Google Cloud Professional Data Engineer Practice Exam

What the Google Cloud Professional Data Engineer exam covers

• Designing Data Processing Systems133 questions
• Ingesting and Processing the Data159 questions
• Storing the Data131 questions
• Preparing and Using Data for Analysis120 questions
• Maintaining and Automating Data Workloads153 questions

Free Google Cloud Professional Data Engineer sample questions

A sample of 10 questions with answers and explanations. Sign up free to practice all 696.

1. Question 1Designing Data Processing Systems

   Which service is a serverless, petabyte-scale data warehouse for analytics?

   • ABigQueryCorrect
   • BCloud DNS
   • CMemorystore
   • DCloud SQL
   ✓ Correct answer: A

   BigQuery is Google Cloud's serverless, petabyte-scale data warehouse specifically designed for analytics workloads. It provides fully managed SQL query capabilities with automatic scaling and no infrastructure management. BigQuery uses columnar storage and parallel execution to achieve fast analytical queries on massive datasets.

   Why the other options are wrong

   • BCloud DNS is incorrect because Cloud DNS is a domain name system service for DNS management, not a data warehouse.
   • CMemorystore is incorrect because Memorystore is an in-memory cache service optimized for low-latency operational access, not analytical workloads.
   • DCloud SQL is incorrect because Cloud SQL is a managed relational database for operational use cases, not designed for petabyte-scale analytics.
2. Question 2Designing Data Processing SystemsSelect all that apply

   Which TWO factors most influence choosing a data store? (Choose TWO)

   • Aaccess pattern and latency/throughput needsCorrect
   • BWhether the data is structured, semi-structured, or unstructuredCorrect
   • CThe logo color of the team
   • DThe number of office locations
   ✓ Correct answer: A, B

   The two most critical factors when choosing a data store are: (1) access pattern determines whether you need operational/transactional access or analytical batch queries, and (2) latency/throughput requirements define acceptable response times and throughput capacity. Additionally, data structure (relational, columnar, key-value, graph) significantly influences which technology best fits.

   Why the other options are wrong

   • CThe logo color of the team is incorrect because The logo color of the team is irrelevant to data store selection.
   • DThe number of office locations is incorrect because The number of office locations does not influence data store choice.
3. Question 3Designing Data Processing SystemsSelect all that apply

   Which TWO factors guide choosing a Google Cloud data store? (Choose TWO)

   • AThe team's logo color
   • BThe office location
   • Caccess pattern (analytical vs operational) and latency/throughputCorrect
   • DStructured vs unstructured dataCorrect
   ✓ Correct answer: C, D

   Choosing a data store depends on whether you need analytical scans (BigQuery) or operational transactions (Cloud SQL/Spanner), combined with your latency/throughput SLAs. Data structure determines technology fit: structured data suits relational stores, while unstructured benefits from object storage. These two factors together define the optimal platform.

   Why the other options are wrong

   • AThe team's logo color is incorrect because The team's logo color is irrelevant to technical architecture decisions.
   • BThe office location is incorrect because Office location does not influence data store selection.
4. Question 4Designing Data Processing Systems

   Which is best for massive analytical SQL over petabytes?

   • ACloud SQL
   • BMemorystore
   • CBigQueryCorrect
   • DCloud DNS
   ✓ Correct answer: C

   BigQuery is purpose-built for massive analytical SQL queries over petabyte-scale datasets. Its columnar architecture, automatic parallelization, and serverless scaling enable fast SQL analytics on huge volumes without infrastructure management.

   Why the other options are wrong

   • ACloud SQL is incorrect because Cloud SQL cannot efficiently handle petabyte-scale analytical workloads.
   • BMemorystore is incorrect because Memorystore is for caching, not analytics.
   • DCloud DNS is incorrect because Cloud DNS manages DNS, not data analytics.
5. Question 5Ingesting and Processing the Data

   Which serverless service runs both batch and streaming data pipelines based on Apache Beam?

   • ACloud DNS
   • BMemorystore
   • CDataflowCorrect
   • DCloud Functions
   ✓ Correct answer: C

   Dataflow is Google Cloud's fully managed streaming and batch processing service built on Apache Beam. It abstracts away infrastructure concerns, automatically scaling workers to handle variable load while enabling unified batch/streaming pipelines with exactly-once processing semantics.

