Mirroring for Google BigQuery in Microsoft Fabric is now generally available. For organizations running critical workloads on BigQuery, this means a simpler, faster, and lower-cost path to bringing your data into the rest of your analytics estate, with production support, an enterprise SLA, and no pipelines to build or maintain.
Bringing BigQuery data into OneLake, no pipelines required
Mirroring is Fabric's turnkey replication capability. Instead of building and operating ETL pipelines to move BigQuery data into an analytics environment, you point Fabric at your BigQuery dataset and let Fabric handle the rest. Snapshots, ongoing change capture, and schema evolution are all managed for you, with no scheduling, no transformation logic, and no refresh windows.
The mirrored data lands in OneLake as open Delta tables, which means it is immediately available to every Fabric workload, including Power BI (with Direct Lake), the SQL analytics endpoint, Data Engineering notebooks, Data Science, Real-Time Intelligence, and Copilot, as well as any Delta-compatible engine outside of Fabric.
The following short walkthrough shows how to set up Mirroring for Google BigQuery end to end, from creating the mirrored database item in your Fabric workspace to validating that changes flow through into OneLake.
Figure: Video walkthrough of setting up Mirroring for Google BigQuery in Microsoft Fabric, from creating the mirrored database item to validating replication.
What you can do with Mirroring for Google BigQuery
Once your data is in OneLake, the options open up:
- Replicate BigQuery tables continuously with no ETL and no operational overhead.
- Query BigQuery data side by side with data from Oracle, Snowflake, SQL Server, Cosmos DB, Databricks, and other mirrored sources in a single Fabric warehouse.
- Build Direct Lake semantic models directly on top of mirrored BigQuery tables, with the freshness customers expect from DirectQuery and the performance they expect from Import, without either compromise.
- Let schema evolution happen automatically. New columns, type changes, and new tables on the BigQuery side flow through to OneLake without touching a pipeline.
- Combine BigQuery data with real-time streams, notebooks, and Copilot experiences to power a unified AI-ready foundation.
Key capabilities
- Support for BigQuery datasets in any Google Cloud region.
- Connectivity options that fit real enterprise environments, including Virtual Network (VNET) support and on-premises Data Gateway (OPDG, v3000.286.6 or greater).
- Automatic schema evolution as source tables change over time.
- Enterprise-ready security and governance out of the box. Microsoft Purview, sensitivity labels, workspace roles, and OneLake data access controls all apply to mirrored data.
- No Fabric replication compute cost and no network ingress fees. You only pay for the analytics compute you actually run.
For the full list of prerequisites, supported scenarios, and current limitations, see the Mirroring for Google BigQuery documentation.
Why this matters
Historically, teams running BigQuery that wanted their data available across the broader Microsoft ecosystem had two options, build custom ingestion, or accept the latency and cost of cross-cloud querying. Neither scales well. Mirroring changes the pattern:
- Zero-ETL by design. Continuous replication with no pipelines to author or babysit.
- Multi-cloud without lock-in. BigQuery data lands in the open Delta Lake format, alongside data from every other cloud and every other Mirroring source.
- Best-in-class Power BI performance. Direct Lake over mirrored tables gives customers sub-second reporting on always-current data.
- AI and Copilot ready. OneLake is the same substrate that powers Copilot in Fabric, so mirrored BigQuery data is instantly available to the AI experiences your teams are building.
- Lower total cost. Zero replication compute, zero ingress fees, and materially reduced downstream capacity consumption in customer-reported workloads.
What customers are saying
Many large enterprises have been successfully running Mirroring for Google BigQuery on production workloads throughout preview:
"Microsoft Fabric Mirroring for Google BigQuery lets us replicate terabyte-scale production data directly into OneLake with no ingestion delay. Schema changes that once took one to two weeks are now available almost immediately giving our marketing teams data that directly drives near real time campaign and content decisions."
— Raj Kapil, MSCI Business Technology
"Mirroring and Direct Lake fundamentally changed our BI economics. We cut background capacity consumption by 83%, reduced our monthly compute bill by 75%, and eliminated manual refresh management entirely. Our data now updates the moment GCP updates, with zero engineering overhead, while we keep full control of table management on the GCP side."
— Christopher Bonnard, Head of BI, Galeries Lafayette
Get started
Mirroring for Google BigQuery is available today, in every Fabric region. If you have BigQuery data and a Fabric workspace, you can replicate into OneLake in minutes.
Microsoft Employee