GCP Professional Architect PBQs 2025

Simulate real exam scenarios with interactive case studies and architecture challenges

PBQ 1: EHR Healthcare HIPAA Migration

Scenario: Migrate an on-prem patient portal to GCP while meeting HIPAA compliance. The system handles 50K daily users with 99.99% uptime requirements.

Requirements:

  • Data residency for EU patient records
  • DDoS protection for public APIs
  • Audit logging for all PHI access
  • Disaster recovery with 15-minute RPO

Recommended Architecture:

Compute: Regional GKE clusters with auto-repair across 3 zones
Data: Cloud SQL with cross-region replicas for EU residency
Security: Cloud Armor + VPC Service Controls perimeter
DR: Scheduled snapshots with Cloud Storage Transfer
Difficulty: ★★★★☆ (4/5) | Exam Relevance: ★★★★★ (5/5)
EHR Healthcare GCP Architecture
graph TD A[Patient Portal] -->|HTTPS| B(Global Load Balancer) B --> C[GKE Cluster: us-central1] B --> D[GKE Cluster: europe-west1] C --> E[(Cloud SQL: Primary)] D --> F[(Cloud SQL: Replica)] E --> G[VPC Service Controls] F --> G G --> H{Cloud Storage} H --> I[Cloud DLP] class A,B,C,D,E,F,G,H,I gcp-service; classDef gcp-service fill:#4285F4,stroke:#1a73e8,color:white;

Key Services: GKE, Cloud SQL, VPC Service Controls, Cloud DLP

PBQ 2: Mountkirk Games Real-Time Analytics

Scenario: Design a pipeline processing 100K events/sec from multiplayer games. Must anonymize PII and support real-time dashboards.

Constraints:

  • <1s latency for cheat detection
  • 90-day data retention
  • SQL access for business teams

Recommended Architecture:

Ingestion: Pub/Sub with 1000+ subscriptions
Processing: Dataflow with DLP API integration
Storage: Bigtable (real-time) + BigQuery (analytics)
Cost Control: Partitioned tables + reservations
Difficulty: ★★★☆☆ (3/5) | Exam Relevance: ★★★★☆ (4/5)
Mountkirk Games Real-Time Analytics
graph LR A[Game Clients] --> B[Pub/Sub] B --> C[Dataflow] C --> D[Cloud DLP] D --> E[Bigtable] D --> F[BigQuery] E --> G[Real-time Dashboards] F --> H[Looker] style A fill:#4285F4,stroke:#1a73e8 style B fill:#EA4335,stroke:#b31412 style C fill:#FBBC05,stroke:#e37400 style D fill:#34A853,stroke:#0d652d style E fill:#4285F4,stroke:#1a73e8 style F fill:#EA4335,stroke:#b31412 style G fill:#FBBC05,stroke:#e37400 style H fill:#34A853,stroke:#0d652d

Key Services: Pub/Sub, Dataflow, Bigtable, BigQuery

PBQ 3: HRL Global Live Streaming

Scenario: Helicopter Racing League needs <5s latency streaming to fans worldwide. Video feeds come from 200+ sensors per helicopter.

Challenges:

  • Peak traffic during races (10x normal)
  • Emerging markets with spotty connectivity
  • Real-time telemetry overlay

Recommended Architecture:

Video: Live Stream API + Cloud CDN with edge caching
Telemetry: Pub/Sub + Dataflow for real-time processing
Emerging Markets: Regional caches + offline mode
Difficulty: ★★★★☆ (4/5) | Exam Relevance: ★★★★☆ (4/5)
Helicopter Racing League Global Live Streaming
graph TB A[Helicopter Sensors] --> B[Live Stream API] A --> C[Pub/Sub] B --> D[Cloud CDN] C --> E[Dataflow] E --> F[BigQuery] E --> G[Bigtable] D --> H[Global Users] style A fill:#4285F4,stroke:#1a73e8 style B fill:#EA4335,stroke:#b31412 style C fill:#FBBC05,stroke:#e37400 style D fill:#34A853,stroke:#0d652d style E fill:#4285F4,stroke:#1a73e8 style F fill:#EA4335,stroke:#b31412 style G fill:#FBBC05,stroke:#e37400 style H fill:#34A853,stroke:#0d652d

Key Services: Live Stream API, Cloud CDN, Dataflow

PBQ 4: TerramEarth Dealership Integration

Scenario: Connect 500+ dealerships' on-prem systems to GCP for vehicle diagnostics. Many locations have unreliable internet.

Requirements:

  • Offline data collection
  • 100MB/day bandwidth limit per site
  • Secure PII transmission

Recommended Architecture:

Edge: Firebase with offline persistence
Connectivity: Cloud IoT Core for batch uploads
Security: IAP for zero-trust access
Difficulty: ★★★☆☆ (3/5) | Exam Relevance: ★★★★☆ (4/5)
TerramEarth Dealership Integration
graph LR A[Dealership Systems] --> B[Firebase] B --> C[Cloud IoT Core] C --> D[Pub/Sub] D --> E[Dataflow] E --> F[BigQuery] style A fill:#4285F4,stroke:#1a73e8 style B fill:#EA4335,stroke:#b31412 style C fill:#FBBC05,stroke:#e37400 style D fill:#34A853,stroke:#0d652d style E fill:#4285F4,stroke:#1a73e8 style F fill:#EA4335,stroke:#b31412

Key Services: Firebase, Cloud IoT Core, IAP

PBQ 5: $100K/Month Bill Reduction

Scenario: A company's GCP bill jumped from $50K to $150K/month after migration. Identify waste and implement fixes.

Findings:

  • 24/7 n2-standard-32 VMs running at 5% CPU
  • Multi-region Cloud SQL replicas unused
  • No committed use discounts

Recommended Fixes:

Compute: Switch to preemptible VMs + Autoscaling
Database: Remove unused replicas + enable CUDs
Storage: Move cold data to Nearline
Difficulty: ★★★☆☆ (3/5) | Exam Relevance: ★★★★★ (5/5)
pie title Monthly Savings "Preemptible VMs" : 35 "Committed Use" : 25 "Autoscaling" : 20 "Storage Tiering" : 15 "Right-Sizing" : 5

Key Services: Committed Use Discounts, Preemptible VMs, Autoscaler

Master All 5 Case Studies

Unlock complete PBQ sets with detailed feedback and progress tracking

Start Free 7-Day Trial
Download on App Store Get on Google Play