Google Cloud Database Engineer Professional Practice Exam

1. 1.An application needs a relational database with strong consistency, horizontal write scaling across continents, and a 99.999% availability SLA. Which Google Cloud service fits?

   • AAlloyDB for PostgreSQL
   • BBigtable
   • CCloud SQL for PostgreSQL
   • DSpannerCorrect answer

   Why

   Spanner provides externally consistent transactions with horizontal scaling across regions and offers a 99.999% availability SLA in multi-region configurations. Cloud SQL scales writes vertically on a single primary, AlloyDB scales reads horizontally but writes on one primary, and Bigtable is a NoSQL wide-column store without relational transactions across arbitrary rows.

2. 2.A time-series workload ingests millions of sensor readings per second and queries them by device and time range with single-digit millisecond latency. Which database fits?

   • ABigtableCorrect answer
   • BBigQuery
   • CCloud SQL for MySQL
   • DSpanner

   Why

   Bigtable is designed for very high write throughput and low-latency lookups on a row key, which suits a device identifier combined with a timestamp. Cloud SQL cannot sustain that ingest rate, Spanner is far more expensive for this pattern and is optimized for transactional relational access, and BigQuery is an analytical warehouse rather than a low-latency key lookup store.

3. 3.A team runs a PostgreSQL application that needs much better performance on both transactional and analytical queries, while remaining PostgreSQL compatible. Which service should they evaluate?

   • ASpanner with the PostgreSQL interface
   • BCloud SQL for PostgreSQL
   • CBigtable
   • DAlloyDB for PostgreSQLCorrect answer

   Why

   AlloyDB for PostgreSQL is PostgreSQL compatible and adds a columnar engine and an optimized storage layer that accelerate analytical queries alongside transactional work. Cloud SQL for PostgreSQL offers compatibility without the columnar acceleration, Spanner requires schema and application changes, and Bigtable is not PostgreSQL compatible at all.

4. 4.An application must store semi-structured documents, sync them to mobile clients offline, and scale automatically without capacity planning. Which service fits?

   • ABigtable
   • BMemorystore for Redis
   • CFirestoreCorrect answer
   • DCloud SQL for MySQL

   Why

   Firestore stores documents, provides offline synchronization for mobile and web SDKs, and scales automatically without provisioning. Cloud SQL is relational and has no offline sync, Bigtable is a wide-column store without document sync, and Memorystore is an in-memory cache rather than a durable document store.

5. 5.A read-heavy application repeatedly issues the same expensive query against Cloud SQL, and the team wants sub-millisecond responses for repeat reads. Which service should be added?

   • AMemorystore for Redis as a cache layerCorrect answer
   • BMigrating the workload to Bigtable
   • CA larger Cloud SQL machine type
   • DAdditional Cloud SQL read replicas

   Why

   Memorystore provides a managed in-memory cache that serves repeated reads in sub-millisecond time, relieving the database. Adding read replicas reduces primary load but does not reach sub-millisecond latency, increasing the instance size does not cache anything, and moving to Bigtable changes the data model entirely.

Practice all 65 questions

Frequently asked questions