Re: One web server, multiple domains

On Tue, 28 Oct 2008 09:26:27 -0400, Lew Pitcher <lpitcher@xxxxxxxxxxxx> wrote:

On October 28, 2008 03:23, in comp.os.linux.networking, sb5309@xxxxxxxxx
(sb5309@xxxxxxxxx) wrote:

A web server typically hosts multiple domains, eg "abcd.com",
"123456.com" etc.

When I ping "abcd.com" or "123456.com", it returns a single IP
address.

What mechanism is used to identify which domain to serve ?

Domain identification is a part of the HTTP protocol that the client
application (the web browser) populates.

When the client (web browser) "clicks" on a hyperlink, the browser will
disect the hyperlink to determine the name of the server. For instance, if
the hyperlink were http://www.yahoo.com/index.html, then the brower would
first determine that the server to contact is www.yahoo.com.

The client then opens a communications path to that server, and asks for the
web page using the entire name from the url. In my yahoo example, the
browser would connect to TCP port 80 of www.yahoo.com and ask for the
www.yahoo.com/index.html webpage.

Now, there may be more than one domain name that points to the machine that
www.yahoo.com points to. So, the web server on that machine looks at the
text of the request ("GET www.yahoo.com/index.html") and determines that it
has to retrieve the "index.html" page from the "www.yahoo.com" profile.

That's it. It's done by the web server, in response to details sent by the
web browser.

Just some supporting evidence from the web-server log:
2008-10-29 00:13:53 +1100 scatter.mine.nu 208.36.144.9 "GET /robots.txt HTTP/1.0
2008-10-29 00:15:56 +1100 bugsplatter.id.au 74.6.8.122 "GET /kernel/boxen/deltre
2008-10-29 00:15:58 +1100 bugsplatter.id.au 74.6.8.122 "GET /kernel/boxen/deltre
2008-10-29 00:23:29 +1100 bugsplatter.id.au 65.55.108.211 "GET /robots.txt HTTP/
2008-10-29 00:23:30 +1100 bugsplatter.id.au 65.55.108.211 "GET /kernel/hwmon/adm
2008-10-29 00:44:29 +1100 bugsplatter.id.au 67.195.37.169 "GET / HTTP/1.0" 200 4
2008-10-29 00:44:44 +1100 bugsplatter.id.au 67.195.37.169 "GET /bugs.css HTTP/1.
2008-10-29 00:48:47 +1100 bugsplatter.mine.nu 66.249.65.46 "GET /test/boxen/semp
2008-10-29 00:53:37 +1100 bugsplatter.id.au 65.55.108.212 "GET /netdraw/netdrawm
2008-10-29 01:05:22 +1100 bugsplatter.mine.nu 66.249.65.46 "GET /test/boxen/semp

There's three domain names in there ;)

Grant.
--
http://bugsplatter.id.au
.

Relevant Pages

  • Re: Installing a JRE plug-in
    ... You are correct that the JSP server runs Java. ... Sorry for not being explicit about client versus server. ... referring to plugins in a client side web browser. ...
    (Debian-User)
  • Re: How to allow a webbrowser to restart a session after it has expired?
    ... If the form is a server form, the URL that the form is posted to will be the same as the original URL, so it will have the station id in the querystring. ... imbedded web browser on the main form. ... I can use the Session[] ...
    (microsoft.public.dotnet.framework.aspnet)
  • Re: How to allow a webbrowser to restart a session after it has expired?
    ... production server running Windows Server 2003, ... imbedded web browser on the main form. ... The asp.net app reads the station id from the request to start the ... I can use the Session[] ...
    (microsoft.public.dotnet.framework.aspnet)
  • How to allow a webbrowser to restart a session after it has expired?
    ... production server running Windows Server 2003, ... imbedded web browser on the main form. ... The asp.net app reads the station id from the request to start the ... I can use the Session[] ...
    (microsoft.public.dotnet.framework.aspnet)
  • Re: anybody found an av that can remove xml exploit virus?
    ... a policy enforcer and which runs the web browser within ... mitigate the effectiveness of the malware payload of the exploit. ... Again, this is a client-side exploit in the web browser, not something ... to fix on the server other than to remedy the corrupted web pages. ...
    (alt.comp.anti-virus)