Nodejs Clustering and expressjs sessions

一笑奈何 提交于 2019-12-18 14:17:05

问题


I'm trying to build nodejs application which will take advantage of multicore machines ( a.k.a. clustering ) and I got a question about sessions. My code looks like this:

var cluster = exports.cluster = require('cluster');
var numCPUs = require('os').cpus().length;

if (cluster.isMaster) {

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

  cluster.on('exit', function(worker, code, signal) {
    console.log('worker ' + worker.process.pid + ' died. Trying to respawn...');
    cluster.fork();
  });

} else {

//spawn express etc

}

My question is: Everytime a single user hits random node instance or for example the first time he opens the page and hits node N4 and till his session expires, he hit node N4 on every request? For those who didn't understand my question, I will try to explain what I'm worried about: A user enters my page, he login on node N3, then I set req.session.userdata to a random data, he refreshes the page and he hit Node N4, will I be able to access req.session.userdata from different Node? That mean there is a chance for the user to get randomly logged out or I'm just not understanding how clustering with express works?


回答1:


You're correct that the in memory session store in Connect/Express is unsuitable for supporting more than one instance. The solution is to implement a session store with a backing database. My recommendation is connect-redis, and example code is at Session Undefined - Using Connect-Redis / ExpressJS / Node

But there are dozens of options.



来源:https://stackoverflow.com/questions/17626279/nodejs-clustering-and-expressjs-sessions

易学教程内所有资源均来自网络或用户发布的内容,如有违反法律规定的内容欢迎反馈
该文章没有解决你所遇到的问题?点击提问,说说你的问题,让更多的人一起探讨吧!