1️⃣ GCP Fundamentals — Deep Dive¶
🌍 Google Cloud Platform Overview¶
Google Cloud Platform (GCP) is a global cloud infrastructure that provides computing, storage, networking, and advanced services to build, deploy, and scale applications.
Core Service Categories¶
| Category | Services | Use Cases |
|---|---|---|
| Compute | Compute Engine, GKE, Cloud Run, App Engine | VMs, containers, serverless apps |
| Storage | Cloud Storage, Persistent Disk, Filestore | Object storage, block storage, file systems |
| Networking | VPC, Cloud Load Balancing, Cloud CDN | Network isolation, traffic distribution |
| Databases | Cloud SQL, Firestore, Bigtable, Spanner | Relational, NoSQL, wide-column, global DB |
| AI/ML | Vertex AI, AutoML, Vision API, Natural Language | Machine learning, computer vision, NLP |
| DevOps | Cloud Build, Artifact Registry, Cloud Deploy | CI/CD, container registry, deployment |
| Security | IAM, Secret Manager, Security Command Center | Identity, secrets, security posture |
| Observability | Cloud Monitoring, Cloud Logging, Cloud Trace | Metrics, logs, distributed tracing |
Key Differentiators¶
┌─────────────────────────────────────────────────────────┐
│ Google's Infrastructure Advantages │
├─────────────────────────────────────────────────────────┤
│ ✓ Same infrastructure as Gmail, YouTube, Search │
│ ✓ Private global fiber network (100+ Tbps) │
│ ✓ Live migration of VMs (no downtime) │
│ ✓ Custom machine types (granular resource control) │
│ ✓ Per-second billing (cost optimization) │
│ ✓ Sustained use discounts (automatic savings) │
│ ✓ Preemptible/Spot VMs (up to 91% discount) │
└─────────────────────────────────────────────────────────┘
🌎 Regions & Zones¶
What is a Region?¶
A region is an independent geographic location consisting of multiple zones. Each region is completely isolated from other regions for fault tolerance and stability.
Global Region Coverage (2026)¶
| Region | Location | Zones | Key Features |
|---|---|---|---|
us-central1 |
Iowa, USA | 4 | Low latency to US central |
us-east1 |
South Carolina, USA | 3 | Low latency to US east coast |
us-west1 |
Oregon, USA | 3 | Low latency to US west coast |
europe-west1 |
Belgium | 3 | EU data residency |
europe-west2 |
London, UK | 3 | UK data residency |
asia-south1 |
Mumbai, India | 3 | Low latency to India |
asia-southeast1 |
Singapore | 3 | APAC hub |
asia-northeast1 |
Tokyo, Japan | 3 | Low latency to Japan |
australia-southeast1 |
Sydney | 3 | Australia data residency |
southamerica-east1 |
São Paulo, Brazil | 3 | South America coverage |
40+ regions globally with continuous expansion.
Region Selection Criteria¶
┌──────────────────────────────────────────────────────────┐
│ How to Choose a Region │
├──────────────────────────────────────────────────────────┤
│ │
│ 1. LATENCY │
│ └─→ Choose region closest to your users │
│ │
│ 2. DATA RESIDENCY & COMPLIANCE │
│ └─→ GDPR (EU), data sovereignty requirements │
│ │
│ 3. SERVICE AVAILABILITY │
│ └─→ Not all services available in all regions │
│ │
│ 4. COST │
│ └─→ Pricing varies by region (Iowa < London) │
│ │
│ 5. DISASTER RECOVERY │
│ └─→ Multi-region for business continuity │
└──────────────────────────────────────────────────────────┘
Region Selection Example¶
Scenario: E-commerce app with global users
┌─────────────────────────────────────────────────────┐
│ User Base → Recommended Region │
├─────────────────────────────────────────────────────┤
│ North America → us-central1 (primary) │
│ Europe → europe-west1 (secondary) │
│ Asia → asia-southeast1 (secondary) │
└─────────────────────────────────────────────────────┘
Architecture:
- Multi-region deployment
- Global Load Balancer
- Cloud CDN for static content
- Cross-region replication for data
What is a Zone?¶
A zone is a deployment area within a region that represents a single physical data center or a cluster of data centers.
