Microsoft has announced the general availability (GA) of Azure Cosmos DB mirroring in Microsoft Fabric with support for private networks, allowing enterprises to replicate operational data securely while maintaining strict network isolation. The update enables organizations to mirror Cosmos DB data into Microsoft Fabric and OneLake without exposing databases to the public internet. 

The feature addresses a major enterprise requirement: combining real-time analytics with strong network security for cloud databases.

Secure Mirroring Without Public Network Exposure

With the new capability, companies using Azure Virtual Networks (VNETs) or private endpoints can mirror their operational Cosmos DB data into Fabric workspaces while preserving their existing network security configurations. 

Previously, some mirroring scenarios required opening access through public endpoints. The new support ensures that organizations can replicate data to Fabric without compromising security boundaries.

This makes the feature especially attractive for industries with strict compliance requirements such as finance, healthcare, and government.

Real-Time Data Replication Into Fabric

Mirroring enables Cosmos DB data to be continuously replicated into Microsoft Fabric’s OneLake storage layer in near real time. 

Once mirrored, the data can be used across Fabric services for analytics and AI workloads, including:

• Power BI dashboards

• Spark notebooks

• Data science pipelines

• Machine learning training

Because the system replicates data automatically, developers no longer need to build complex ETL pipelines to move operational data into analytics environments.

Bridging Operational and Analytical Workloads

Azure Cosmos DB mirroring is part of Microsoft’s push toward HTAP (Hybrid Transactional/Analytical Processing) architectures.

In this model:

• Operational workloads continue running in Cosmos DB

• Analytical workloads operate on the mirrored copy in Fabric

This separation allows organizations to analyze live data without affecting application performance or consuming additional database request units. 

Designed for AI and Real-Time Analytics

The feature also supports AI and modern analytics scenarios. Once data is mirrored into Fabric, organizations can run:

• Real-time analytics

• Fraud detection pipelines

• Personalization engines

• IoT data analysis

These capabilities enable companies to build AI-powered applications directly on operational data streams.

Part of Microsoft’s Fabric Data Platform Vision

The update aligns with Microsoft’s broader strategy around Microsoft Fabric, its unified data platform that integrates data engineering, analytics, and AI workloads into a single environment.

By connecting Azure Cosmos DB—Microsoft’s globally distributed NoSQL database—with Fabric’s analytics capabilities, the company aims to reduce architectural complexity while enabling real-time, AI-ready data pipelines.

Source: Microsoft