Prepare and test your skills

Prepare and test your skills

Free sample question, answered and explained below. The Professional Data Engineer practice bank has 1,353 more like it. Try 10 free: instant score, no account.
An enterprise is architecting a globally distributed mobile platform deployed across the United States, Europe, and Asia. The application encompasses two distinct operational data paths:
According to CAP theorem principles and Google Cloud database characteristics, which architectural design correctly implements these requirements?
This architecture leverages a specialized multi-database pattern combining a non-relational distributed NoSQL database configured for eventual consistency with Cloud Spanner, Google Cloud's fully managed, globally distributed, horizontally scalable, and strongly consistent relational database.
The CAP theorem dictates that a distributed system cannot simultaneously achieve Consistency, Availability, and Partition Tolerance under all network conditions. Dividing workloads across an AP-leaning eventually consistent NoSQL data store and a CP-leaning strongly consistent distributed engine (Cloud Spanner) ensures optimal performance and correctness for each domain.
This architecture routes all application operations—including ephemeral social counters, presence markers, and financial ledger modifications—through a single multi-region Cloud Spanner cluster configured for strict synchronous read-write transactions.
The workload consisted exclusively of high-stakes transactional records (such as securities trading or banking) where every single data point mandates absolute serializability and external consistency across continents.
Under the CAP theorem, opting for global strong consistency ($C$) inherently trades off lower write latencies and total partition availability ($A$), making it an anti-pattern for high-volume social activity metrics.
This proposal relies on object storage (Cloud Storage) distributed across regions, cached at edge points of presence (PoPs) using Cloud CDN to serve and record application data via HTTP caching semantics.
The goal was to deliver a static or warm-standby disaster recovery website where user-specific dynamic transactions are halted and replaced by read-only cached content during an infrastructure outage.
Confusing application-tier edge caching with distributed database state management leads to architectural failure when stateful, transactional, or high-throughput mutational workloads are involved.
This pattern utilizes Cloud SQL (a regional relational database engine such as MySQL or PostgreSQL) operating in a single primary region, paired with asynchronous read replicas deployed in secondary geographical regions.
This design would be suitable for a centralized, low-volume corporate system that operates primarily out of a single region and only requires remote read acceleration for static reporting tools with relaxed RPO requirements.
Relational databases utilizing traditional single-master asynchronous replication cannot provide true global active-active write availability or multi-region strong transactional guarantees under CAP constraints.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated