Microsoft Fabric Open Mirroring

Turn database changes into continuously mirrored Fabric data.

Omni Loader captures committed changes, produces Fabric Open Mirroring change files, and publishes them through OneLake or a controlled staging path.

5
Source families

Oracle, SQL Server, PostgreSQL, MySQL, and Sybase.

Parquet
Mirroring files

Emit Fabric-compatible files with durable progress.

1 path
Backfill + CDC

Seed the mirror, catch up, and stay current.

Fabric-native delivery

Publish the shape Open Mirroring expects.

The Fabric sink groups source changes into target table specifications and advances its cursor only after its delivery contract succeeds.

Initial snapshot backfill
Insert, update, and delete change files
OneLake or local staging workflows
Per-destination lag and checkpoint

One source reader

Fabric does not add another capture process.

Use the same captured stream for Fabric and other destinations. A Fabric delay accumulates on its own durable cursor.

Independent Fabric backlog and retry
Parallel fan-out to other destinations
Capture mode chosen by source access
Control-plane throughput and health

Keep exploring

See the connected parts of the platform.

Keep Fabric current from the logs you already write.

Tell us the source, table set, Fabric path, and freshness goal. We will map backfill, CDC, and delivery.