   Why the other options are wrong

   • ACloud DNS is incorrect because Cloud DNS manages DNS records, not data pipelines.
   • BMemorystore is incorrect because Memorystore is an in-memory cache, not a pipeline engine.
   • DCloud Functions is incorrect because Cloud Functions is for event-driven functions, not batch/streaming pipelines.
6. Question 6Ingesting and Processing the DataSelect all that apply

   Which TWO services form a typical real-time streaming pipeline on Google Cloud? (Choose TWO)

   • Adataflow for stream processing/transformationCorrect
   • BCloud DNS for processing
   • CMemorystore for stream transformation
   • Dpub/Sub for ingestionCorrect
   ✓ Correct answer: A, D

   A typical real-time streaming pipeline uses Pub/Sub as the message queue for ingesting high-volume events, and Dataflow to consume and transform those messages in real-time. This combination provides scalable, serverless event-driven architecture with automatic scaling.

   Why the other options are wrong

   • BCloud DNS for processing is incorrect because Cloud DNS is a DNS service, not a stream processor.
   • CMemorystore for stream transformation is incorrect because Memorystore is for caching, not stream transformation.
7. Question 7Ingesting and Processing the Data

   How does Pub/Sub help when a subscriber repeatedly fails to process certain messages?

   • ASilently drop all messages
   • BConfigure a dead-letter topic to capture messages that exceed max delivery attemptsCorrect
   • CConvert messages to DNS records
   • DStop the entire subscription permanently
   ✓ Correct answer: B

   Pub/Sub's dead-letter topic feature automatically routes messages that exceed the maximum delivery attempts threshold to a separate topic. This enables visibility into problematic messages, debugging of subscriber failures, and recovery workflows without blocking the main pipeline.

   Why the other options are wrong

   • ASilently drop all messages is incorrect because Silently dropping messages loses data and prevents troubleshooting.
   • CConvert messages to DNS records is incorrect because Converting messages to DNS records is not a valid Pub/Sub feature.
   • DStop the entire subscription permanently is incorrect because Stopping the subscription prevents processing of all messages, not just problematic ones.
8. Question 8Ingesting and Processing the Data

   Which file formats are efficient columnar formats often used in data pipelines?

   • ADOCX and PDF
   • BMP3 and WAV
   • Cparquet and ORC (columnar); Avro is row-based but schema-richCorrect
   • DPNG and JPEG
   ✓ Correct answer: C

   Parquet and ORC are columnar formats optimized for analytical queries, offering excellent compression and predicate pushdown for big data pipelines. Avro is row-oriented but includes schema evolution capabilities, making these three formats industry standards for data pipeline efficiency and schema flexibility.

   Why the other options are wrong

   • ADOCX and PDF is incorrect because DOCX and PDF are document formats, not suitable for data pipelines.
   • BMP3 and WAV is incorrect because MP3 and WAV are audio formats, not data pipeline formats.
   • DPNG and JPEG is incorrect because PNG and JPEG are image formats, not data exchange formats.
9. Question 9Ingesting and Processing the DataSelect all that apply

   Which TWO build a streaming analytics pipeline on Google Cloud? (Choose TWO)

   • ACloud DNS for processing
   • BMemorystore for transformation
   • Cpub/Sub for ingestionCorrect
   • Ddataflow streaming, writing to BigQuery/BigtableCorrect
   ✓ Correct answer: C, D

   A streaming analytics pipeline ingests high-volume events via Pub/Sub, processes/transforms them in Dataflow with real-time aggregations, and writes results to BigQuery (for analytics) or Bigtable (for operational access). This architecture enables immediate insights on streaming data.

   Why the other options are wrong

   • ACloud DNS for processing is incorrect because Cloud DNS is a DNS service, not part of analytics pipelines.
   • BMemorystore for transformation is incorrect because Memorystore is for caching, not transformation.
10. Question 10Ingesting and Processing the Data

    Which Datastream capability captures ongoing changes from source databases?

    • AChange data capture (CDC) replicationCorrect
    • BIP assignment
    • CDNS resolution
    • DDisk encryption
    ✓ Correct answer: A

    Datastream's primary capability is change data capture, which continuously streams data modifications from source databases. CDC captures INSERT/UPDATE/DELETE operations as they occur, enabling real-time data synchronization and replication-based analytics workflows.

    Why the other options are wrong

    • BIP assignment is incorrect because IP assignment is a network function, not Datastream functionality.
    • CDNS resolution is incorrect because DNS resolution is name resolution, not CDC.
    • DDisk encryption is incorrect because Disk encryption is security, not CDC functionality.

Google Cloud Professional Data Engineer practice exam FAQ