Fabric Mirroring for Azure SQL Managed Instance now in Public Preview

What is the fastest way to analyze my operational data from disparate sources?

With all of us witnessing the need for data driven decisions in businesses, it has been hard to be agile in analyzing the right data at the right time, due to it being siloed at various sources, due to complicated replication technologies and due to sparse resources and knowledge needed to set up replication. This situation pushed companies to only do holistic analytics for priority projects, where they could allocate resources. Exploration for non-obvious opportunities is often below the bar to prioritize and execute in an agile enough way to generate competitive advantage. Fabric Mirroring is a solution to this problem.

What is Fabric Mirroring?

Fabric mirroring is designed to facilitate advanced analytics scenarios by seamlessly synchronizing data from operational databases such as Azure SQL Managed Instance into Microsoft Fabric’s OneLake.

Mirroring continuously copies the chosen data from your Azure SQL Managed Instance into Microsoft Fabric’s OneLake with a handful of setup clicks and with near real-time latency. This enables you to leverage the full suite of analytics capabilities within Microsoft Fabric, making it easier to derive insights and drive business decisions.

Unlike alternative database replication tools requiring experts, Mirroring is a first-party feature in Fabric, specifically tailored for Fabric as a destination, ensuring that setting up and running the replication is as simple and smooth as possible.

With this public preview, Azure SQL Managed Instance joins the already available mirroring capabilities for Azure SQL Database, Azure Cosmos Database, Snowflake - for unified and familiar experience across data sources.

Benefits of Fabric Mirroring

Fabric Advanced Analytics Capabilities within arm's reach

Before Fabric Mirroring: Having to manually stitch data sources together and use various tools for advanced scenarios introduced learning curve and complexity in organization's daily operations, with significant focus spent on ensuring that scenarios are set up and run properly.

With Fabric Mirroring: With mirrored data being always in sync with its origin and a "first-class citizen" in Fabric OneLake, your operational data from SQL Managed Instance is able to be cross referenced with any other data in Fabric OneLake, for unified analytics benefit across your data sources. Other Fabric scenarios such as data science experiments are unlocked as well as direct lake mode in Power BI for fast data access and Copilot capability. 

Ease of Use

Before Fabric Mirroring: Setting up data synchronization for analytics required technical expertise and complex configurations to set up complex ETL (Extract, Transform, Load) setups to move data from operational databases to analytics platforms. Booking the experts to configure the setups also meant longer lead times to get to insights. Customers also had to manage multiple tools and processes to keep their data up-to-date. Simpler setups relied on manual data exports, which were also time consuming and prone to errors.

With Fabric Mirroring: The setup is straightforward and integrated within the Microsoft Fabric environment. Users can enable mirroring with just a few clicks, without needing extensive technical knowledge. This ease of use allows businesses to focus more on analyzing data rather than managing data movement. It also unlocks next levels of agility in being able to quickly set up analytics scenarios without writing business cases to justify the investment into ETL and waiting for experts to become available.

Simplified Maintenance and Monitoring

Before Fabric Mirroring: Ongoing maintenance and monitoring of data synchronization setups were labor-intensive. IT teams had to constantly check the health of the synchronization and ensure that data was being accurately copied.

With Fabric Mirroring: Microsoft Fabric handles the heavy lifting. The platform provides built-in monitoring capabilities, allowing users to easily track the status of their mirrored data. This reduces the operational burden on IT teams and ensures that any issues are promptly addressed.

Cost Efficiency

Before Fabric Mirroring: Implementing data synchronization solutions often involved additional costs for hardware, software, and third-party services.

With Fabric Mirroring: The feature is built into the Azure SQL Managed Instance service and Fabric, eliminating the need for extra resources. This integration helps reduce overall costs while providing a robust solution for data synchronization. Microsoft does not charge for compute used to transfer the data, nor for ingress fees. There is a generous (terabytes) amount of storage offered for mirroring, based on your Fabric capacity size.

How Fabric mirroring works

Fabric Mirroring for Azure SQL Managed Instance works by reading changes from database transaction log file, in a similar way to Change Data Capture technology, except that Mirroring does not save the data back into change tables, but publishes it to Fabric OneLake. This ensures great reliability and performance with least amount of resource tax on the source database.

Moreover, unlike CDC, DDL’s like add/drop column are also supported on actively mirrored tables. 

Once the data is mirrored into OneLake, it will be in delta format, ready for immediate consumption across all Fabric experiences and features like Power BI with new Direct Lake mode, Data Warehouse, Data Engineering, Lakehouse, KQL Database, Notebooks and co-pilots.

Conclusion

Fabric Mirroring for Azure SQL Managed Instance represents a significant step forward in enabling advanced analytics scenarios. By automating and simplifying the process of copying operational data into Microsoft Fabric, it allows businesses to gain insights in an agile way and with minimal effort. As this feature is currently in Public Preview, we encourage our customers to explore its capabilities and provide feedback to help us refine and improve the service.

In summary, Fabric Mirroring not only enhances your analytics capabilities but also makes it easier and more cost-effective to manage data synchronization.

We look forward to hearing your thoughts and feedback on Fabric Mirroring. Please share your experiences in the comments below!

Availability notes

Mirroring for Azure SQL Managed Instance is available for Azure SQL Managed Instances deployed in Azure regions except: East US 2; West US 2; Central US; West US. These regions will be enabled for SQL Managed Instance mirroring in near future.

Mirroring for SQL MI is only available for your Managed Instances with update policy configured as "Always up to date".

Learn more

To get started and learn more about Mirroring Azure SQL Managed Instance in Fabric, its pre-requisites, setup, FAQ’s, current limitations, and tutorial, click here to read all about it. 

Try Azure SQL Managed Instance for free.

Subscribe to SQL MI LinkedIn Newsletter today!

Try new SSMS 21 Preview with Copilot with your SQL MI.

What's new in Azure SQL Managed Instance?

Fabric Mirroring also exists for Azure SQL Database and is now Generally Available. Read about it here