Web sites associated with the same IP address - http

Question

How does a web server recognize which URL to serve when there are multiple web sites (hostnames) associated with the same IP address?

Answer 1

With the Apache web server, a set of virtualhosts is defined that contain parameters to match a request to a directory. Using a basic, default set up for an Ubuntu web server, you would have a file in the /etc/apache2/sites-enabled directory like this:
<VirtualHost>
ServerName example.com
DocumentRoot /var/www/example.com/httpdocs
</VirtualHost>
This tells apache that all requests to example.com that arrive at this machine should be routed to the /var/www/example.com/httpdocs folder. Another entry to example.org could point it to a different folder.
Also relevant is the /etc/hosts file and the apache a2ensite command.

Answer 2

This is done in apache by using NameVirtualHost
You first declare what IP and port in httpd.conf to use like:
NameVirtualHost 127.0.0.1:80
Then in your virtual host block, you do:
<VirtualHost 127.0.0.1:80>
ServerName your_domain
DocumentRoot path_to_your_app
....
</VirtualHost>
This will allow you to have multiple hosts on one IP. But be warned that if you access the IP directly, it will direct the request to the first virtual host.

Answer 3

Through the use of virtual hosts.
This is an example configuration from nginx
http {
index index.html;
server {
server_name www.domain1.com;
access_log logs/domain1.access.log main;
root /var/www/domain1.com/htdocs;
}
server {
server_name www.domain2.com;
access_log logs/domain2.access.log main;
root /var/www/domain2.com/htdocs;
}
}
Essentially, when a user requests a resource the server checks the host field of the request and responds accordingly.

Answer 4

HTTP 1.1 defines a header called the "Host" header.
Before Apache or any other server sees the request, the client browser creates the http 1.1 request headers and sends the request to the server you are asking for your browser to contact.
Once the request hits apache the server looks at the Host header portion of the HTTP request headers. You can observe this by using wireshark, liveHttpHeaders, HTTPfox or whatever http dissecting/packet capturing tool you like. The host header in HTTP 1.1 (Host: header is not defined for HTTP 1.0 or 0.9) is formatted as follows:
Host: www.example.com\r\n
When apache looks at this header it parses it and goes through the existing VirtualHosts table that is used for mapping matching host headers to directories or actions defined.
That is to say if you had a NameVirtualHost for www.example.com that points to /some/path/example.com/
NameVirtualhost stuff here
<VirtualHost 127.0.0.1:80>
ServerName www.example.com
DocumentRoot /some/path/example.com
....
</VirtualHost>
your apache would take the following request:
GET /index.html HTTP/1.1\r\n <-- version is a key part
Connection: close\r\n
Host: www.example.com\r\n <-- key part
Accept: blah\r\n
Another: blah\r\n
read the wiki page for more on header format.
Apache would see that the host header contains www.example.com and serve up the file
/some/path/example.com/index.html because that directory and filename matches the requested resource and it is the directory that is to be used for serving requests with the host header containing www.example.com.
That is how it works.

Answer 5

Depends on the type of server. Apache uses a .htaccess file and you could also configure virtual hosts. If you're trying to do something specific, you may want to edit your question to include exactly what you're looking for and what software you're using to host.