A pipeline currently uses a Dataflow Gen2 to move a straightforward, unmodified extract of daily exhibit-ticket sales from a source database into a Warehouse staging table — no transformation logic is applied at any step. The load is one of the slowest steps in the nightly schedule. What change is most likely to improve its performance, without changing what data ends up in the staging table?
Select an answer to reveal the explanation.
Short Explanation
Dataflow Gen2 is built for shaping and transforming data with a full query engine underneath, which is overhead you're paying for even when there's nothing to shape. A Copy activity, by contrast, is a lean, purpose-built mover for exactly this case — data going from A to B unchanged. When there's genuinely no transformation happening, swapping to Copy plays to its strength and skips the engine you don't need.
Full Explanation
Choosing the right orchestration tool for the job is itself a performance decision: Dataflow Gen2 runs a full Power Query mash-up engine designed for transformation and shaping, and that engine carries overhead even when a mash-up does nothing but pass data through unchanged, whereas a pipeline's Copy activity is purpose-built for efficient bulk data movement with none of that transformation overhead — so for a genuinely untransformed extract-and-load, Copy is the better-performing choice. Renaming columns doesn't unlock a different, faster execution path inside Dataflow Gen2; it's still routed through the same mash-up engine, so it wouldn't meaningfully change throughput. Isolating the Dataflow Gen2 in its own workspace changes organizational and access boundaries, not the per-item engine overhead that's actually the bottleneck here. Converting the staging table into a shortcut over the live source database changes the architecture entirely — it means never materializing a stable copy for the night's load, which isn't the same problem as speeding up a copy, and a Warehouse table typically isn't the shortcut target being described anyway. A caveat: this trade-off only holds while the step truly has zero transformation logic — the moment real shaping, joins, or type conversions are needed, Dataflow Gen2's capabilities may outweigh the raw-copy speed advantage of a Copy activity. A concrete check: compare the activity's duration and data-read/write throughput in pipeline monitoring before and after the swap for the same source and volume.