DP-700 Ingest and Transform Data Practice Test 3 — 15 Questions

Question 1

Refer to the exhibit. You are configuring a Dataflow Gen2 to move data. Based on the JSON configuration, what happens if the incoming data contains an extra column not defined in the destination schema?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

The ingestion process fails.. With schema validation set to 'strict', Dataflow Gen2 enforces a rigid mapping between source and sink schemas. If the incoming data does not match the defined schema exactly, the ingestion process will fail. This configuration is critical for maintaining high data quality and preventing 'schema drift' in downstream analytical tables, ensuring that automated processes do not break due to unexpected changes in the upstream data structure.

Answer

The extra column is automatically dropped.

Answer

The extra column is added to the destination schema.

Answer

The extra column is logged as a warning.

Question 2

You are using a Copy Activity in Microsoft Fabric Data Factory. You need to ensure that the data is encrypted both in transit and at rest. Which setting should you verify?

Accepted Answer

Check the 'Use Encrypted Connection' property in the Linked Service. Microsoft Fabric automatically handles encryption at rest for all data stored in OneLake. For data in transit, you must ensure that your connections use encrypted protocols like HTTPS or TLS. By default, Fabric connectors are configured to use secure channels, but verifying these settings ensures compliance with corporate security mandates regarding data in transit protection during ingestion from external sources into your Fabric environment.

Answer

Enable 'Force TLS 1.2' on the source server

Answer

Manually encrypt the files using a PGP key

Answer

Configure an Azure Key Vault secret for the data

Question 3

You are performing a large-scale data migration into a Fabric Lakehouse using a Spark Notebook. You notice that the job is failing with an 'Out of Memory' error. What is the most effective way to address this while utilizing PySpark?

Accepted Answer

Use df.repartition() to redistribute data. Out of Memory (OOM) errors in Spark often result from unbalanced partitions or attempting to load too much data into a single executor. By increasing the number of partitions (repartitioning) or adjusting the cluster configuration, you redistribute the workload across the Spark executors. This is a fundamental skill for data engineers to master, as it allows for processing massive datasets efficiently without requiring constant hardware upgrades or job failures.

Answer

Increase the cluster node count significantly

Answer

Convert the data to JSON format before loading

Answer

Disable the Spark broadcast join optimization

Question 4

You need to automate the ingestion of files from an Azure Blob Storage container into a Fabric Lakehouse. Which THREE components are required to build this automated pipeline? (Choose three)

Accepted Answer

A Trigger to define the execution schedule. An automated pipeline requires a trigger to start the process, a Linked Service to connect to the source storage, and the Copy Activity to perform the actual movement. These three components form the core backbone of any Fabric data pipeline. Mastery of these components is essential for orchestrating repeatable and scalable data workflows that can handle modern data engineering requirements across complex enterprise environments.

Answer

A Managed VNet configuration

Answer

A SQL Server Integration Services (SSIS) project

Question 5

You are configuring a Fabric pipeline to perform incremental loads from a source system. Which approach is considered the best practice to track changes efficiently?

Accepted Answer

Use a watermark column for filtering. Using a watermarking column (such as a 'LastModified' timestamp or an incrementing ID) is the standard industry practice for incremental loading. By querying only rows with a value greater than the previous high-water mark, you significantly reduce the amount of data transferred and processed. This approach minimizes source system load and execution time, making it essential for building performant, cost-effective pipelines that handle large-scale data updates over time.

Answer

Truncate and reload the entire dataset every time

Answer

Compare every row using a hash function

Answer

Delete all data older than one month

Question 6

Refer to the exhibit. Given the retry policy configuration, what happens if the source SQL database is temporarily unavailable during the first attempt?

Accepted Answer

The pipeline retries 3 times with 60-second intervals.. The retry policy defined in the activity configuration instructs the Fabric pipeline to automatically attempt the copy operation again after the specified interval. This is a critical feature for building resilient pipelines that can gracefully handle transient network errors or brief service outages without failing the entire job. It ensures higher pipeline success rates and reduces the need for manual monitoring and intervention in complex, real-world cloud environments.

Answer

The pipeline fails immediately.

Answer

The pipeline retries indefinitely.

Answer

The pipeline executes the next activity immediately.

Question 7

You are designing a data transformation layer using Spark in Microsoft Fabric. Which THREE of the following are primary benefits of using Delta Lake format over traditional Parquet files? (Choose three)

Accepted Answer

ACID transaction support. Delta Lake adds a transaction log (the Delta Log) to the Parquet file format. This enables ACID transactions, preventing data corruption and partial writes. It also supports schema evolution, allowing the data structure to change over time without breaking downstream applications. Finally, 'Time Travel' allows users to query previous versions of data, which is invaluable for auditing, debugging, and reproducing analytical results. These features are critical for modern, reliable data platforms.

Answer

Smaller storage footprint than Parquet

Answer

Faster cold-start execution for Spark clusters

Question 8

You are monitoring a Fabric pipeline. You need to identify which activity failed within the pipeline execution. Which tool should you use?

