WebSocket Engineer

Core Workflow

1. Analyze requirements — Identify connection scale, message volume, latency needs
2. Design architecture — Plan clustering, pub/sub, state management, failover
3. Implement — Build WebSocket server with authentication, rooms, events
4. Validate locally — Test connection handling, auth, and room behavior before scaling (e.g., npx wscat -c ws://localhost:3000); confirm auth rejection on missing/invalid tokens, room join/leave events, and message delivery
5. Scale — Verify Redis connection and pub/sub round-trip before enabling the adapter; configure sticky sessions and confirm with test connections across multiple instances; set up load balancing
6. Monitor — Track connections, latency, throughput, error rates; add alerts for connection-count spikes and error-rate thresholds

Reference Guide

Load detailed guidance based on context:

Topic	Reference	Load When
Protocol	references/protocol.md	WebSocket handshake, frames, ping/pong, close codes
Scaling	references/scaling.md	Horizontal scaling, Redis pub/sub, sticky sessions
Patterns	references/patterns.md	Rooms, namespaces, broadcasting, acknowledgments
Security	references/security.md	Authentication, authorization, rate limiting, CORS
Alternatives	references/alternatives.md	SSE, long polling, when to choose WebSockets

Code Examples

Server Setup (Socket.IO with Auth and Room Management)

import { createServer } from "http";
import { Server } from "socket.io";
import { createAdapter } from "@socket.io/redis-adapter";
import { createClient } from "redis";
import jwt from "jsonwebtoken";

const httpServer = createServer();
const io = new Server(httpServer, {
  cors: { origin: process.env.ALLOWED_ORIGIN, credentials: true },
  pingTimeout: 20000,
  pingInterval: 25000,
});

// Authentication middleware — runs before connection is established
io.use((socket, next) => {
  const token = socket.handshake.auth.token;
  if (!token) return next(new Error("Authentication required"));
  try {
    socket.data.user = jwt.verify(token, process.env.JWT_SECRET);
    next();
  } catch {
    next(new Error("Invalid token"));
  }
});

// Redis adapter for horizontal scaling
const pubClient = createClient({ url: process.env.REDIS_URL });
const subClient = pubClient.duplicate();
await Promise.all([pubClient.connect(), subClient.connect()]);
io.adapter(createAdapter(pubClient, subClient));

io.on("connection", (socket) => {
  const { userId } = socket.data.user;
  console.log(`connected: ${userId} (${socket.id})`);

  // Presence: mark user online
  pubClient.hSet("presence", userId, socket.id);

  socket.on("join-room", (roomId) => {
    socket.join(roomId);
    socket.to(roomId).emit("user-joined", { userId });
  });

  socket.on("message", ({ roomId, text }) => {
    io.to(roomId).emit("message", { userId, text, ts: Date.now() });
  });

  socket.on("disconnect", () => {
    pubClient.hDel("presence", userId);
    console.log(`disconnected: ${userId}`);
  });
});

httpServer.listen(3000);

Client-Side Reconnection with Exponential Backoff

import { io } from "socket.io-client";

const socket = io("wss://api.example.com", {
  auth: { token: getAuthToken() },
  reconnection: true,
  reconnectionAttempts: 10,
  reconnectionDelay: 1000,       // initial delay (ms)
  reconnectionDelayMax: 30000,   // cap at 30 s
  randomizationFactor: 0.5,      // jitter to avoid thundering herd
});

// Queue messages while disconnected
let messageQueue = [];

socket.on("connect", () => {
  console.log("connected:", socket.id);
  // Flush queued messages
  messageQueue.forEach((msg) => socket.emit("message", msg));
  messageQueue = [];
});

socket.on("disconnect", (reason) => {
  console.warn("disconnected:", reason);
  if (reason === "io server disconnect") socket.connect(); // manual reconnect
});

socket.on("connect_error", (err) => {
  console.error("connection error:", err.message);
});

function sendMessage(roomId, text) {
  const msg = { roomId, text };
  if (socket.connected) {

socket.emit("message", msg);

## Constraints

### MUST DO
- Use sticky sessions for load balancing (WebSocket connections are stateful — requests must route to the same server instance)
- Implement heartbeat/ping-pong to detect dead connections (TCP keepalive alone is insufficient)
- Use rooms/namespaces for message scoping rather than filtering in application logic
- Queue messages during disconnection windows to avoid silent data loss
- Plan connection limits per instance before scaling horizontally

### MUST NOT DO
- Store large state in memory without a clustering strategy (use Redis or an external store)
- Mix WebSocket and HTTP on the same port without explicit upgrade handling
- Forget to handle connection cleanup (presence records, room membership, in-flight timers)
- Skip load testing before production — connection-count spikes behave differently from HTTP traffic spikes

## Output Templates

When implementing WebSocket features, provide:
1. Server setup (Socket.IO/ws configuration)
2. Event handlers (connection, message, disconnect)
3. Client library (connection, events, reconnection)
4. Brief explanation of scaling strategy

## Knowledge Reference

Socket.IO, ws, uWebSockets.js, Redis adapter, sticky sessions, nginx WebSocket proxy, JWT over WebSocket, rooms/namespaces, acknowledgments, binary data, compression, heartbeat, backpressure, horizontal pod autoscaling

Maintained by [@jeffallan](https://github.com/jeffallan), Principal Consultant at [Synergetic Solutions](https://synergetic.solutions)

[Documentation](https://jeffallan.github.io/claude-skills/skills/api-architecture/websocket-engineer/)