Question

I've set up Node.js and Nginx on my server. Now I want to use it, but, before I start there are 2 questions:

  1. How should they work together? How should I handle the requests?
  2. There are 2 concepts for a Node.js server, which one is better:

    a. Create a separate HTTP server for each website that needs it. Then load all JavaScript code at the start of the program, so the code is interpreted once.

    b. Create one single Node.js server which handles all Node.js requests. This reads the requested files and evals their contents. So the files are interpreted on each request, but the server logic is much simpler.

It's not clear for me how to use Node.js correctly.

Was it helpful?

Solution

Nginx works as a front end server, which in this case proxies the requests to a node.js server. Therefore you need to setup an nginx config file for node.

This is what I have done in my Ubuntu box:

Create the file yourdomain.com at /etc/nginx/sites-available/:

vim /etc/nginx/sites-available/yourdomain.com

In it you should have something like:

# the IP(s) on which your node server is running. I chose port 3000.
upstream app_yourdomain {
    server 127.0.0.1:3000;
    keepalive 8;
}

# the nginx server instance
server {
    listen 80;
    listen [::]:80;
    server_name yourdomain.com www.yourdomain.com;
    access_log /var/log/nginx/yourdomain.com.log;

    # pass the request to the node.js server with the correct headers
    # and much more can be added, see nginx config options
    location / {
      proxy_set_header X-Real-IP $remote_addr;
      proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
      proxy_set_header Host $http_host;
      proxy_set_header X-NginX-Proxy true;

      proxy_pass http://app_yourdomain/;
      proxy_redirect off;
    }
 }

If you want nginx (>= 1.3.13) to handle websocket requests as well, add the following lines in the location / section:

proxy_http_version 1.1;
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";

Once you have this setup you must enable the site defined in the config file above:

cd /etc/nginx/sites-enabled/ 
ln -s /etc/nginx/sites-available/yourdomain.com yourdomain.com

Create your node server app at /var/www/yourdomain/app.js and run it at localhost:3000

var http = require('http');

http.createServer(function (req, res) {
    res.writeHead(200, {'Content-Type': 'text/plain'});
    res.end('Hello World\n');
}).listen(3000, "127.0.0.1");
console.log('Server running at http://127.0.0.1:3000/');

Test for syntax mistakes:

nginx -t

Restart nginx:

sudo /etc/init.d/nginx restart

Lastly start the node server:

cd /var/www/yourdomain/ && node app.js

Now you should see "Hello World" at yourdomain.com

One last note with regards to starting the node server: you should use some kind of monitoring system for the node daemon. There is an awesome tutorial on node with upstart and monit.

OTHER TIPS

You can also setup multiple domain with nginx, forwarding to multiple node.js processes.

For example to achieve these:

These ports (4000 and 5000) should be used to listen the app requests in your app code.

/etc/nginx/sites-enabled/domain1

server {
    listen 80;
    listen [::]:80;
    server_name domain1.com;
    access_log /var/log/nginx/domain1.access.log;
    location / {
        proxy_pass    http://127.0.0.1:4000/;
    }
}

In /etc/nginx/sites-enabled/domain2

server {
    listen 80;
    listen [::]:80;
    server_name domain2.com;
    access_log /var/log/nginx/domain2.access.log;
    location / {
        proxy_pass    http://127.0.0.1:5000/;
    }
}

You can also have different urls for apps in one server configuration:

In /etc/nginx/sites-enabled/yourdomain:

server {
    listen 80;
    listen [::]:80;
    server_name yourdomain.com;

    location ^~ /app1/{
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header Host $http_host;
        proxy_set_header X-NginX-Proxy true;
        proxy_pass    http://127.0.0.1:3000/;
    }

    location ^~ /app2/{
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header Host $http_host;
        proxy_set_header X-NginX-Proxy true;
        proxy_pass    http://127.0.0.1:4000/;
    }
}

Restart nginx:

sudo service nginx restart

Starting applications.

