mirroring vs the bronze layer

Mirroring is a good fit for a bronze layer in most cases. Even though the Mirrored Database item shows up alongside warehouse artifacts in the UI, the mirrored data physically lands in OneLake as Delta Parquet, continuously replicated from source with no transformation. That is exactly what a bronze layer is supposed to be.

The usual pattern is to create the Mirrored Database, then create a Lakehouse and add OneLake shortcuts to the mirrored tables. From there your silver and gold layers read through the shortcut without any duplicated storage. Just keep in mind mirroring only covers the supported sources (Azure SQL DB / MI, Fabric SQL DB, Snowflake, Cosmos DB, plus Open Mirroring for custom sources), so it will handle only the parts of your bronze that come from those systems.

If this helped, please give a thumbs up and mark it as the accepted solution.

Best,
Shai Karmani

Let's connect in LinkedIn

Hello db042190​,

The key distinction is between the Fabric item and the architectural layer.

Bronze is a logical layer for raw data. Database Mirroring can serve that role because it continuously replicates supported source data into OneLake as Delta tables.

So I would not interpret its placement under Data Warehousing as meaning Mirroring comes after Bronze. The Fabric item category and its role in a medallion architecture are separate concepts.

One caveat is that not all Mirroring works the same way. Database Mirroring copies the data into OneLake, while Metadata Mirroring uses shortcuts to work with the source data.

https://learn.microsoft.com/en-us/fabric/onelake/onelake-medallion-lakehouse-architecture

Hi db042190​ 

What I would also say is that the medallion architecture always is reliant on what your use case is. Sometimes you might need bronze, silver, gold. Sometimes you might just need bronze and gold. It all depends on your data, your requirements, your transformations, and how this fits into your overall ways of working.