[10065] in North American Network Operators' Group
Re: multihoming without BGP
daemon@ATHENA.MIT.EDU (Marc Slemko)
Fri Jun 13 10:12:12 1997
Date: Fri, 13 Jun 1997 08:09:35 -0600 (MDT)
From: Marc Slemko <marcs@znep.com>
To: Dana Hudes <dhudes@netra.graphnet.com>
cc: nanog@merit.edu
In-Reply-To: <33A14A43.E8ABC158@graphnet.com>
On Fri, 13 Jun 1997, Dana Hudes wrote:
> Paul A Vixie wrote:
> >
> >
> > But there's no guarantee that you need separate addresses per home
> > page.
> > If you don't count Lynx or Mosaic as part of your target audience,
> > then you
> > can depend on the "Host:" keyword sent in queries by *all* modern
> > browsers.
>
>
> I had to jump in here with a correction and clarification. We
> are using Netscape Enterprise Server 2.0. The 'software virtual
> server' feature doesn't work with any version of Netscape Navigator
> up to and including Communicator Preview 5. The server sees a
> HTTP 1.0 not HTTP 1.1 request. In my logs I do see some webcrawlers
> sending HTTP 1.1 so it isn't completely a server issue.
> Netscape Enterprise 3.0 is coming out RSN but I don't know that it will
> be any better. If it is I'll speak up here.
This does not belong on nanog, but to set the facts straight: if Netscape
Enterprise 2.0 supports Host: based virtual hosts but only for hosts that
make HTTP/1.1 requests that is lame. Versions of Navigator have sent
Host: headers for some time; so has MSIE. Last I knew, AOL's proxies
didn't yet because they were too lazy. The vast majority of clients do
send Host: headers even though they aren't HTTP/1.1. HTTP/1.1 clients
_have_ to send the Host: header, but HTTP/1.0 clients may and most do.
I do not advocate using them yet for most websites, but that is another
issue.
