Which component should be used to minimize refresh changes to the Orders table when refreshing from OneLake?

Which component should be used to minimize refresh changes to the Orders table when refreshing from OneLake?

• A. Azure Data Factory pipeline with dataflow
• B. Power BI dataset refresh
• C. SQL job with truncate and reload
• D. Azure Synapse pipeline

Answer: (A)

Incremental loading with an upsert into the target is what this question tests. When refreshing from OneLake, you want to apply only the changes (new and updated orders) rather than rewriting the entire Orders table. A pipeline that uses a mapping data flow in Azure Data Factory can read just the rows that have changed since the last load (using a watermark or a last-modified indicator) and then merge those rows into the destination with an upsert operation. This preserves existing data, minimizes the amount of data written, and avoids the overhead and risks of truncating and reloading the whole table. Power BI dataset refresh focuses on updating BI visuals and does not modify the source Orders table to minimize refresh changes. A SQL job that truncates and reloads would replace the entire table, maximizing changes rather than minimizing them. Azure Synapse pipelines can also do incremental loads, but the mapping data flow approach in Data Factory is the most direct way to implement delta merges from OneLake with built-in upsert behavior, aligning exactly with minimizing refresh changes.

Incremental loading with an upsert into the target is what this question tests. When refreshing from OneLake, you want to apply only the changes (new and updated orders) rather than rewriting the entire Orders table. A pipeline that uses a mapping data flow in Azure Data Factory can read just the rows that have changed since the last load (using a watermark or a last-modified indicator) and then merge those rows into the destination with an upsert operation. This preserves existing data, minimizes the amount of data written, and avoids the overhead and risks of truncating and reloading the whole table.

Power BI dataset refresh focuses on updating BI visuals and does not modify the source Orders table to minimize refresh changes. A SQL job that truncates and reloads would replace the entire table, maximizing changes rather than minimizing them. Azure Synapse pipelines can also do incremental loads, but the mapping data flow approach in Data Factory is the most direct way to implement delta merges from OneLake with built-in upsert behavior, aligning exactly with minimizing refresh changes.