Real-Time Chat System

A scalable real-time chat system using Node.js, WebSocket, Kafka, Redis, and Cassandra.

Setup

Required Secret Files

Before starting the services, you need to create the following secret files:

secrets/
├── jwt_secret.txt        # Secret key for JWT token signing (random string)
├── db_password.txt       # Password for MySQL database
├── db_root_password.txt  # Password for MySQL root user

Example secret file contents:

# jwt_secret.txt
your-256-bit-secret

# db_password.txt
your-db-password

# db_root_password.txt
your-db-root-password

Running the Services

1. Create the required secret files as shown above
2. Start the services:

docker-compose up

3. Run the test client:

cd test-client
npm install
npm start

Test Client Usage

The test client provides a command-line interface for testing the chat system. Here are the available commands:

# Channel Creation
/create MyChannel              # Create a new group channel
/create "Team Chat" user1 user2 user3  # Create channel with participants

# Basic messaging
/join <channelId>     # Join a channel
Hello, world!         # Send a message to current channel
/leave <channelId>    # Leave a channel

# Channel management
/read <channelId>     # Mark a channel as read
/msg <message>        # Explicitly send a message (alternative to just typing)

# Presence
/status <status>      # Update presence status (online, away, busy, etc.)
/quit                 # Exit the client

When you start the client, you'll see a list of your available channels. You can join any channel and start sending messages immediately.

Architecture

The system consists of several components:

• Chat Server: Handles WebSocket connections and real-time message delivery
• API Server: Handles authentication and server discovery
• Message Dispatcher: Routes messages to chat servers
• Presence Service: Tracks user online status
• Service Discovery: Zookeeper for server registration and discovery
• Message Queue: Kafka for reliable message delivery
• Cache: Redis for user to chat server mapping
• Database: Cassandra for persistent storage

Features

• Real-time messaging using WebSockets
• Channel-based communication (group & direct messages)
• Message persistence in Cassandra
• Distributed message delivery via Kafka
• Scalable architecture with multiple chat servers
• Service discovery using Zookeeper
• JWT-based authentication
• User presence tracking
• Message read receipts
• Channel participant management
• Command-line interface for testing
• Automatic channel list updates
• Message batching for better performance

Message Types

Client to Server

• create_channel: Create a new channel
• join_channel: Join an existing channel
• leave_channel: Leave a channel
• chat_message: Send a message to a channel
• mark_channel_read: Mark messages in a channel as read

Server to Client

• channels_list: List of user's channels
• channel_joined: Confirmation of joining a channel
• channel_left: Confirmation of leaving a channel
• chat_message: Incoming message
• error: Error notification