Design and Implement Data Pipelines DP-750 Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   When designing a multi-task data pipeline, what is the primary purpose of defining a clear order of operations?

   • ATo ensure each task has access to the output of its upstream dependencies
   • BTo allow tasks to run in parallel without any coordination
   • CTo eliminate the need for error handling in the pipeline
   • DTo minimize the number of tasks required in the pipeline
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   In a data pipeline, which sequence of operations is most appropriate for a typical batch ETL workload?

   • AWrite results → Load raw data → Transform data → Validate data
   • BTransform data → Load raw data → Write results → Validate data
   • CLoad raw data → Transform data → Validate data → Write results
   • DValidate data → Load raw data → Transform data → Write results
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A team is building a Lakeflow job with three notebook tasks: extract, transform, and load. The extract task pulls data from a source system, the transform task cleans and aggregates the data, and the load task writes the data to a Delta table. The team wants to ensure that the transform task does not start until the extract task has successfully completed, and the load task does not start until the transform task has successfully completed. What should they configure?

   • ASet the load task to depend on the extract task, and set the transform task to depend on the load task.
   • BSet the transform task to depend on the extract task, and set the load task to depend on the transform task.
   • CCombine all three tasks into a single notebook and run them sequentially in cells.
   • DSet all three tasks to run in parallel and use a shared variable to coordinate the order.
   Reveal Solution Discussion Select an answer first
5. 9application · medium

   A data engineering team maintains a batch pipeline that ingests raw clickstream files, performs a series of transformations, and writes aggregated results to a Delta table. The team wants to reduce development time and make the pipeline easier to maintain by reusing the same transformation logic across multiple environments. They do not require custom Python libraries or interactive debugging. Which approach should they choose?

   • AImplement the pipeline as a series of SQL views in a single notebook and rely on the notebook's scheduled run.
   • BCreate a Lakeflow Spark Declarative Pipeline that defines the transformations as declarative steps.
   • CBuild a notebook-based pipeline with separate cells for each transformation and schedule it as a job.
   • DUse a single notebook that contains all transformation logic and run it manually whenever new data arrives.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A data pipeline processes daily sales files. Occasionally, a file is missing or arrives late. The team wants the pipeline to retry transient failures, but if a file is truly missing after a certain number of attempts, the pipeline should fail and alert the operations team. They are using a Lakeflow job with notebook tasks. What should they configure?

   • ASet a retry policy on the notebook task that retries up to 3 times with a 5-minute wait, and enable the 'fail on retry exhaustion' option so the job fails after the retries are exhausted.
   • BSet the notebook task to run continuously and never fail, so it keeps trying until the file appears.
   • CUse a Delta Live Tables pipeline with a 'skip missing files' option and no retry policy.
   • DConfigure the notebook to catch all exceptions and write a success status even if the file is missing.
   Reveal Solution Discussion Select an answer first