Co-Author: Maraki Ketema, Principal Product Manager
Unlocking Data Value at Scale: Mirroring in Microsoft Fabric
In the modern data era, speed, scale, and simplicity are no longer luxuries—they're expectations. Organizations want to harness the power of their operational data in real time, without the overhead of complex ETL pipelines or latency-filled data movement. This is where Mirroring in Microsoft Fabric comes in, redefining how data is synchronized and activated across ecosystems.
Data is complex. It’s often scattered across multiple systems, stored in various formats, locked in silos and changing all the time — making it difficult to harness its full potential. Bringing this data together to power AI and BI workloads typically requires time-consuming ETL processes, custom pipelines, and deep technical expertise. There's no simple way to get started...until now.
What is Mirroring in Microsoft Fabric?
When should you use it—and more importantly, what value does it bring to you and your organization?
Let’s break it down.
Mirroring enables you to create a near real-time replica of your operational database directly within OneLake in Microsoft Fabric. OneLake is a fully managed, SaaS-based data lake built into Microsoft Fabric, designed to unify all your data—structured and unstructured—across domains. It provides a single, secure, and governed location where data from various sources can be seamlessly synchronized, stored, and analyzed, making it easier to unlock insights and power AI and analytics at scale. Mirroring is a zero-ETL, fully managed, and high-performance data movement mechanism that keeps your analytical and transactional systems stay in sync—minus the hassle.
Mirroring in Fabric provides a modern way of accessing and synchronizing your existing data estate continuously and seamlessly from any database or data warehouse into OneLake in Microsoft Fabric.
When to use Mirroring:
- Analyze operational data in near real-time
Example: Data from Azure Cosmos DB, Azure SQL Database, Azure Database for PostgreSQL, SQL Server or Snowflake needs to be analyzed with Fabric toolset (Power BI, SQL Notebooks, etc.).
- Avoid managing ETL pipelines - Mirroring eliminates the need to schedule and manage traditional ingestion pipelines.
- For a single source of truth across both operational and analytical systems
Mirroring keeps your data synchronized with the source system, so you’re always analyzing the latest data.
- Building a Lakehouse architecture and want real-time updates
Mirrored data lands in OneLake, so it’s seamlessly integrated with other Fabric workloads like Data Engineering, Data Science, and Real-Time Analytics.
Once mirrored, the data is replicated into OneLake in a Delta Lake format, making it immediately query able via Fabric experiences like Lakehouse, Data Warehouse, Power BI, Notebooks, or Spark.
What if your data isn't in any of the supported systems
Or what if it's spread across legacy databases, custom apps, and bespoke solutions—some of which are no longer even supported (Ex: SQL Server 2008).
This example highlights a challenge many customers face: a tangled mix of legacy data sources, on-prem systems, modern cloud platforms, CSV exports, and custom-built integrations. In this kind of environment, enhancing your data estate can feel overwhelming.
That’s where Open Mirroring comes in
Open Mirroring is Microsoft’s approach that allows any application or data provider to write change data directly into a mirrored database within Fabric. It gives you the flexibility to define your own mirroring source. Just convert your data into Parquet or CSV and load it into OneLake. From there, our replication engine takes over—automatically converting the data to Delta format and keeping an up-to-date replica in OneLake.
No matter where your data originates, Open Mirroring makes it accessible, current, and ready to power any Fabric workload.
What happens if your source data changes
No problem—Open Mirroring handles that too. Simply upload your change files in Parquet or CSV with the relevant watermark, and our replication technology will ensure the corresponding tables and rows in OneLake stay perfectly in sync.
Open Mirroring in Fabric is designed to be extensible, customizable, and open. This capability enables any applications and data ISVs to write change data directly into a Mirrored Database in Microsoft Fabric using the Open Mirroring public API.
Use Open Mirroring when
Your data resides in legacy systems, bespoke data solutions, and/or external data applications.
Open Mirroring is designed not just for our customers, but also for our partners. We're building an ecosystem that supports Open Mirroring sources from leading ISVs, including: Striim, Oracle GoldenGate, MongoDB ,CluedIn , DataStax (coming soon).
If Mirroring is supported for your source, we suggest using it as the preferred option. For data sources that are not supported by native mirroring, opt for Open Mirroring.
Key comparison: Mirroring vs Open Mirroring
| Feature | Mirroring | Open Mirroring |
| Data Movement | Yes (replicated into OneLake) | Yes (replicated into OneLake) |
| Latency | Near real-time | Depends on source system |
| Format in Fabric | Delta Lake | Delta Lake |
| Supported Sources | Snowflake, Azure SQL Database, Azure Cosmos DB, Azure SQL Managed Instance, Azure Database for PostgreSQL, Azure Databricks Catalog with more coming soon on the roadmap | Any source or application |
| Governance | Full control within OneLake | Full control within OneLake |
| Performance | High (optimized for Fabric) | High (optimized for Fabric) |
| Pricing Model | Compute to Replicate + Storage in Fabric is free (1TB/ CU limit) CU based consumption model to perform analytics on top on the stored data | Compute to Replicate + Storage in Fabric is free (1TB/ CU limit) CU based consumption model to perform analytics on top on the stored data |
Pricing considerations
Mirroring:
- Storage in OneLake: enjoy free Mirroring storage for replicas up to a certain limit based on the purchased capacity SKU you provision. For example, if you purchase F64, you will get 64 free terabytes worth of storage. OneLake storage is billed only when the free Mirroring storage limit is exceeded, or the provisioned compute capacity is paused*.
- *Free Mirroring storage can only be used for storing Mirroring replica data—no other data. Free Mirroring storage is also only for purchased capacities; it does not come with the Fabric free trial.
- Pay for Compute in Fabric for downstream analytics.
Benefits
- Real-time analytics on operational data with near-zero latency.
- No ETL pipelines required data is mirrored automatically.
- Supports multiple sources like Azure SQL Database, Azure Cosmos DB, and Snowflake etc., (Check the table for the full list).
- Integrated with Microsoft Fabric workloads such as Power BI, Spark, and SQL via OneLake.
- Efficient storage model with no duplicate storage costs.
- Consumption-based pricing—you pay only for compute when used.
- Faster time to insight with reduced data engineering effort.
- Always up to date through continuous synchronization with source systems.
- Enables better collaboration across teams with a single version of data truth.
Open Mirroring
- Storage in OneLake : enjoy free Mirroring storage for replicas up to a certain limit based on the purchased compute capacity SKU you provision. For example, if you purchase F64, you will get 64 free terabytes worth of storage. OneLake storage is billed only when the free Mirroring storage limit is exceeded, or the provisioned compute capacity is paused*.
- *Free Mirroring storage can only be used for storing Mirroring replica data—no other data. Free Mirroring storage is also only for purchased capacities; it does not come with the Fabric free trial.
- Pay for Compute in Fabric for downstream analytics.
Benefits
In addition to all the benefits from Mirroring, Open Mirroring offers additional advantages.
- Cheaper for read-heavy data sources.
- Great for unique data requirements, regulated environments.
Customer and Partner value
Both Mirroring and Open Mirroring embody Microsoft Fabric’s core principle of 'one copy of data, multiple analytical experiences.' In practice it will follow these examples:
For customers
- Faster time to insight with real-time data ingestion across multiple data sources and no ETL setup required, not to mention the no-code approach.
- Lower TCO by reducing infrastructure sprawl and eliminating tooling overlap.
- Agility to unify internal and external data seamlessly, while maintaining governance.
For Partners
- Frictionless integrations for cross-cloud and multi-vendor data ecosystems.
- Accelerated co-innovation with customers adopting multi-cloud strategies.
Closing
As data grows in volume, variety, and velocity, the traditional copy-and-load model is no longer effective. Mirroring and Open Mirroring in Microsoft Fabric offer two complementary approaches: providing control when you need performance, simplicity and offering flexibility when your data needs to stay where it is.
Mirroring, Open Mirroring is more than just a technical upgrade, it's a strategic enabler. By meeting your data wherever it resides, Fabric empowers you to modernize analytics, reduce data friction, and unlock new business models—all while ensuring trust and governance at scale.