Centralised session info to remove dependency of stickiness!

You want to:

• report a bug/enhancement

Current behaviour

When node clusters are used, it is required to maintain stickiness explicitly even though shared adapter is used socket.io-redis. When client sends multiple requests, if subsequent requests doesn’t go to the same node worker where initial request was made, socket.io throws an error: Session ID unknown. How about caching this session information in a centralised place (May be in a redis adapter itself) and make it work from any worker ?

Currently there are implementations available to make connections go to the same worker based of Client’s IP Address hashing. But these have to be managed explicitly and also might not help load balancing when users share common IP address. (Ex: An internal app being used within a same network).

[johnmela]

I am also working through this problem and would prefer not to use sticky sessions. I am curious to know what the synchronization issues were pre 1.0.x that led to the deprecation of shared session support?

[darrachequesne]

For future readers:

We have added some explanation here: https://socket.io/docs/v4/using-multiple-nodes/#Why-is-sticky-session-required

Within a Node.js cluster, you can now use the cluster adapter and the @socket.io/sticky package:

const cluster = require("cluster");
const http = require("http");
const { Server } = require("socket.io");
const numCPUs = require("os").cpus().length;
const { setupMaster, setupWorker } = require("@socket.io/sticky");
const { createAdapter, setupPrimary } = require("@socket.io/cluster-adapter");

if (cluster.isMaster) {
  console.log(`Master ${process.pid} is running`);

  const httpServer = http.createServer();

  // setup sticky sessions
  setupMaster(httpServer, {
    loadBalancingMethod: "least-connection",
  });

  // setup connections between the workers
  setupPrimary();

  // needed for packets containing buffers (you can ignore it if you only send plaintext objects)
  // Node.js < 16.0.0
  cluster.setupMaster({
    serialization: "advanced",
  });
  // Node.js > 16.0.0
  // cluster.setupPrimary({
  //   serialization: "advanced",
  // });

  httpServer.listen(3000);

  for (let i = 0; i < numCPUs; i++) {
    cluster.fork();
  }

  cluster.on("exit", (worker) => {
    console.log(`Worker ${worker.process.pid} died`);
    cluster.fork();
  });
} else {
  console.log(`Worker ${process.pid} started`);

  const httpServer = http.createServer();
  const io = new Server(httpServer);

  // use the cluster adapter
  io.adapter(createAdapter());

  // setup connection with the primary process
  setupWorker(io);

  io.on("connection", (socket) => {
    /* ... */
  });
}

Documentation:

• https://socket.io/docs/v4/using-multiple-nodes/#Using-Node-js-Cluster
• https://socket.io/docs/v4/cluster-adapter/

Please reopen if needed.