Zone Characteristics¶
Region: us-central1 (Iowa)
│
├── us-central1-a ← Zone A (Data Center 1)
│ ├── Independent power supply
│ ├── Independent cooling
│ ├── Independent networking
│ └── Isolated failure domain
│
├── us-central1-b ← Zone B (Data Center 2)
│ └── Physically separate from Zone A
│
├── us-central1-c ← Zone C (Data Center 3)
│ └── Physically separate from Zones A & B
│
└── us-central1-f ← Zone F (Data Center 4)
└── Additional capacity and redundancy
Zone Isolation Benefits¶
┌────────────────────────────────────────────────────┐
│ Why Multiple Zones Matter │
├────────────────────────────────────────────────────┤
│ │
│ ⚡ Power Failure in Zone A │
│ → Zones B, C, F continue operating │
│ │
│ 🔥 Fire/Natural Disaster in Zone B │
│ → Zones A, C, F unaffected │
│ │
│ 🔧 Maintenance in Zone C │
│ → Zones A, B, F handle traffic │
│ │
│ 🌐 Network Issue in Zone F │
│ → Zones A, B, C maintain connectivity │
└────────────────────────────────────────────────────┘
High Availability Architecture¶
Single Zone Deployment (NOT RECOMMENDED)¶
┌─────────────────────────────────────┐
│ Region: us-central1 │
│ │
│ ┌───────────────────────────────┐ │
│ │ Zone: us-central1-a │ │
│ │ │ │
│ │ ┌─────────────────────────┐ │ │
│ │ │ Application Instances │ │ │
│ │ │ • VM-1 │ │ │
│ │ │ • VM-2 │ │ │
│ │ │ • Database │ │ │
│ │ └─────────────────────────┘ │ │
│ └───────────────────────────────┘ │
└─────────────────────────────────────┘
❌ Single Point of Failure
❌ No redundancy
❌ Downtime during zone failure
Multi-Zone Deployment (RECOMMENDED)¶
┌──────────────────────────────────────────────────────────────┐
│ Region: us-central1 │
│ │
│ ┌──────────────────┐ ┌──────────────────┐ ┌────────────┐│
│ │ Zone A │ │ Zone B │ │ Zone C ││
│ │ │ │ │ │ ││
│ │ ┌──────────────┐ │ │ ┌──────────────┐ │ │ ┌────────┐││
│ │ │ VM-1 │ │ │ │ VM-2 │ │ │ │ VM-3 │││
│ │ │ App Instance │ │ │ │ App Instance │ │ │ │ App │││
│ │ └──────────────┘ │ │ └──────────────┘ │ │ └────────┘││
│ │ │ │ │ │ ││
│ │ ┌──────────────┐ │ │ ┌──────────────┐ │ │ ││
│ │ │ DB Primary │◄─┼──┼─┤ DB Replica │ │ │ ││
│ │ └──────────────┘ │ │ └──────────────┘ │ │ ││
│ └──────────────────┘ └──────────────────┘ └────────────┘│
│ ▲ ▲ ▲ │
│ └─────────────────────┴────────────────────┘ │
│ Load Balancer (Regional) │
└──────────────────────────────────────────────────────────────┘
✓ High Availability (99.99% SLA)
✓ Automatic failover
✓ Zero downtime during zone failure
✓ Load distribution across zones
Real-World HA Pattern¶
┌─────────────────────────────────────────────────────────────┐
│ Production Architecture: E-commerce Platform │
└─────────────────────────────────────────────────────────────┘
┌──────────────────────┐
│ Global Load │
│ Balancer │
└──────────┬───────────┘
│
┌────────────────┼────────────────┐
│ │ │
┌─────────▼────────┐ ┌────▼──────────┐ ┌──▼──────────┐
│ Zone A │ │ Zone B │ │ Zone C │
│ │ │ │ │ │
│ ┌──────────────┐ │ │ ┌───────────┐│ │ ┌─────────┐ │
│ │ GKE Node │ │ │ │ GKE Node ││ │ │ GKE Node│ │
│ │ • Frontend │ │ │ │ • Frontend││ │ │ •Frontend│ │
│ │ • Backend │ │ │ │ • Backend ││ │ │ •Backend│ │
│ └──────────────┘ │ │ └───────────┘│ │ └─────────┘ │
│ │ │ │ │ │
│ ┌──────────────┐ │ │ ┌───────────┐│ │ │
│ │ Cloud SQL │◄┼─┼─┤ Read ││ │ │
│ │ Primary │ │ │ │ Replica ││ │ │
│ └──────────────┘ │ │ └───────────┘│ │ │
│ │ │ │ │ │
│ ┌──────────────┐ │ │ ┌───────────┐│ │ ┌─────────┐ │
│ │ Redis │ │ │ │ Redis ││ │ │ Redis │ │
│ │ Cache │ │ │ │ Cache ││ │ │ Cache │ │
│ └──────────────┘ │ │ └───────────┘│ │ └─────────┘ │
└──────────────────┘ └───────────────┘ └─────────────┘
Benefits:
• 99.99% uptime SLA
• Automatic failover in seconds
• Load balanced across zones
• Database replication for reads
• Distributed caching
🌐 Global Infrastructure¶
Google Cloud operates one of the largest and most advanced private networks in the world.
