为什么非集群应用程序在Node.js中表现优于集群应用程序？

Question

Many people said cluster module makes node.js app more faster than regular one. But what i experienced here is somehow confusing. I create two scripts, first is the regular http server without using cluster module. Second script is http server using cluster module.

I am using apache benchmark to send high request to these servers. And here are the results:

Attempt#1:
non-cluster: 15,418 req/sec
cluster: 10,333 req/sec

Attempt#2:
non-cluster: 12,563 req/sec
cluster: 9,874 req/sec

how can non-cluster script outperform cluster script?

here is the github repo for the scripts.

non-cluster script:

const http = require('http');

const server = http.createServer((req, res) => {
  res.end('Hello World!');
});

server.listen(process.env.SERVER_PORT, () => {
  console.log('Server started on port ' + process.env.SERVER_PORT);
});

cluster script:

const cluster = require('cluster');
const http = require('http');
const numCPUs = require('os').cpus().length;

if (cluster.isMaster) {
  console.log(`Master ${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 {
  // Workers can share any TCP connection
  // In this case it is an HTTP server
  http.createServer((req, res) => {
    res.writeHead(200);
    res.end('Hello World!');
  }).listen(process.env.SERVER_PORT);

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

My node version is 7.10 on Ubuntu 16.04.

Answer 1

I try to set the scheduling policy to SCHED_NONE on cluster script

cluster.schedulingPolicy = cluster.SCHED_NONE

Now cluster script constantly outperforms the non-cluster script. yeeah