nodejs.proposal.md

Clustered Threading cluster_threads module like cluster module that supports sharing (or different ports) a port for listening #48350

What is the problem this feature will solve?

Clustering of threads, basically similar to clustered thread pools with features same as Cluster module.

https://github.com/ganeshkbhat/loadbalancer/blob/4f15c79434797ccaf17aeae6f6a3f4c47f0e5703/nodejs.proposal.md

What is the feature you are proposing to solve the problem?

Clustered threads module: The module Clustered threads module which is basically Clustering of threads will be similar to clustered thread pools with features same as inbuilt Cluster module based on child_process.fork() features with minor differences.

Clustering of threads, basically similar to Cluster module where each thread based on specification can:

1. listen to same port
2. listen to different ports

Background:

nodejs module: cluster cluster

1. The worker processes are spawned using the child_process.fork() method, so that they can communicate with the parent via IPC and pass server handles back and forth.

2. The cluster module supports two methods of distributing incoming connections.

   • The first one (and the default one on all platforms except Windows) is the round-robin approach, where the primary process listens on a port
   • The second approach is where the primary process creates the listen socket and sends it to interested workers.

3. Because server.listen() hands off most of the work to the primary process, there are three cases where the behavior between a normal Node.js process and a cluster worker differs:

   a.) server.listen({fd: 7}) Because the message is passed to the primary, file descriptor 7 in the parent will be listened on, and the handle passed to the worker, rather than listening to the worker's idea of what the number 7 file descriptor references.

   b.) server.listen(handle) Listening on handles explicitly will cause the worker to use the supplied handle, rather than talk to the primary process.

   c.) server.listen(0) Normally, this will cause servers to listen on a random port. However, in a cluster, each worker will receive the same "random" port each time they do listen(0). In essence, the port is random the first time, but predictable thereafter. To listen on a unique port, generate a port number based on the cluster worker ID.

CHILD PROCESS PORT SHARING:

• shares the same port process.same.port.js:

const cluster = require('node:cluster');
const http = require('node:http');
const numCPUs = require('node:os').availableParallelism();
const process = require('node:process');

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

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

  cluster.on('exit', (worker, code, signal) => {
    console.log(`worker ${worker.process.pid} died`);
  });
} else {
  http.createServer((req, res) => {
    res.writeHead(200);
    res.end('hello world\n');
  }).listen(8000);

  console.log(`Worker ${process.pid} started`);
}

• shares a different port process.different.port.js:

const cluster = require('node:cluster');
const http = require('node:http');
const numCPUs = require('node:os').availableParallelism();
const process = require('node:process');

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

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

  cluster.on('exit', (worker, code, signal) => {
    console.log(`worker ${worker.process.pid} died`);
  });
} else {
  http.createServer((req, res) => {
    res.writeHead(200);
    res.end('hello world\n');
  }).listen(process.pid);

  console.log(`Worker ${process.pid} started`);
}

RECOMMENDED AND EXPECTED WORKER THREAD CLUSTER MODULE AND PORT SHARING like cluster module:

• shares the same port:

const cluster_thread = require('node:cluster_thread');
const http = require('node:http');
const numCPUs = require('node:os').availableParallelism();
const process = require('node:process');
let workers = [];

if (cluster_thread.isPrimary) {
  console.log(`Primary ${process.pid} is running`);

  // Fork workers.
  for (let i = 0; i < numCPUs; i++) {
    workers[i] = new cluster_thread.Worker(__filename, {
        workerData: script,
    });
  }

  workers[i].on('exit', (worker, code, signal) => {
   console.log(`worker ${worker.process.pid} died`);
  });
} else {
  http.createServer((req, res) => {
    res.writeHead(200);
    res.end('hello world\n');
  }).listen(8000);

  console.log(`Worker ${process.pid}, ${cluster_thread.worker.threadId} started`);
}

• shares a different port:

const cluster_thread = require('node:cluster_thread');
const http = require('node:http');
const numCPUs = require('node:os').availableParallelism();
const process = require('node:process');
let workers = [];

if (cluster_thread.isMainThread) {
  console.log(`Primary ${process.pid} is running`);

  // Fork workers.
  for (let i = 0; i < numCPUs; i++) {
    workers[i] = new cluster_thread.Worker(__filename, {
        workerData: script,
    });
  }

  workers[i].on('exit', (worker, code, signal) => {
    console.log(`worker ${worker.process.pid} died`);
  });
} else {
  http.createServer((req, res) => {
    res.writeHead(200);
    res.end('hello world\n');
  }).listen(cluster_thread.worker.threadId);

  console.log(`Worker ${process.pid}, ${cluster_thread.worker.threadId} started`);
}

What alternatives have you considered?

nodejs module: cluster cluster basd on child_process will be more resource intensive than worker_threads based cluster_threads module