BenchMark Quizzes by MDFT Pro

Schedule the Population of Medallion Layers

You work for MDFT Pro, a well-known training agency that provides certification preparation courses. Mark, the Data Engineer, is implementing a medallion architecture in Microsoft Fabric to organize the company’s student enrollment and course completion data.

The architecture uses three layers: bronze (raw data from various sources), silver (cleansed and validated data), and gold (aggregated data for reporting). Each medallion is stored in a separate lakehouse, and each layer requires different transformations that must be populated in sequence to ensure data quality. The bronze lakehouse ingests raw data, the silver lakehouse cleanses and standardizes it, and the gold lakehouse creates aggregated views for business intelligence reports.

What should Mark use to schedule the population of the medallion layers across the separate lakehouses?

Choose the correct answer from the options below.

Explanations for each answer:

A data pipeline that calls other data pipelines is correct. A master data pipeline that invokes other pipelines (using Invoke Pipeline) is the best approach when each medallion (bronze, silver, gold) is stored in a separate lakehouse. It lets you orchestrate cross-lakehouse pipelines in the correct sequence, manage dependencies and retries, and centralize error handling across the separate lakehouses.
A notebook is incorrect. A notebook can perform transformations within a single lakehouse, but it lacks orchestration features needed to coordinate multiple pipelines across different lakehouses. Notebooks are best for individual transformation tasks, not cross-lakehouse sequencing.
An Apache Spark job is incorrect. A Spark job can run transformations, but it doesn't provide orchestration across multiple lakehouses. You would still need an orchestration layer to ensure the jobs run in sequence across bronze, silver, and gold lakehouses and to handle failures and retries.
Multiple data pipelines running in parallel is incorrect. Scheduling multiple independent pipelines (one per lakehouse) without a coordinating pipeline or orchestration will not guarantee correct sequencing across the separate lakehouses. Without centralized coordination you risk data consistency issues if downstream lakehouse pipelines start before upstream ones finish.

Learn more about the Invoke Pipeline activity:

Invoke Pipeline Activity

Practice Exam

Schedule the Population of Medallion Layers

Discuss this question on social media: