Event Driven Notification System (EDNS v1.0.0)

A production-grade event-driven notification system built with Go that demonstrates asynchronous processing, reliable message delivery, real-time notifications, and a scalable microservice architecture.

The system is built around independent services communicating through Redis Streams and Redis Pub/Sub, making it easy to extend with additional notification channels.

---

Table of Contents

• Overview
• Architecture
• Notification Flow
• Features
• Project Structure
• Tech Stack
• Running the Project
• Reliability & Observability
• Future Improvements

---

Overview

EDNS is composed of three independent services:

• API Service
• Worker Service
• WebSocket Service

Notifications are created through the REST API, processed asynchronously by workers, and delivered through multiple notification channels.

---

Architecture

                Client
                  │
                  ▼
            API Service
                  │
        PostgreSQL + Redis Streams
                  │
                  ▼
           Worker Service
           │             │
           ▼             ▼
      SMTP Email   Redis Pub/Sub
                         │
                         ▼
                 WebSocket Service
                         │
                         ▼
                 Connected Clients

---

Notification Flow

1. Client creates a notification through the REST API.
2. API stores notification data in PostgreSQL.
3. Notification ID is published to Redis Streams.
4. Worker consumes the message using Redis Consumer Groups.
5. Worker processes each delivery channel.
6. Email notifications are sent through SMTP.
7. WebSocket notifications are published through Redis Pub/Sub.
8. WebSocket Service pushes notifications to all connected user sessions.
9. Failed deliveries are retried and eventually moved to the Dead Letter Queue (DLQ).

---

Features

Core

• REST API
• PostgreSQL Persistence
• Repository-Service Architecture
• Configuration Management
• Docker & Docker Compose

Event Processing

• Redis Streams
• Redis Consumer Groups
• Reliable Message Acknowledgement
• Automatic Retry Logic
• Dead Letter Queue (DLQ)
• Pending Message Recovery (XAUTOCLAIM)
• Multiple Worker Support

Notification Channels

• Email Notifications (SMTP)
• WebSocket Notifications
• Redis Pub/Sub
• Multiple Active Connections Per User

Observability

• Structured Logging (Zap)
• Health Checks
• Readiness Checks
• Prometheus Metrics

Reliability

• Graceful Shutdown
• Fault Tolerant Workers
• Delivery Status Tracking
• Notification Status Synchronization

---

Project Structure

api-service/
worker-service/
websocket-service/

internal/
├── config/
├── events/
├── logger/
├── models/
├── observability/
├── repository/
├── stream/
└── validation/

configs/
deploy/
migrations/

---

Tech Stack

Language

• Go

Framework

• Gin

Database

• PostgreSQL
• pgx

Messaging

• Redis Streams
• Redis Pub/Sub

Real-Time Communication

• Gorilla WebSocket

Email

• SMTP
• Mailpit

Observability

• Prometheus
• Grafana
• Zap Logger

Containerization

• Docker
• Docker Compose

---

Running the Project

Production

make docker-up-prod

Development

make docker-up-dev

Available services:

Service	Address
API	http://localhost:8080
Worker	http://localhost:8081
WebSocket	ws://localhost:8082/api/ws
Mailpit	http://localhost:8025
Prometheus	http://localhost:9090
Grafana	http://localhost:3001

---

Reliability & Observability

Reliability

• Redis Consumer Groups
• Automatic Retries
• Dead Letter Queue
• Pending Message Recovery
• Graceful Shutdown
• Multiple Worker Support

Observability

• Structured Logging
• Health Endpoint
• Readiness Endpoint
• Prometheus Metrics

---

System Design Concepts

• Event-Driven Architecture
• Producer-Consumer Pattern
• Repository Pattern
• Service Layer
• Distributed Workers
• Redis Streams
• Redis Consumer Groups
• Redis Pub/Sub
• Dead Letter Queue
• Retry Mechanism
• WebSocket Connection Management
• Fault Tolerance
• Horizontal Scalability

---

Future Improvements

• Unit & Integration Tests
• Grafana Dashboards
• OpenTelemetry Tracing
• CI/CD Pipeline
• Kubernetes Deployment
• Authentication & Authorization
• Rate Limiting
• Additional Notification Channels