Skip to Content

Bronze Lakehouse

Last modified by Dimitri Rupp on 2026/06/11 09:30
On this page

Bronze Lakehouse

Information

Medallion-Stage: Bronze — Raw Ingestion Layer
Quality goal: As close to the source as possible

Overview

The Bronze Lakehouse layer contains data from various source systems with the data quality matching as close to the source as possible. Data is stored in distinct Lakehouses depending on the source system. Depending on the source, the orchestration method used to get data into these Lakehouses can be a Data Pipeline, Gen2 Dataflow or a Synapse Notebook.

Source-to-Bronze Flow

The diagram below shows which source system feeds which dedicated Bronze Lakehouse, and which orchestration method is used.

flowchart LR
    subgraph SRC["**Source systems**"]
        direction TB
        S1[BMD
ERP]
        S2[Rimo DWH
SPL Software]
        S3[Board
Planning DB]
        S4[SharePoint
Lists & Files]
        S5[H3A
Database]
    end

    subgraph BRZ["**Bronze Lakehouses**"]
        direction TB
        L1[(BMD Bronze)]
        L2[(Rimo Bronze)]
        L3[(Board Bronze)]
        L4[(SharePoint Bronze)]
        L5[(H3A Bronze)]
    end

    S1 -->|Data Pipeline| L1
    S2 -->|Data Pipeline| L2
    S3 -->|Gen2 Dataflow| L3
    S4 -->|Gen2 Dataflow / Notebook| L4
    S5 -->|Data Pipeline| L5

    BRZ --> SIL[Silver Lakehouse
cleansed & conformed]

    classDef src fill:#e3f2fd,stroke:#1976d2,stroke-width:1px
    classDef brz fill:#f4a261,stroke:#b85c00,stroke-width:1px,color:#000
    classDef sil fill:#c0c0c0,stroke:#555,stroke-width:1px

    class S1,S2,S3,S4,S5 src
    class L1,L2,L3,L4,L5 brz
    class SIL sil

The five Bronze Lakehouses

🗄️ BMD

Data from the BMD database — SPL Tele's ERP system. Hosts financial postings, customer and vendor master data, and accounting transactions.

→ BMD Bronze Lakehouse

🗄️ Rimo

Data from the Rimo DWH — SPL Tele's own software product. Provides project, resource and operational data used by Power Users.

→ Rimo Lakehouse

🗄️ Board

Data from the Board database — used for planning and controlling. Loaded via Gen2 Dataflow.

→ Board Bronze Lakehouse

🗄️ SharePoint

Data from SharePoint sites — typically lists, libraries and Excel files maintained by business teams. Used to bring decentralised data into the Lakehouse without a dedicated source database.

→ SharePoint Bronze Lakehouse

🗄️ H3A

Data from the H3A database — third-party operational system. Loaded via Data Pipeline.

→ H3A Bronze Lakehouse

Orchestration methods

Three methods are in use, chosen per source system:

MethodTypical useStrength
Data PipelineDatabase sources (BMD, Rimo, H3A)Native Fabric orchestration; copy + transform; parameterizable
Gen2 DataflowSources reachable via Power Query (Board, SharePoint Lists)Low-code; reuses existing Power Query knowledge
Synapse NotebookComplex transformations, file parsing, custom logicFull PySpark / Python control; ideal for non-trivial SharePoint payloads

Design principles

  • One Lakehouse per source system — keeps Bronze stable and the lineage obvious.
  • Schema close to source — no joins, no enrichment in Bronze; only raw or near-raw tables.
  • Idempotent loads — pipelines and dataflows can be re-run without breaking history.
  • Append where possible, full-load where cheaper — chosen per table based on volume and change pattern.

Where Bronze fits in the bigger picture

Information

Bronze → Silver → Gold. Bronze is only the raw landing zone. Cleansing, conforming and business-friendly modelling happens in the Silver and Gold Lakehouses.

Verwandte Themen

Kuratiert — kann direkt im Editor ergänzt werden://

Automatisch vorgeschlagen über gemeinsame Tags (architektur, fabric): 

Status: Migriert — Team-Review ausstehend · Owner: (festlegen) · Letzter Review: 2026-06-11