DWA (Data Warehouse Automation) – v2 Release Notes
Version: 2025.6.1
Release Date: June 16, 2025
This release introduces comprehensive support for Microsoft Fabric, enabling the full automation suite to run on Microsoft's latest analytics platform. It also includes a significant new OpenLineage integration for enhanced data traceability and a new JWT authentication method for improved, unified security across APIs.
SQL Execution Engine
A summary of all changes made to the execution engine.
🚀 Features
- Added direct support for Microsoft Fabric workloads, including connectivity using Service Principal authentication for secure, automated access.
- Introduced OpenLineage integration to capture start and end events for all executions, providing robust job monitoring and data lineage tracking.
🐛 Bug Fixes
- Improved error reporting by removing a control mechanism that could hide underlying database errors during execution.
🔧 Maintenance
- Standardised the naming of the environment configuration file from
mq.settings.envtomq.env.
SQL Generator API
A summary of all changes made to the SQL generation logic and templates.
🚀 Features
- Implemented full SQL generation compatibility for Microsoft Fabric, including a new
fabricdialect and updated logic for all core data loading patterns. - Deployed a new, secure JWT OAUTH2.0 authentication method, allowing for a coordinated and unified user base across all APIs.
- Added new "Order by" parameters to the AttributeLoader, Publisher, and RelationLoader patterns for more control over data processing output.
🐛 Bug Fixes
- Corrected the generation of full ADLS URLs for
OPENROWSETcommands. - Fixed an issue with incorrect characters appearing in logging columns for stage loads.
- Corrected a minor naming issue in generated statistics.
- Modified Snowflake data type conversions to be more tolerant of empty JSON strings, preventing errors in the publisher.
🔧 Maintenance
- All SQL patterns were refactored to use temporary tables (
#temp), improving reliability and alignment with SQL Server best practices. - The relationship loading pattern was updated to align more closely with the behaviour of Synapse and SQL Server.
- Refactored internal context handling within SQL templates for cleaner and more maintainable code.