node app1.js

var http = require('http');
http.createServer(function (req, res) {
    res.writeHead(200, {'Content-Type': 'text/plain'});
    res.end('Hello from app1!\n');
}).listen(3000, "127.0.0.1");
console.log('Server running at http://127.0.0.1:3000/');

node app2.js

var http = require('http');
http.createServer(function (req, res) {
    res.writeHead(200, {'Content-Type': 'text/plain'});
    res.end('Hello from app2!\n');
}).listen(4000, "127.0.0.1");
console.log('Server running at http://127.0.0.1:4000/');

I proxy independent Node Express applications through Nginx.

Thus new applications can be easily mounted and I can also run other stuff on the same server at different locations.

Here are more details on my setup with Nginx configuration example:

Deploy multiple Node applications on one web server in subfolders with Nginx

Things get tricky with Node when you need to move your application from from localhost to the internet.

There is no common approach for Node deployment.

Google can find tons of articles on this topic, but I was struggling to find the proper solution for the setup I need.

Basically, I have a web server and I want Node applications to be mounted to subfolders (i.e. http://myhost/demo/pet-project/) without introducing any configuration dependency to the application code.

At the same time I want other stuff like blog to run on the same web server.

Sounds simple huh? Apparently not.

In many examples on the web Node applications either run on port 80 or proxied by Nginx to the root.

Even though both approaches are valid for certain use cases, they do not meet my simple yet a little bit exotic criteria.

That is why I created my own Nginx configuration and here is an extract:

upstream pet_project {
  server localhost:3000;
}

server {
  listen 80;
  listen [::]:80;
  server_name frontend;

  location /demo/pet-project {
    alias /opt/demo/pet-project/public/;
    try_files $uri $uri/ @pet-project;
  }

  location @pet-project {
    rewrite /demo/pet-project(.*) $1 break;

    proxy_set_header X-Real-IP $remote_addr;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_set_header Host $proxy_host;
    proxy_set_header X-NginX-Proxy true;

    proxy_pass http://pet_project;
    proxy_redirect http://pet_project/ /demo/pet-project/;
  }
}

From this example you can notice that I mount my Pet Project Node application running on port 3000 to http://myhost/demo/pet-project.

First Nginx checks if whether the requested resource is a static file available at /opt/demo/pet-project/public/ and if so it serves it as is that is highly efficient, so we do not need to have a redundant layer like Connect static middleware.

Then all other requests are overwritten and proxied to Pet Project Node application, so the Node application does not need to know where it is actually mounted and thus can be moved anywhere purely by configuration.

proxy_redirect is a must to handle Location header properly. This is extremely important if you use res.redirect() in your Node application.

You can easily replicate this setup for multiple Node applications running on different ports and add more location handlers for other purposes.

From: http://skovalyov.blogspot.dk/2012/07/deploy-multiple-node-applications-on.html

Node.js with Nginx configuration.

$ sudo nano /etc/nginx/sites-available/subdomain.your_domain.com

add the following configuration so that Nginx acting as a proxy redirect to port 3000 traffic from the server when we come from “subdomain.your_domain.com”

upstream subdomain.your_domain.com {
  server 127.0.0.1:3000;
}
server {
  listen 80;
  listen [::]:80;
  server_name subdomain.your_domain.com;
  access_log /var/log/nginx/subdomain.your_domain.access.log;
  error_log /var/log/nginx/subdomain.your_domain.error.log debug;
  location / {
    proxy_set_header X-Real-IP $remote_addr;
    proxy_set_header X-Forwarder-For $proxy_add_x_forwarded_for;
    proxy_set_header Host $http_host;
    proxy_set_header X-NginX-Proxy true;
    proxy_pass http://subdomain.your_domain.com;
    proxy_redirect off;
  }
}

answering your question 2:

I would use option b simply because it consumes much less resources. with option 'a', every client will cause the server to consume a lot of memory, loading all the files you need (even though i like php, this is one of the problems with it). With option 'b' you can load your libraries (reusable code) and share them among all client requests.

