Understanding Nginx Server Block

In this tutorial, we will discuss about Nginx server block and how the nginx decides which server block will handle the request.

Nginx Server Block Structure

This is what the Nginx server block looks like:

server {
    listen 80;
    server_name example.com;

    . . .

    location / {

        . . .

    }

}

There many directives you can put inside a server block, but there are 2 main directives that we will discuss here. These 2 directives are:

1. listen directive, it will be the address and port where the Nginx will listen to.
2. server_name directive, it will be like your host domain.

These directives decides which server block will Handle the incoming request.

How Nginx Decide Which Server Block is Chosen

This is step by step how the Nginx decide which server block will be choosen:

1. First Nginx will see to which ip and port the request comes. Nginx will match the ip and port with listen directive.
2. If there are many server block with the same listen directives, Nginx then see the host header from the request and then match them with server_name directive.
3. If there are many server block with the same server_name directive that match the host header, Nginx will choose the first server block with that name.
4. If there are no server_name directive match the host header, the nginx will choose the server block who has default_server parameter.

Nginx Server Block Usage Example

To have better understanding of how the Nginx evaluate and choose the server block, we will provide some example for you.

First Example

Let’s assume that our server has *.example.com domain, resolvable by DNS, pointing to our server, and this is the configuration of Nginx in our server.

server {
    listen 127.0.0.1:80;
    . . .
}

server {
    listen 80;
    . . .
}

If we have request from a client to http://example.com, the second server block will be chosen. If we write the listen directive without the adress, then the Nginx will not bind to specific address and can serve from any interfaces. The first server block cannot serve the request because it bind to localhost and can only serve a request from localhost.

Second Example

Let’s assume that our server has *.example.com domain, resolvable by DNS, pointing to our server, and this is the configuration of Nginx in our server.

server {
    listen 80;
    server_name first.example.com;
    . . .
}

server {
    listen 80;
    server_name second.example.com;
    . . .
}

If we have request from a client to http://second.example.com, second server block will be chosen. Both server block has the same listen directive, so Nginx will match the server_name directive with the host header on the client request.

Third Example

Let’s assume that our server has *.example.com domain, resolvable by DNS, pointing to our server, and this is the configuration of Nginx in our server.

server {
    listen 80;
    server_name first.example.com;
    . . .
}

server {
    listen 80;
    server_name second.example.com;
    . . .
}

server {
    listen 80 default_server;
    . . .
}

If we have request from a client to http://example.com, third server block will be chosen. The Nginx will try to match the server_name directive with the host header on the client request first, but as you can see no server_name matching with the host header. Nginx will then search the default_server parameter and choose the server block.

For more advance configuration with the listen and server_name directive, you can always see the Nginx Documentation here: Module ngx_http_core_module.

Thats it! Hopefully now you understand how Nginx decide which server block will handle the request fro these explanation. Next, I will recommend you to find out how the Nginx choose the location block inside the server block. Please read my other tutorial here: Understanding Nginx Location Block.