WebSocket in Java: Advanced Real-Time Applications

WebSocket in Java: Creating Advanced Real-Time Applications

Modern web applications increasingly require instant data transfer between client and server. Traditional HTTP, based on request-response, is inefficient for tasks where latency is critical: online chats, stock quotes, collaborative document editing, or game servers. The WebSocket protocol solves this problem by establishing a persistent bidirectional connection. In this article, we will explore advanced techniques for working with WebSocket in Java, going beyond basic examples.

1. Architecture of an Advanced WebSocket Application in Java

A basic WebSocket example in Java using the @ServerEndpoint annotation is good for learning, but production solutions require a more robust architecture. Let's look at the key components.

1.1. Choosing the Technology Stack

For enterprise applications, the de facto standard is Spring Boot with the spring-boot-starter-websocket module. It provides integration with STOMP (Simple Text Oriented Messaging Protocol) — a sub-protocol over WebSocket that adds message routing, queues, and subscriptions.

1.2. Basic WebSocket Configuration in Spring Boot

Let's start with a configuration class:

@Configuration@EnableWebSocketMessageBrokerpublic class WebSocketConfig implements WebSocketMessageBrokerConfigurer {

    @Override    public void configureMessageBroker(MessageBrokerRegistry config) {        // Built-in broker for /topic and /queue destinations        config.enableSimpleBroker("/topic", "/queue");        // Prefix for messages from client to server        config.setApplicationDestinationPrefixes("/app");        // Prefix for user-specific queues (point-to-point)        config.setUserDestinationPrefix("/user");    }

    @Override    public void registerStompEndpoints(StompEndpointRegistry registry) {        // Endpoint for connection via SockJS (fallback for older browsers)        registry.addEndpoint("/ws")                .setAllowedOrigins("*")                .withSockJS();    }}

This code configures connection endpoints, routing prefixes, and a built-in message broker. Using STOMP significantly simplifies working with subscriptions and queues.

2. Advanced Message Processing: From Chat to Stock Market Data

Now let's move on to implementing specific scenarios. Let's consider three typical cases: group chat, personal notifications, and data streaming.

2.1. Group Chat with Message History

A simple chat is broadcasting messages to a topic. Let's complicate the task: add message history saving and filtering by rooms.

@Controllerpublic class ChatController {

    @Autowired    private ChatMessageRepository repository; // JPA repository

    @MessageMapping("/chat.sendMessage")    @SendTo("/topic/chat/{roomId}")    public ChatMessage sendMessage(@Payload ChatMessage message,                                   @DestinationVariable String roomId,                                   SimpMessageHeaderAccessor headerAccessor) {        // Save to database        message.setTimestamp(Instant.now());        repository.save(message);        // Add sender info from session        String username = (String) headerAccessor.getSessionAttributes().get("username");        message.setSender(username);        return message;    }

    @MessageMapping("/chat.addUser")    @SendTo("/topic/chat/{roomId}")    public ChatMessage addUser(@Payload ChatMessage message,                               @DestinationVariable String roomId,                               SimpMessageHeaderAccessor headerAccessor) {        headerAccessor.getSessionAttributes().put("username", message.getSender());        return message;    }}

Here we use