Advanced
Prepare and test your skills
Advanced
Prepare and test your skills
Active-Active Multi-Region Ingestion with Spanner
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 designing a mission-critical financial data ingestion and processing system on Google Cloud. The system must withstand both zonal and complete regional infrastructure failures with an RPO close to zero and an automated, minimal RTO without requiring manual failover procedures.
Which architecture meets these requirements?
Explanations
Deploy a Global External Application Load Balancer routing traffic to regional ingestion services across two regions, buffer events using global Pub/Sub topics, and persist state to a multi-region Cloud Spanner instance.
Deploy a Global External Application Load Balancer routing traffic to regional ingestion services across two regions, buffer events using global Pub/Sub topics, and persist state to a multi-region Cloud Spanner instance.
What the Multi-Region Active-Active Architecture Is
This architecture implements a fully distributed active-active multi-region deployment pattern on Google Cloud. It utilizes globally distributed and multi-regional managed services that natively provide fault tolerance and data synchronization across separate geographic failure domains without requiring manual failover intervention.
How It Meets the Zero RPO and Minimal RTO Requirements
- Automated Regional Routing (Minimal RTO): A Global External Application Load Balancer provides a single anycast IP front end that automatically detects regional backend outages and dynamically steers client ingestion traffic to healthy regional Compute Engine or GKE clusters.
- Global Ingestion Buffering: Cloud Pub/Sub is a globally scoped service where topics can accept messages from any region, ensuring high availability and continuous ingestion during infrastructure disruptions.
- Synchronous Cross-Region Consistency (RPO ~ 0): A multi-region Cloud Spanner instance uses synchronous replication and distributed consensus across regions, ensuring strong consistency (external consistency) and zero data loss in the event of an entire regional failure.
Key Advantages and Capabilities
- Built-in Redundancy: Avoids single points of failure (SPOFs) at networking, compute, messaging, and database layers.
- Zero-Touch Automated Failover: Health checks and global load balancing seamlessly reroute incoming data streams during disasters without DNS propagation delays.
- High Availability SLA: Multi-region Cloud Spanner configurations deliver industry-leading 99.999% availability SLAs.
- Separation of Concerns: Decouples data computation from durable distributed storage.
Why This Is Optimal
Choosing global and multi-regional resources over regional or zonal constructs minimizes operational complexity and eliminates the risk of human error during disaster recovery events.
Deploy compute instances in a single zone attached to Regional Persistent Disks configured with synchronous replication across regions, fronted by an internal load balancer.
Deploy compute instances in a single zone attached to Regional Persistent Disks configured with synchronous replication across regions, fronted by an internal load balancer.
What Regional Persistent Disks Are
Regional Persistent Disk (Regional PD) and Hyperdisk Balanced High Availability are Google Cloud storage features that provide synchronous block storage replication at the infrastructure layer.
Why It Doesn't Meet the Requirement
- Replication Boundary: Regional Persistent Disks replicate data synchronously between two zones in the same region, not across multiple regions. They cannot replicate data cross-region because synchronous disk writes across regions would incur prohibitive network latency.
- Single-Zone Compute Bottleneck: Running compute instances in a single zone creates a single point of failure (SPOF) for processing.
- Lack of Global Routing: An internal load balancer is regionally scoped and cannot distribute external client traffic or manage cross-region failover.
Key Characteristics and Limitations
- Intra-Region HA: Regional PD is designed strictly for zonal disaster recovery and high availability within a single region.
- Storage Overhead: Uses double the storage capacity to maintain active replication across two local zones.
- Vulnerable to Regional Outages: A regional failure renders both disk replicas inaccessible.
This Answer Would Be Correct If
The workload required synchronous, zero-RPO storage failover between two zones within the same Google Cloud region for stateful single-region VM workloads.
Deploy a Global External Application Load Balancer routing to ingestion workers that stream raw transactions directly into a BigQuery multi-region dataset (such as the US multi-region).
Deploy a Global External Application Load Balancer routing to ingestion workers that stream raw transactions directly into a BigQuery multi-region dataset (such as the US multi-region).