But be ware that if you have multiple cores you should tweak node.js to use all of them.

I made a repository in Github which you can clone, vagrant-node-nginx-boilerplate

basically the node.js app at /var/www/nodeapp is

var http = require('http');
http.createServer(function (req, res) {
  res.writeHead(200, {'Content-Type': 'text/plain'});
  res.end('Hello World\n');
}).listen(4570, '127.0.0.1');

console.log('Node Server running at 127.0.0.1:4570/');

and the nginx config at /etc/nginx/sites-available/ is

server {
        listen 80 default_server;
        listen [::]:80 default_server;

        root /var/www/nodeapp;
        index index.html index.htm;

        server_name localhost;

        location / {
          proxy_pass http://127.0.0.1:4570;
          proxy_http_version 1.1;
          proxy_set_header Upgrade $http_upgrade;
          proxy_set_header Connection 'upgrade';
          proxy_set_header Host $host;
          proxy_cache_bypass $http_upgrade;
        }
}

You could also use node.js to generate static files into a directory served by nginx. Of course, some dynamic parts of your site could be served by node, and some by nginx (static).

Having some of them served by nginx increases your performance..

We can easily setup a Nodejs app by Nginx acting as a reverse proxy.
The following configuration assumes the NodeJS application is running on 127.0.0.1:8080,

  server{
     server_name domain.com sub.domain.com; # multiple domains

     location /{ 
      proxy_pass http://127.0.0.1:8080;  
      proxy_set_header Host $host;
      proxy_pass_request_headers on;  
     }

     location /static/{
       alias /absolute/path/to/static/files; # nginx will handle js/css
     }
   } 

in above setup your Nodejs app will,

  • get HTTP_HOST header where you can apply domain specific logic to serve the response. '
  • Your Application must be managed by a process manager like pm2 or supervisor for handling situations/reusing sockets or resources etc.

  • Setup an error reporting service for getting production errors like sentry or rollbar

NOTE: you can setup logic for handing domain specific request routes, create a middleware for expressjs application

Nginx can act as a reverse proxy server which works just like a project manager. When it gets a request it analyses it and forwards the request to upstream(project members) or handles itself. Nginx has two ways of handling a request based on how its configured.

  • serve the request
  • forward the request to another server

    server{
     server_name mydomain.com sub.mydomain.com;
    
     location /{ 
      proxy_pass http://127.0.0.1:8000;  
      proxy_set_header Host $host;
      proxy_pass_request_headers on;  
     }
    
     location /static/{
       alias /my/static/files/path;
     }
    

    }

Server the request

With this configuration, when the request url is mydomain.com/static/myjs.js it returns the myjs.js file in /my/static/files/path folder. When you configure nginx to serve static files, it handles the request itself.

forward the request to another server

When the request url is mydomain.com/dothis nginx will forwards the request to http://127.0.0.1:8000. The service which is running on the localhost 8000 port will receive the request and returns the response to nginx and nginx returns the response to the client.

When you run node.js server on the port 8000 nginx will forward the request to node.js. Write node.js logic and handle the request. That's it you have your nodejs server running behind the nginx server.

If you wish to run any other services other than nodejs just run another service like Django, flask, php on different ports and config it in nginx.

You can run nodejs using pm2 if you want to manage each microservice means and run it. Node will be running in a port right just configure that port in nginx(/etc/nginx/sites-enabled/domain.com)

server{
    listen 80;
    server_name domain.com www.domain.com;

  location / {
     return 403;
  }
    location /url {
        proxy_pass http://localhost:51967/info;
    }
}

Check whether localhost is running or not by using ping.

And

Create one single Node.js server which handles all Node.js requests. This reads the requested files and evals their contents. So the files are interpreted on each request, but the server logic is much simpler.

This is best and as you said easier too

Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top