Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Professional Cloud Database Engineer practice bank has an additional 769 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
Your organization runs a mission-critical AlloyDB for PostgreSQL cluster in a single Google Cloud region supporting both heavy OLTP writes and latency-sensitive analytical queries across two separate databases (Finance and Inventory). You need to design the intra-region architecture to meet the following requirements:
Which cluster architecture should you implement?
This approach deploys a non-HA (zonal) primary instance in one zone alongside a single-node read pool instance located in another zone, relying on the read pool to serve analytical reads and cover zonal outages.
The cluster were a development or test environment where cost savings are prioritized over availability and extended downtime during zonal failures is acceptable.
To achieve true intra-region resilience across both reads and writes, primary HA and multi-node read pools must both be configured rather than treating a read pool as a substitute for a primary standby node.
This architecture enables high availability on the primary instance for write resilience while creating one single-node read pool that shares the exact same machine series as the primary and serves both Finance and Inventory queries.
C4, C4A, or Z3 versus N2 on the primary) to optimize price-performance or work around regional capacity limits without restarting the primary.The cluster served a single analytical database where read downtime during a node failure was tolerable and budget constraints restricted the read pool to one node.
Decoupling read pool count, node count, and machine types from the primary instance is a key optimization strategy in AlloyDB for balancing cost, cache hit ratios, and high availability.
An AlloyDB high availability (HA) primary instance pairs an active node and a standby node across two different zones in the same region, backed by a regional log persistor and regional storage service. Pairing this with multiple multi-node read pool instances separates transactional writes from analytical reads while providing zonal fault tolerance across every tier of the cluster.
C4 or Z3) from the primary instance without requiring primary restarts.This design isolates heavy OLTP write workloads on the HA primary, guarantees full regional resilience for both writes and reads, and maximizes analytical query speed through dedicated caches and selective columnar acceleration.
This configuration provisions an HA primary instance with an active node and a standby node in separate zones, turns on the columnar engine directly on the primary instance, and attempts to use the standby node as a read-scaling endpoint for analytical workloads.
The workload consisted exclusively of OLTP transactions requiring zonal high availability and zero data loss, with no requirement to offload or horizontally scale analytical read queries.
Even when hot standby features run the standby as a replica with warm caches to speed up failover, read pool instances remain the architectural component built to offload, load-balance, and scale read-intensive workloads horizontally.
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