Deployment Diagram

Status: Complete | Last Updated: 2025-11-20 | Version: 1.0.0

Back to: Diagrams Index | Architecture Index

Overview

The deployment diagram shows how Smart Supply Pro is built, containerized, and deployed to cloud infrastructure, including the CI/CD pipeline, Docker workflow, and production topology.

Build & Deployment Pipeline

graph LR Dev["👨‍💻 Developer
Workstation"] Git["📦 GitHub
Repository
main branch"] CI["⚙️ GitHub Actions
CI/CD Pipeline"] Build["🔨 Maven Build
Compile
Test
Package JAR"] Docker["🐳 Docker
Build Multi-stage
Image"] Registry["📋 Container
Registry
(ghcr.io)"] Fly["☁️ Fly.io
Cloud Platform"] subgraph Production["Production Environment"] App["🚀 Running App
(Spring Boot)
Health Checks"] Nginx["🌐 Nginx
Reverse Proxy
TLS/HTTPS
Static Files"] Regions["🌍 Regions
(iad, ord, lhr)"] end Oracle[("💾 Oracle
Autonomous DB
Wallet Auth")] Dev -->|git push| Git Git -->|Webhook trigger| CI CI -->|Checkout| Git CI -->|Build| Build Build -->|Docker build| Docker Docker -->|Push| Registry Registry -->|Deploy| Fly Fly -->|Run| Nginx Nginx -->|Route| App App -->|SQL| Oracle Fly -->|Distribute| Regions style Dev fill:#e3f2fd style Git fill:#bbdefb style CI fill:#90caf9 style Build fill:#64b5f6 style Docker fill:#42a5f5 style Registry fill:#2196f3 style Fly fill:#1976d2 style Production fill:#e8f5e9 style App fill:#81c784 style Nginx fill:#66bb6a style Oracle fill:#1b5e20

Detailed Component Description

Development & Source Control

Developer Workstation - Local development environment - Runs tests and builds locally - Commits code to Git repository - Pushes changes to GitHub main branch

GitHub Repository - Source code repository - Includes: Source code, Dockerfile, pom.xml, configuration files - Main branch triggers automated CI/CD pipeline on push - Repository secrets: Docker credentials, deployment tokens

CI/CD Pipeline (GitHub Actions)

1-ci-test.yml - Triggered on every push to main branch - Steps: 1. Checkout code 2. Set up Maven (Java 17) 3. Run Maven build (compile, test, package) 4. Run security scanning 5. Build Docker image 6. Push image to ghcr.io (GitHub Container Registry)

docs-pipeline.yml - Generates API documentation - Runs Redocly for OpenAPI documentation - Generates architecture diagrams (Mermaid) - Publishes documentation to GitHub Pages

2-deploy-ghpages.yml - Deploys API docs and architecture docs to GitHub Pages - Makes documentation publicly accessible

Build Artifact

Maven Build Output - Compiles Java source code - Runs unit and integration tests - Creates JAR file: inventory-service-0.0.1-SNAPSHOT.jar - JAR includes: Classes, dependencies, embedded Tomcat

Docker Containerization

Multi-Stage Dockerfile - Stage 1 (Builder): Compiles Java code, runs tests, creates JAR - Stage 2 (Runtime): Uses minimal base image, copies JAR, configures entry point - Result: Optimized container image with only runtime dependencies - Image size: ~500MB (JDK 17 + Spring Boot + dependencies)

Docker Image Features - Base image: eclipse-temurin:17-jre-alpine (lightweight) - Entry point: java -jar application.jar - Exposed ports: 8080 (application), 8081 (health checks) - Environment variables: Database credentials, OAuth2 keys (injected at runtime)

Container Registry

GitHub Container Registry (ghcr.io) - Stores built Docker images - Images tagged with Git commit SHA - Also tagged as “latest” - Authentication via GitHub credentials - Integrated with GitHub repository

Fly.io Deployment

fly.toml Configuration - App name: smart-supply-pro (or similar) - Organization and region settings - Machine resources: CPU, memory (tuned for constraints) - Port configuration: 8080 for application - Environment variables and secrets management - Health checks: Regular polling of /actuator/health

Deployment Strategy - Type: Rolling deployment (zero-downtime updates) - New versions gradually replace old instances - Health checks ensure instance is ready before routing traffic

Regions & Scaling - Primary regions: iad (Northern Virginia), ord (Chicago), lhr (London) - Auto-scaling based on CPU/memory - Multiple instances for high availability - Load balancing across regions

Nginx Reverse Proxy - Handles TLS/HTTPS termination - Routes requests to Spring Boot application - Serves static frontend files (if applicable) - Compression (GZIP) for response optimization - Forwards headers properly for authentication

Application Runtime

Spring Boot Application - Runs on embedded Tomcat (port 8080) - Configures itself via environment variables - Establishes database connection pool - Initializes security context

Health Checks - Fly.io pings /actuator/health regularly - Application responds with UP/DOWN status - Failed checks trigger instance restart - Prevents stuck instances from receiving traffic

Database Connection

Oracle Autonomous Database - Wallet-based authentication (secure credentials) - JDBC URL configuration: jdbc:oracle:thin:@... - Connection pooling: HikariCP (optimized pool size) - TLS encryption for network traffic - Automatic backups and high availability

Environment Progression

Development → GitHub → CI Pipeline → Container Registry → Fly.io Production

  1. Developer commits to main branch locally
  2. Push to GitHub triggers CI pipeline
  3. Pipeline builds, tests, and packages application
  4. Docker image created and pushed to registry
  5. Fly.io detects new image and deploys
  6. Rolling deployment gradually updates instances
  7. Health checks confirm new version is running

Key Features

  • Fully Automated: Push code → everything else happens automatically
  • Zero-Downtime Deployments: Rolling updates ensure no service interruption
  • Security: Docker image scanning, secret management via Fly.io
  • High Availability: Multiple regions and instances
  • Observability: Health checks, logs, and metrics
  • Scalability: Horizontal scaling via Fly.io machines