Skip to content
Bert Verbeek edited this page Mar 16, 2026 · 6 revisions

BC2ADLS - Business Central to Azure Data Lake Storage

Welcome to the bc2adls documentation! This tool enables you to export incremental data from Dynamics 365 Business Central to Microsoft Fabric or Azure Data Lake Storage.

� Get it from the Marketplace

The bc2adls extension is available as a free app on the Microsoft Marketplace (AppSource). You can install it directly into your Business Central environment:

📥 Download bc2adls from Microsoft Marketplace

Alternatively, the extension is also available as a PTE (Per-Tenant Extension) from the GitHub Releases page. See PTE App vs AppSource App for a detailed comparison of the differences between the two versions (mainly object number changes).

�📋 Table of Contents

Introduction

The bc2adls tool provides a powerful solution for exporting Business Central data to external storage platforms. It supports incremental data exports, ensuring that only changed data is transferred, which optimizes performance and reduces costs.

Export Options

bc2adls supports three different export options, each designed for specific use cases:

1. 🗄️ Azure Data Lake Storage (ADLS Gen2)

Export data to Azure Data Lake Storage Gen2 with full CDM (Common Data Model) support and Azure Synapse Analytics integration.

Best for:

  • Organizations already using Azure infrastructure
  • Advanced data transformation with Azure Synapse pipelines
  • Integration with multiple Azure services
  • CDM format requirements for Power BI consumption

Key Components:

  • Business Central extension (businessCentral/app)
  • Azure Synapse pipelines (synapse/)
  • CDM manifest support (CSV and Parquet formats)

📖 Read the Azure Data Lake Setup Guide


2. 📊 Microsoft Fabric with Notebooks

Export data directly to Microsoft Fabric Lakehouse and use notebooks to process delta files into Delta Parquet tables.

Best for:

  • Organizations using Microsoft Fabric
  • Direct integration with Power BI and Fabric ecosystem
  • Simplified setup without Azure Synapse
  • Modern lakehouse architecture

Key Components:

  • Business Central extension (businessCentral/app)
  • Fabric notebooks for data consolidation (fabric/)
  • Delta Parquet table format

📖 Read the Microsoft Fabric Setup Guide


3. 🔄 Microsoft Fabric - Open Mirroring

Use the native Fabric mirroring capability to replicate Business Central data continuously.

Best for:

  • Near real-time data replication
  • Minimal configuration requirements
  • Fully managed solution by Microsoft Fabric
  • Organizations wanting the simplest setup

Key Features:

  • Native Fabric integration
  • Automatic continuous replication
  • No notebook required

📖 Read the Fabric Open Mirroring Guide


Getting Started

Quick Start Guide

  1. Choose your export option based on your infrastructure and requirements
  2. Install the Business Central extension from the releases
  3. Configure Azure/Fabric resources according to your chosen option
  4. Set up authentication using Service Principal
  5. Configure export settings in Business Central
  6. Schedule exports using Job Queue entries
  7. Set up data consolidation (for ADLS or Fabric with notebooks)

Prerequisites

  • Dynamics 365 Business Central (Online or On-Premises)
  • Azure subscription (for ADLS and Fabric options)
  • Appropriate permissions in Business Central and Azure/Fabric

Documentation Pages

Setup and Configuration

Operations

Advanced Topics

Reference

  • PTE App vs AppSource App - Differences between the PTE and Marketplace versions (object number mapping)
  • FAQs - Frequently asked questions and troubleshooting
  • Changelog - Release notes and version history
  • Permission Sets - Understanding BC permissions for bc2adls

Resources

Video Resources

Community and Support


Next Steps

  1. Choose your best option (Fabric is the newest architecture)
  2. Follow the appropriate setup guide
  3. Install and configure the Business Central extension
  4. Test with a few tables before full deployment
  5. Join the community and share your experience!

This project is maintained by the community. Special thanks to the original creators Soumya Dutta and Henri Schulte.

Clone this wiki locally