You plan to process the following three datasets by using Fabric: Dataset1: This dataset will be added to Fabric and will have a unique primary key between the source and the destination. The unique…

This question tests knowledge of Microsoft Fabric item types (Dataflow Gen2, Notebook, Warehouse) and which built-in capabilities best match specific data loading and transformation requirements, with emphasis on minimizing development effort.

Design and implement data ingestion and transformation

Question

You plan to process the following three datasets by using Fabric: Dataset1: This dataset will be added to Fabric and will have a unique primary key between the source and the destination. The unique primary key will be an integer and will start from 1 and have an increment of 1. Dataset2: This dataset contains semi-structured data that uses bulk data transfer. The dataset must be loaded in one process between the source and the destination. The data transformation process will include the use of custom visuals to understand and work with the dataset in development mode. Dataset3: This dataset is in a lakehouse. The data will be bulk loaded. The data transformation process will include row-based windowing functions during the loading process. You need to identify which type of item to use for the datasets. The solution must minimize development effort and use built-in functionality, when possible. What should you identify for each dataset? To answer, select the appropriate options in the answer area. NOTE: Each correct selection is worth one point.

Explanation

Approach. Dataset1 maps to Dataflow Gen2: its Power Query engine includes a built-in 'Add Index Column' feature (Starting Index: 1, Increment: 1) that generates a unique integer primary key with zero custom code - exactly matching the auto-increment requirement while minimizing development effort. Dataset2 maps to Notebook: Notebooks natively handle semi-structured data (JSON, Parquet, etc.) via PySpark, support atomic bulk loads in a single execution, and allow custom Python visualization libraries (matplotlib, plotly, seaborn) to explore and understand data during interactive development - matching the 'custom visuals in development mode' requirement. Dataset3 maps to Notebook: since the data already lives in a Lakehouse, PySpark Notebooks access it directly without extra connectors, support large bulk loads via Spark parallelism, and have built-in support for row-based windowing functions through the PySpark Window API (equivalent to SQL's ROWS BETWEEN framing), making this the lowest-effort built-in solution.

Concept tested. Microsoft Fabric item selection - matching Dataflow Gen2 (built-in index column for auto-increment keys), Notebook (semi-structured bulk loading with visualization and PySpark Window functions), and Lakehouse/Warehouse trade-offs based on data location, bulk load patterns, and transformation complexity.

Reference. Microsoft Learn - Microsoft Fabric: Choose a data store | DP-700 Exam: Implement and Manage an Analytics Solution using Microsoft Fabric

Topics

#Fabric Data Integration#Fabric Data Warehouse#Dataflow Gen2#Spark Notebooks

Community Discussion

No community discussion yet for this question.

Full DP-700 Practice