Accepted Answer

Fabric Pipeline Monitoring view. The 'Pipeline Monitoring' view in the Fabric workspace provides a centralized dashboard to view the status, start time, end time, and duration of all pipeline activities. Clicking on a specific pipeline run allows you to drill down into individual activities to inspect error logs, which is essential for diagnosing failures. This visibility is crucial for maintaining operational excellence and ensuring that data pipelines meet their SLAs in a production environment.

Answer

Power BI Desktop

Answer

Azure Resource Manager (ARM) templates

Answer

OneLake file explorer

Question 9

Refer to the exhibit. You are loading this file into a Lakehouse. You notice that the column headers contain special characters, and the data is failing to load correctly. What should you do?

Accepted Answer

Disable the 'first_row_as_header' setting and use custom schema mapping. When dealing with CSV files, column headers must be compatible with the destination system's naming conventions. Often, special characters in source headers are invalid for column names in the target Lakehouse tables. By sanitizing headers or choosing to skip them and using custom column names, you ensure the load completes successfully. This is a common real-world challenge when integrating data from disparate source systems into a unified analytical Lakehouse.

Answer

Change the file encoding to ASCII

Answer

Increase the pipeline retry count

Answer

Convert the CSV to binary format

Question 10

You are building a pipeline and need to ensure that a downstream activity only runs if a previous activity succeeds. Which pipeline feature should you use?

Accepted Answer

An 'On Success' connector. The 'On Success' constraint on the pipeline connector links two activities in a dependency chain. This ensures that the second activity starts only if the first one finishes without errors. This is fundamental for building reliable, production-grade workflows where the order of operations and the integrity of the data processing chain are paramount for successful data delivery to the end users.

Answer

A ForEach activity

Answer

A Wait activity

Answer

A Filter activity

Question 11

You need to access data from an external Azure Data Lake Storage (ADLS) Gen2 account in a Fabric Lakehouse without duplicating the physical data. Which feature should you use?

Accepted Answer

Shortcut. Shortcuts provide a way to reference data stored in external locations without the need for data movement or replication. This approach reduces storage costs and ensures that the Fabric environment always sees the most current version of the source data. It is a fundamental concept in OneLake for maintaining a single source of truth across different cloud providers.

Answer

Dataflow Gen2

Answer

Copy Activity

Answer

Mirroring

Question 12

You are building a Dataflow Gen2 to ingest a 50 GB CSV file from an Azure Storage account into a Fabric Lakehouse. You want to ensure the ingestion uses the 'Fast Copy' feature. What must you do?

Accepted Answer

Ensure no transformations are used in the Power Query steps.. Fast Copy in Dataflows Gen2 allows for high-throughput data movement by bypassing the Power Query mashup engine for the data movement phase. It requires that the source connector supports it and that no complex transformations are applied within the Power Query steps that would force the engine to process rows individually.

Answer

Enable the 'Allow Schema Evolution' setting in the destination.

Answer

Use a Managed Private Endpoint for the connection.

Answer

Set the destination to 'Append' instead of 'Replace'.

Question 13

You are optimizing a Spark Notebook in Fabric that processes large Parquet files. Which TWO actions will improve the performance of data transformations?

Accepted Answer

Enable V-Order on the Delta tables.. Optimizing Spark performance in Fabric involves managing how data is distributed across the cluster and how it is physically stored. Using V-Order ensures that files are optimized for Fabric's compute engines, while proper partitioning prevents data skew and allows for efficient parallel processing across the available Spark nodes in the workspace.

Answer

Increase the 'isSequential' property to True.

Answer

Convert all files to the Avro format for faster reads.

Answer

Disable the Spark 'High Concurrency' mode.

Question 14

Your organization wants to implement near real-time data ingestion from an Azure SQL Database into Fabric for reporting. You must minimize the impact on the source database's performance and avoid manual pipeline scheduling. Which solution should you implement?

Accepted Answer

Fabric Mirroring for Azure SQL Database.. Mirroring in Fabric provides a low-latency, continuous synchronization path from supported databases. It uses the source database's transaction log to capture changes (CDC), which is much more efficient than traditional polling and does not require the user to manage complex schedules or orchestration pipelines to keep data synchronized.

Answer

A Pipeline with a Copy Activity using a 'Tumbling Window' trigger.

Answer

A Dataflow Gen2 with incremental refresh enabled.

Answer

A Spark Notebook using the 'JDBC' connector in a loop.

Question 15

Refer to the exhibit. You are reviewing a JSON snippet from a Fabric Data Factory pipeline. What is the result of executing this activity if the target table already exists?

Accepted Answer

The existing data in the table will be replaced by the source data.. The JSON configuration specifies a Copy Activity that pulls data from an Amazon S3 source and writes it to a Lakehouse table. The tableAction property is set to Overwrite, which instructs the engine to drop the existing data in the target table and replace it with the new incoming dataset. This is a common pattern for full-load ingestion.

Answer

The activity will append new rows to the existing table data.

Answer

The pipeline will fail because the table already exists.

Answer

A new table with a timestamp suffix will be created.