What the BigQuery Multi-Region Option Is
This architecture attempts to use a Global External Application Load Balancer alongside BigQuery multi-region datasets to ingest streaming transactions into a broad geographical boundary (e.g., US or EU).
Why It Doesn't Meet the Requirement
- Misconception About BigQuery Multi-Region: In BigQuery, a multi-region location designation does not provide automated cross-region replication or cross-region disaster recovery. The underlying data is stored redundantly across zones within a single primary region in that geographic boundary.
- Regional Outage Vulnerability: If the specific primary region hosting the BigQuery dataset suffers a catastrophic outage, analytical queries and streaming inserts become unavailable until service in that region is restored.
- Transaction Processing Constraints: BigQuery is an analytical data warehouse (OLAP) rather than an operational database designed for low-latency, transactional state processing with automated transactional failover.
Key Characteristics and Limitations
- Zonal Redundancy Only: BigQuery automatically replicates data across two zones in a single region.
- No Automatic Region Replication: Cross-region availability in BigQuery requires configuring scheduled dataset copies to another region.
This Answer Would Be Correct If
The goal was to run large-scale analytical SQL queries across petabytes of historical logs within a broad geographic location without requiring real-time OLTP transaction support and automated multi-region failover.
Deploy a Regional External Application Load Balancer in the primary region, publish streaming events to regional Pub/Sub topics, and store state in a Cloud SQL Enterprise instance configured with High Availability (HA).
Deploy a Regional External Application Load Balancer in the primary region, publish streaming events to regional Pub/Sub topics, and store state in a Cloud SQL Enterprise instance configured with High Availability (HA).
What the Regional Cloud SQL HA Architecture Is
This pattern deploys a regional active-passive high availability (HA) configuration within a single Google Cloud region, utilizing a regional load balancer and Cloud SQL with automated synchronous replication between two zones.
Why It Doesn't Meet the Requirement
- Fails Regional Disaster Tolerance: The Regional External Application Load Balancer and regional Cloud SQL HA configuration are scoped strictly to a single region. If the entire primary region experiences an outage, the load balancer cannot route traffic and Cloud SQL cannot fail over.
- Requires Manual Intervention: Cloud SQL HA only provides automatic failover between zones in the same region, not across regions. Cross-region disaster recovery requires cross-region read replicas, which demand manual promotion, application connection string reconfiguration, and external DNS failover.
Key Characteristics and Limitations
- Zonal Resilience Only: Cloud SQL HA protects against single-zone failures via synchronous standby instances, achieving high availability within a region.
- Regional Scope: Regional load balancers and regional instances stop accepting traffic during full regional failures.
- Recovery Delay: Manual cross-region promotion introduces extended Recovery Time Objectives (RTO).
This Answer Would Be Correct If
The workload only required high availability against zonal failures within a single geographical region while minimizing cross-region egress and operational cost.
Try 10 interactive questions like this one, free
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
- 1,353 practice questions across all 5 exam domains
- Every answer explained: why it is right, and why the others are not
- Scout, an AI tutor that answers your follow-up questions
- Readiness tracking by domain, so you know when to book the exam
Related questions
Keep the momentum going with these hand-picked practice scenarios
AdvancedProfessional Data Engineer
Cloud Composer Cross-Region Disaster Recovery
Disaster recovery and fault tolerance
AdvancedProfessional Data Engineer
OnVault Snapshot Pool Offloading
Disaster recovery and fault tolerance
AdvancedProfessional Data Engineer
BI Engine Materialized View Smart Tuning
BigQuery features for business intelligence (e.g., BI Engine, materialized views)
AdvancedProfessional Data Engineer
BigQuery Baseline and Autoscaling Slots
Monitoring planned usage
AdvancedProfessional Data Engineer
Cloud Composer Orchestration for Dataproc Serverless
Processing data
AdvancedProfessional Cloud DevOps Engineer
Cloud Storage Bucket Lock WORM Retention
Exporting and retaining logs (e.g., routing to BigQuery, Pub/Sub, Cloud Storage)
Last updated Validated