Network Architecture¶
┌────────────────────────────────────────────────────────────────┐
│ Google's Global Network Infrastructure │
└────────────────────────────────────────────────────────────────┘
┌──────────────────────────────────────────────────────┐
│ Edge Network (200+ Points of Presence) │
│ • CDN Edge Locations │
│ • Cloud Armor (DDoS Protection) │
│ • Global Load Balancing Entry Points │
└────────────────┬─────────────────────────────────────┘
│
┌────────────────▼─────────────────────────────────────┐
│ Premium Tier Network (Google's Private Backbone) │
│ • 100+ Tbps capacity │
│ • Subsea cables owned by Google │
│ • Low latency, high reliability │
│ • Traffic stays on Google network │
└────────────────┬─────────────────────────────────────┘
│
┌────────────────▼─────────────────────────────────────┐
│ Regional Networks (40+ Regions) │
│ • VPC Networks │
│ • Regional Load Balancers │
│ • Private Service Connect │
└────────────────┬─────────────────────────────────────┘
│
┌────────────────▼─────────────────────────────────────┐
│ Zonal Resources (120+ Zones) │
│ • Compute Instances │
│ • Persistent Disks │
│ • GKE Clusters │
└──────────────────────────────────────────────────────┘
Network Tiers¶
Premium Tier (Default)¶
User Request Flow (Premium Tier):
┌──────────┐
│ User │ (Mumbai, India)
│ Browser │
└────┬─────┘
│
▼
┌─────────────────┐
│ Nearest Edge │ ← Enters Google network immediately
│ Location │ (Mumbai PoP)
└────┬────────────┘
│
▼ Google Private Network (Low latency, high reliability)
┌─────────────────┐
│ Google │
│ Backbone │ ← Traffic stays on Google's network
│ Network │
└────┬────────────┘
│
▼
┌─────────────────┐
│ Destination │ (us-central1)
│ Region │
└─────────────────┘
Benefits:
✓ Lowest latency
✓ Highest reliability
✓ Global load balancing
✓ Better performance
✓ SLA guarantees
Standard Tier¶
User Request Flow (Standard Tier):
┌──────────┐
│ User │ (Mumbai, India)
│ Browser │
└────┬─────┘
│
▼ Public Internet
┌─────────────────┐
│ ISP Network │ ← Uses public internet
│ (Multiple hops)│
└────┬────────────┘
│
▼ Public Internet
┌─────────────────┐
│ Regional │ ← Enters Google network at region
│ Entry Point │ (us-central1)
└────┬────────────┘
│
▼
┌─────────────────┐
│ Destination │
│ Region │
└─────────────────┘
Trade-offs:
✓ Lower cost
✗ Higher latency
✗ Less reliable
✗ Regional load balancing only
Global Load Balancing¶
┌────────────────────────────────────────────────────────────┐
│ Global HTTP(S) Load Balancer Architecture │
└────────────────────────────────────────────────────────────┘
┌──────────────────┐
│ Anycast IP │
│ (Single Global │
│ IP Address) │
└────────┬─────────┘
│
┌────────────────────┼────────────────────┐
│ │ │
┌────▼─────┐ ┌────▼─────┐ ┌────▼─────┐
│ us-east1 │ │ eu-west1 │ │ asia-se1 │
│ │ │ │ │ │
│ Backend │ │ Backend │ │ Backend │
│ Services │ │ Services │ │ Services │
└──────────┘ └──────────┘ └──────────┘
Features:
• Automatic routing to nearest healthy backend
• Cross-region failover
• SSL termination at edge
• Cloud CDN integration
• DDoS protection with Cloud Armor
Latency Optimization¶
Edge Caching with Cloud CDN¶
Without CDN:
┌──────┐ ┌──────────┐
│ User │──────── 200ms latency ────────────▶│ Origin │
│ │◀─────── Full content ──────────────│ Server │
└──────┘ └──────────┘
With CDN:
┌──────┐ ┌─────────┐ ┌──────────┐
│ User │─ 5ms ──▶│ CDN │─── Cache ────▶│ Origin │
│ │◀────────│ Edge │ Miss │ Server │
└──────┘ └─────────┘ └──────────┘
│
└─ Cache Hit (5ms response)
Performance Improvement:
• 95% faster response time
• Reduced origin load
• Lower bandwidth costs
• Better user experience
Best Practices¶
1. Multi-Zone Deployment¶
# Deploy across multiple zones for HA
gcloud compute instance-groups managed create web-group \
--base-instance-name web \
--template web-template \
--size 3 \
--zones us-central1-a,us-central1-b,us-central1-c
2. Regional Resources¶
# Use regional resources for automatic zone distribution
gcloud compute addresses create web-ip \
--region us-central1 \
--network-tier PREMIUM
3. Health Checks¶
# Configure health checks for automatic failover
gcloud compute health-checks create http web-health-check \
--port 80 \
--check-interval 10s \
--timeout 5s \
--unhealthy-threshold 3 \
--healthy-threshold 2
Cost Optimization¶
Region Pricing Comparison¶
| Region | Compute (n1-standard-1/month) | Storage (GB/month) |
|---|---|---|
| us-central1 (Iowa) | $24.27 | $0.020 |
| us-east1 (S. Carolina) | $24.27 | $0.020 |
| europe-west2 (London) | $28.18 | $0.023 |
| asia-southeast1 (Singapore) | $28.67 | $0.023 |
Tip: Use Iowa (us-central1) for cost-effective deployments when latency permits.