Dataflows Gen2
Use Power Query Online to connect, clean, transform, and land data in supported Fabric destinations.
DP-700: understand when low-code transformation and reusable ETL logic are appropriate.
DP-700 · DP-600 shared topic
Dataflows, Pipelines, and Ingestion Choices
A practical guide to choosing between low-code transformations, orchestration, and direct data movement in Microsoft Fabric.
The core distinction
Think of a Dataflow Gen2 as the worker that transforms data and a pipeline as the manager that coordinates the whole process. Copy Data is the direct transport option when the data should move with little or no transformation.
Pipelines
Use pipelines to orchestrate activities, schedule runs, pass parameters, and monitor ingestion workflows.
DP-700 and DP-600: distinguish transformation work from orchestration and sequencing.
Copy Data
Use Copy Data when the primary requirement is moving data as-is between a source and destination.
DP-700: compare Copy Data with Dataflows Gen2 and notebooks based on transformation needs.
Choosing the right tool
Dataflows transform data, pipelines coordinate work, and notebooks or SQL handle code-heavy transformations.
Both exams test architecture decisions, not just definitions.
Dataflows Gen2 versus pipelines
| Capability | Dataflows Gen2 | Data pipeline |
|---|---|---|
| Primary role | Transform and clean data | Orchestrate and schedule work |
| Interface | Power Query Online | Pipeline canvas |
| Typical logic | Joins, filters, pivots, and shaping | Dependencies, loops, parameters, and triggers |
| Good fit | Reusable low-code ETL | Multi-step ingestion workflows |
Copy Data versus Dataflows Gen2
| Choose Copy Data when… | Choose Dataflows Gen2 when… |
|---|---|
| The source data should move as-is. | The data needs cleaning or filtering during ingestion. |
| Transformation will happen later in a notebook or SQL activity. | You want a visual Power Query workflow. |
| A high-throughput direct movement is the priority. | You need reusable transformations across destinations. |
Exam decision checklist
- • Choose Dataflows Gen2 when visual, reusable transformations are the main requirement.
- • Choose a pipeline when scheduling, dependencies, loops, parameters, or monitoring are central.
- • Choose Copy Data when moving data as-is is more important than transforming it during ingestion.
- • Choose a notebook or SQL activity when the transformation requires code or advanced logic.