High level architecture

Google Cloud Architecture

Your current architecture can be modernized on Google Cloud using managed services while keeping the same logical flow.

Users (Mobile/Desktop)
        │
        ▼
Cloud Load Balancer + Cloud Armor
        │
        ▼
Frontend/UI Layer
(2 Product UI Pods, 2 Order UI Pods)
Hosted on GKE
        │
        ▼
API Gateway
(Google API Gateway / Apigee)
        │
 ┌──────┴─────────┐
 │                │
 ▼                ▼

Product Service    Order Service
   Route              Route

Internal Load Balancer
       │
 ┌─────┴─────┐
 ▼           ▼

Product Namespace
(GKE)
3 Java Pods

Order Namespace
(GKE)
3 Python Pods

       │
       ▼

Cloud SQL
(Product DB)

Cloud SQL
(Order DB)

Read Replicas

---

Component Mapping

Hybrid Architecture	Google Cloud Equivalent
F5 ASM	Cloud Armor
F5 LTM	Global HTTP(S) Load Balancer
UI Instances	GKE Deployments
API Gateway	Google API Gateway or Apigee
HAProxy/Ingress	GKE Service + Internal Load Balancer
Java Pods	GKE Java Deployment
Python Pods	GKE Python Deployment
Databases	Cloud SQL PostgreSQL
ELK	Cloud Logging + Cloud Monitoring or Elastic on GCP
Jenkins	Cloud Build / Jenkins on GKE
Helm	Helm on GKE
Keycloak (UC)	Keycloak on GKE or Identity Platform
Kubernetes	GKE
Container Registry	Artifact Registry

---

Detailed Architecture

1. Edge Layer

Cloud Load Balancer

Equivalent to F5 LTM.

Responsibilities

• SSL termination
• Global traffic distribution
• Health checks
• Cross-region routing

Cloud Armor

Equivalent to F5 ASM.

Responsibilities

• WAF
• DDoS protection
• Rate limiting
• OWASP Top 10 protection
• Geo-blocking

NFR Coverage

NFR	GCP Service
Security	Cloud Armor
Availability	Global Load Balancer
Performance	Edge POPs
Reliability	Health checks

---

2. UI Layer

Google Kubernetes Engine (GKE)

Deploy:

Product UI
  ├─ Pod 1
  └─ Pod 2

Order UI
  ├─ Pod 1
  └─ Pod 2

NFR Coverage

NFR	Implementation
Availability	Multi-pod deployment
Scalability	HPA
Performance	CDN integration
Maintainability	Separate deployments

---

3. API Management Layer

Option 1: Google API Gateway

Suitable for medium-sized workloads.

Option 2: Apigee

Enterprise-grade API management.

Features:

• OAuth
• JWT Validation
• Rate limiting
• Monetization
• API Analytics

NFR Coverage

NFR	Implementation
Security	Authentication
Observability	API Analytics
Scalability	Managed gateway
Governance	API lifecycle management

---

4. Product Domain

Product Namespace (GKE)

product-service
├── Java Pod 1
├── Java Pod 2
└── Java Pod 3

Use:

• Deployment
• HPA
• ClusterIP Service

Instead of HAProxy:

GKE Service
+
Internal Load Balancer

NFR Coverage

NFR	Implementation
Scalability	HPA
Reliability	Multiple replicas
Availability	Self-healing
Maintainability	Microservice isolation

---

5. Order Domain

Order Namespace (GKE)

order-service
├── Python Pod 1
├── Python Pod 2
└── Python Pod 3

Use:

• Deployment
• HPA
• ClusterIP Service

NFR Coverage

NFR	Implementation
Availability	Multi-pod
Scalability	Auto scaling
Reliability	Pod failover
Maintainability	Independent deployments

---

6. Database Layer

Cloud SQL for PostgreSQL

Product Database

Primary
 ├─ Read Replica 1
 └─ Read Replica 2

Order Database

Primary
 ├─ Read Replica 1
 └─ Read Replica 2

Recommended:

• Regional deployment
• Automatic backups
• PITR (Point-in-Time Recovery)

NFR Coverage

NFR	Implementation
Availability	HA Cloud SQL
Reliability	Replication
Recoverability	PITR
Performance	Read replicas

---

7. Logging Architecture

Instead of ELK:

Native GCP

Pods
  │
  ▼
Cloud Logging
  │
  ▼
Log Analytics
(BigQuery)

Optional:

Elastic Cloud
on Google Cloud

NFR Coverage

NFR	Implementation
Auditability	Central logs
Observability	Managed logging
Maintainability	Log retention

---

8. Monitoring Architecture

Cloud Monitoring

Equivalent to Prometheus + Grafana.

GKE Metrics
Cloud SQL Metrics
Load Balancer Metrics
Application Metrics
        │
        ▼
Cloud Monitoring
        │
        ▼
Alert Policies

Optional:

• Managed Prometheus
• Managed Grafana

NFR Coverage

NFR	Implementation
Availability	Alerts
Performance	Metrics
Reliability	SLO monitoring

---

9. CI/CD Pipeline

Recommended GCP Pipeline

Developer
   │
   ▼
GitHub/GitLab
   │
   ▼
Cloud Build
   │
   ▼
Security Scan
(Container Analysis)
   │
   ▼
Artifact Registry
   │
   ▼
Helm Chart
   │
   ▼
GKE Deployment

Alternative:

GitHub
   │
Jenkins
   │
Artifact Registry
   │
Helm
   │
GKE

NFR Coverage

NFR	Implementation
Deployability	Automated pipelines
Security	Image scanning
Recoverability	Rollbacks
Maintainability	GitOps support

---

10. User Center (UC)

Option 1 (Recommended)

Google Cloud Identity Platform

Features:

• OAuth2
• OIDC
• MFA
• SSO
• Social Login

Option 2

Keycloak deployed on GKE.

NFR Coverage

NFR	Implementation
Security	IAM
Compliance	Audit logs
Usability	SSO
Maintainability	Central identity

---

Target Enterprise GCP Architecture

Users
  │
  ▼
Cloud CDN
  │
  ▼
Cloud Load Balancer
  │
  ▼
Cloud Armor
  │
  ▼
API Gateway / Apigee
  │
  ▼
GKE Cluster
 ├── Product UI Pods (2)
 ├── Order UI Pods (2)
 ├── Product Service Java Pods (3)
 └── Order Service Python Pods (3)
  │
  ▼
Cloud SQL PostgreSQL
(Product DB + Replicas)

Cloud SQL PostgreSQL
(Order DB + Replicas)

Logging:
Cloud Logging

Monitoring:
Cloud Monitoring
Managed Prometheus
Managed Grafana

CI/CD:
GitHub → Cloud Build → Artifact Registry → Helm → GKE

Identity:
Identity Platform / Keycloak

This GCP version removes most self-managed infrastructure (F5, HAProxy, ELK servers, monitoring servers) and replaces them with managed Google Cloud services, improving operational efficiency while supporting enterprise targets such as 99.95% availability, auto-scaling, disaster recovery, centralized security, and full observability.