hubs, IoT Hubs, or custom scripts?
3. Map Downstream Consumers: Who and what consumes this data? Document all Power BI reports, dashboards, Grafana instances, and applications that query ADX.
4. Export Your Schema: You’ll need the schema for every table and function. Use the .show and .get commands in the ADX query editor to script your objects.

Example: Scripting a Table Schema

Run this KQL command in your Azure Data Explorer query window to get the creation command for a specific table.

.get table YourTableName schema as csl

This will output the .create table command with all columns, data types, and folder/docstring properties. Save these scripts for each table. Do the same for your functions using .show function YourFunctionName.

Phase 2: The Migration – Data and Schema

With your plan in place, it’s time to create your new home in Fabric and move your data.

Step 1: Create a KQL Database and Eventhouse in Fabric

1. Navigate to your Microsoft Fabric workspace.
2. Select the Real-Time Analytics experience.
3. Create a new KQL Database.
4. Within your KQL Database, Fabric automatically provisions an Eventhouse. This is your primary database for analysis. You can also create “KQL Querysets” which are like saved query collections.

Step 2: Recreate Your Schema

Using the scripts you exported in Phase 1, run the .create table and .create function commands in your new Fabric KQL Database query window.

Step 3: Migrate Your Data

For historical data, the most effective method is exporting from ADX to Parquet format in Azure Data Lake Storage (ADLS) Gen2 and then ingesting into Fabric.

Example: One-Time Data Ingestion with a Fabric Pipeline

1. Export from ADX: Use the .export command in ADX to push your historical table data to a container in ADLS Gen2.Code snippet.
2. Ingest into Fabric: In your Fabric workspace, create a new Data Pipeline.
3. Use the Copy data activity.
   • Source: Connect to your ADLS Gen2 account and point to the exported Parquet files.
   • Destination: Select “Workspace” and choose your KQL Database and target table.
4. Run the pipeline. Fabric will handle the ingestion into your Eventhouse table with optimized performance.

.export async to parquet (
    h@"abfss://your-container@your-storage-account.dfs.core.windows.net/path/to/export"
)
<|
YourTableName

For ongoing data streams, you will re-point your Event Hubs or IoT Hubs from your old ADX cluster to your new Fabric Eventstream or KQL Database connection string.

Phase 3: Update Queries and Reports

Most of your KQL queries will work in Fabric without modification. The primary task here is updating connection strings in your downstream tools.

Connecting Power BI to Fabric Eventhouse:

This is where the integration shines.

1. Open Power BI Desktop.
2. Click Get Data.
3. Search for the KQL Database connector.
4. Instead of a cluster URI, you’ll see a simple dialog to select your Fabric workspace and the specific KQL Database.
5. Select DirectQuery for real-time analysis.

Your existing Power BI data models and DAX measures should work seamlessly once the connection is updated.

Example: Updating an Application Connection

If you have an application using the ADX SDK, you will need to update the connection string.

• Old ADX Connection String: https://youradxcluster.kusto.windows.net
• New Fabric KQL DB Connection String: https://your-fabric-workspace.kusto.fabric.microsoft.com

You can find the exact query URI in the Fabric portal on your KQL Database’s details page.

Embracing the Future

Completing your Synapse to Fabric migration is more than a technical task—it’s a strategic step into the future of data analytics. By consolidating your workloads, you reduce complexity, unlock powerful new AI capabilities, and empower your team with a truly unified platform. Start planning today to ensure you’re ahead of the curve.

Further Reading & Official Resources

For those looking to dive deeper, here are the official Microsoft documents and resources to guide your migration and learning journey:

1. Official Microsoft Documentation: Migrate to Real-Time Analytics in Fabric
2. Microsoft Fabric Real-Time Analytics Overview
3. Quickstart: Create a KQL Database
4. Get data into a KQL database
5. OneLake, the OneDrive for Data
6. Microsoft Fabric Community Forum