Google PCA Design and plan a cloud solution architecture Practice Question
A financial services firm runs a global trading platform on Google Cloud. The architecture must survive the loss of an entire region with a recovery point objective of zero and a recovery time objective of under one minute, and it must keep strong consistency for order records. Which design should the architect recommend?
⚠ Common exam trap
The trap here is treating any multi-region data service as equivalent, when replicas that are asynchronous cannot satisfy a zero recovery point objective.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Deploy the application in two regions behind a global external Application Load Balancer, and store order records in Cloud Spanner with a multi-region instance configuration.
Zero RPO with sub-minute RTO for transactional records requires synchronous replication plus automated failover. Cloud Spanner multi-region configurations replicate synchronously and expose externally consistent reads and writes, so a committed order survives a full region outage. The global external Application Load Balancer provides anycast front ends and health-check-based failover to the surviving region. Together they deliver the availability and consistency guarantees the trading platform needs without manual intervention.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Deploy the application in two regions behind a global external Application Load Balancer, and store order records in Cloud Spanner with a multi-region instance configuration.
Why this is correct
A multi-region Cloud Spanner configuration replicates data synchronously across regions and provides external consistency, so no committed order is lost when a region fails, satisfying the zero RPO. Spanner redirects traffic to surviving replicas automatically, and the global external Application Load Balancer steers users to healthy backends within seconds, meeting the sub-minute RTO. This combination is the standard design for globally consistent, region-fault-tolerant transactional systems on Google Cloud.
- ✗
Deploy the application in two regions behind a global external Application Load Balancer, and store order records in a Cloud SQL for PostgreSQL instance with a cross-region read replica promoted on failover.
Why it's wrong here
Cloud SQL cross-region replicas are asynchronous, so a region loss can lose the most recent transactions, which violates the zero-RPO requirement. Promotion is also a manual or orchestrated operation that typically takes longer than one minute. The load balancer can shift traffic, but the data tier cannot guarantee no data loss, so this design fails the stated objectives even though it looks multi-region.
- ✗
Deploy the application in two regions behind a global external Application Load Balancer, and store order records in a Bigtable instance with a multi-cluster routing policy.
Why it's wrong here
Bigtable multi-cluster routing gives high availability and eventual consistency across clusters, but it does not provide external consistency or multi-row ACID transactions. An order system that must never lose or reorder committed records cannot rely on eventual consistency, and some writes may be acknowledged in one cluster before replicating. Bigtable is excellent for high-throughput key-value time-series workloads, not for strongly consistent financial ledgers.
- ✗
Deploy the application in two regions behind a global external Application Load Balancer, and store order records in a multi-region Cloud Storage bucket mounted as a file system on the application VMs.
Why it's wrong here
Cloud Storage is an object store with eventual visibility semantics for listings and it is not designed to be mounted as a POSIX file system for transactional writes. It offers no multi-row transactions, no ordering guarantees, and no locking, so concurrent order updates would corrupt records. Durability across regions does not translate into consistency or low-latency transactional access, so this option cannot meet either the RPO or the consistency requirement.
Go deeper
Related to this question
Learn chapter
Introduction to Google Cloud Platform
Key term
Spanner
Google Cloud's globally distributed, strongly consistent, and horizontally scalable database service designed for mission-critical, transactional workloads.
Key term
RPO
Recovery Point Objective (RPO) is the maximum acceptable amount of data loss measured in time, defining how recent data must be to resume operations after a disruption.
About these practice questions
One of 807 original PCA practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Google Cloud exam blueprint
This PCA practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the PCA exam.