Setup Nginx as Load Balancer on Ubuntu

HOW TO SETUP NGINX LOAD BALANCING

LOAD BALANCING : Load Balancing is an efficient mechanism which improves the distribution of workloads or traffic across several capable Virtual Private Server.Using multiple components with load balancing instead of a single component may increase reliability and availability through redundancy.The Round Robin Algorithm will sequentially distribute requests across the group of servers and easy to implement.

SETUP : To setup Nginx Load Balancing    

You must have root privilges on your VPS.
Nginx must be installed on your VPS.

By using "apt-get", you can easily install it.

sudo apt-get install nginx

UPSTREAM MODULE : In order to setup Round Robin Load Balancer,we have to use Nginx upstream module and include configuration into the Nginx setting.

sudo nano /etc/nginx/sites-available/default

Now, add the load balancing configuration to the file.

1. Include the Upstream Module

upstream xyz  
{
     server xyz1.example.com;
     server xyz2.example.com;
     server xyz3.example.com;
}

2. Now, reference this module in the configuration. request from http://xyz would be distributed to xyz1.example.com, xyz2.example.com and xyz3.example.com

server
 {
     location / 
     {
        proxy_pass  http://xyz;
     }
}

Now, restart nginx with:  

sudo service nginx restart

Several directives are used to direct site visitors more effectively to equally distribute load across virtual servers.

WEIGHT : Nginx allows us to assign weight to machines in number specifying the proportion of traffic that should be directed to each server.

upstream xyz
{
          server xyz1.example.com weight=1;
          server xyz2.example.com weight=2;
          server xyz3.example.com weight=4;
 }

If a server is known to be inactive, it should be marked as down like 

server xyz3.example.com.com  down;

MAX FAILS : There are two factors associated with the max fails: max_fails and fall_timeout.

Max_fails refers to the maximum number of failed attempts to connect to a server should occur before it is considered inactive. 
Fall_timeout specifies the length of that the server is considered inoperative. 

Now, the configuration look like :
Once the time expires, new attempts to reach the server will start up again.
The default timeout value is 10 seconds.      

upstream xyz
{
       server xyz1.example.com.com max_fails=3  fail_timeout=15s;
       server xyz2.example.com weight=2;
       server xyz3.example.com weight=4;      
}

This is all about using Nginx as a load balancer. Any doubts or suggestions are most welcome in comments.