Power BI: Manage Dataflow Gen1 Versions

Key insights

• Dataflow Gen1 is now a legacy service: Microsoft stopped new feature development and will only apply critical fixes.
  Teams should plan migration because Gen1 will not gain modern scaling or new integrations.
• There are multiple versions and variants to know: Dataflow Gen2 is the modern runtime and Dataflow Gen2 CI/CD adds deployment and version control.
  The ecosystem also distinguishes Internal vs External dataflows and Standard vs Analytical dataflows for different sharing and modeling needs.
• Licensing and compute differ by version: Gen1 works with Power BI Pro/PPU and does not require Fabric capacity, while Gen2 runs on Fabric capacity and may change cost and governance choices.
  Expect better performance but plan capacity and licensing when you migrate.
• How Gen1 works in practice: authors use the online Power Query editor to transform data, then store results in targets like ADLS Gen2 or Dataverse.
  Dataflows support scheduled refreshes and serve datasets, reports, or dashboards via the Dataflows connector.
• Benefits and limits: Gen1 offers familiar Power Query authoring, incremental refresh, and Pro/PPU cost options.
  Its limits include fewer destination types, slower performance at scale, no pipeline integration, and higher refresh costs for large workloads.
• Migration guidance: move critical ETL to Gen2 or Gen2 CI/CD for scalability, Lakehouse/Warehouse destinations, and pipeline support.
  Keep Power Query logic where possible, test refresh performance, and update licensing and deployment plans before cutting over.

Video Overview and Key Takeaway

Reza Rad (RADACAD) [MVP] published a technical video that explains the evolving landscape of Power BI Dataflow Gen1 and its move to legacy status. The video makes clear that Microsoft intends to develop around Dataflow Gen2 and a CI/CD variant built for Microsoft Fabric. Consequently, teams that rely on Gen1 should start planning migration paths while recognizing some short-term benefits of staying with Gen1 for small-scale scenarios.

Moreover, the presentation walks viewers through the practical differences between internal and external dataflows as well as standard and analytical variants. It also highlights how Gen1 remains functional but will receive only critical fixes going forward. Therefore, organizations must weigh cost, performance, and governance when deciding the next steps.

What Dataflow Gen1 Is and Its Multiple Versions

According to the video, Dataflow Gen1 is the original cloud-based Power Query ETL layer inside Power BI that outputs to ADLS Gen2 or Dataverse. It supports familiar authoring patterns, incremental refresh, and connectors that many teams already use, which explains its long adoption. However, Gen1 split into contextual variants: internal versus external and standard versus analytical, which created management complexity for large deployments.

Reza Rad further clarifies that the ecosystem now centers on three main versions: Dataflow Gen1 (legacy), Dataflow Gen2 (Fabric-powered runtime), and Dataflow Gen2 CI/CD for deployment pipelines. While Gen1 works with Pro and Premium Per User (PPU) licensing, Gen2 typically requires Fabric capacity and adds destinations like Lakehouses and Warehouses. Thus, the shift is not only technical but also licensing- and cost-driven.

Why Gen1 Is Being Phased Out and the Trade-offs

The video stresses that Microsoft labeled Gen1 as legacy after eight years of development, shifting investments to Gen2 for scalability and integration with data engineering patterns. On one hand, Gen1 remains attractive because of Pro/PPU affordability and the familiar Power Query authoring experience. On the other hand, organizations face trade-offs: Gen1 can be cheaper for light workloads but often performs worse and lacks pipeline integration, while Gen2 delivers higher scale at the expense of Fabric capacity costs.

Furthermore, the presenter draws attention to destinations and features as a key trade-off. Gen1 primarily targets ADLS Gen2 and has limited destination support, whereas Gen2 supports multiple storage and compute destinations that fit enterprise patterns. Consequently, teams must balance immediate budget constraints against long-term maintainability, governance, and performance requirements.

Migration Challenges and Technical Considerations

The video outlines several migration challenges that organizations commonly face. First, compatibility differences in destination targets and runtime behavior may require re-authoring or tuning of complex transformations, which increases project effort. Second, the move to Gen2 often entails shifting to Fabric capacity and rethinking CI/CD, which introduces new tooling and governance needs for teams that previously relied on lightweight, self-service workflows.

Additionally, Reza Rad points out performance and monitoring considerations. Gen2 improves high-scale compute and pipeline orchestration, but it can also introduce cost complexity because capacity planning and scaling affect costs in different ways. Therefore, teams must assess refresh patterns, query performance, and downstream model dependencies before committing to a migration path.

Practical Guidance for Teams and Next Steps

Reza Rad advises organizations to inventory existing Gen1 assets and categorize them by business criticality, refresh frequency, and complexity before migrating. In particular, low-volume, low-cost use cases may remain on Gen1 for a while, while mission-critical, high-volume pipelines should move to Gen2 and adopt CI/CD practices. This staged approach reduces risk and spreads migration cost over time.

Finally, the presenter recommends close collaboration between data engineers, BI authors, and governance teams to set policies for destinations, capacity, and deployment processes. In short, the transition requires balancing cost, performance, and operational maturity, and it benefits from careful planning and pilot migrations that validate assumptions before a full rollout. Overall, the video provides a practical roadmap that helps teams weigh trade-offs and prepare for a Fabric-centric future.

• Author: Reza Rad (RADACAD) [MVP]
• Main message: Dataflow Gen1 is legacy; plan for Gen2 and CI/CD on Fabric.

Keywords

Power BI Dataflow Gen1 multiple versions, Power BI dataflow versioning, Dataflow Gen1 version control, Manage Power BI dataflow versions, Restore Power BI Dataflow Gen1, Power BI dataflow version history, Compare Power BI dataflow versions, Troubleshoot Dataflow Gen1 version issues