[32404] in North American Network Operators' Group
Re: "...the IPv4 TOS field should be end-to-end...."
daemon@ATHENA.MIT.EDU (Joe Abley)
Tue Nov 21 08:25:29 2000
Date: Tue, 21 Nov 2000 08:25:29 -0500
From: Joe Abley <jabley@automagic.org>
To: JIM FLEMING <jfleming@anet.com>
Cc: Roeland Meyer <rmeyer@mhsc.com>,
"'Shawn McMahon'" <smcmahon@eiv.com>, nanog@merit.edu
Message-ID: <20001121082527.B2383@buddha.automagic.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <068d01c0537c$95e2e920$df00a8c0@ValuedSonyCustomer>; from jfleming@anet.com on Mon, Nov 20, 2000 at 11:33:30PM -0600
Errors-To: owner-nanog-outgoing@merit.edu
On Mon, Nov 20, 2000 at 11:33:30PM -0600, JIM FLEMING wrote:
> In my opinion, the IPv4 TOS field should be end-to-end....
> ...clients should set it....routers should leave it alone....
The IPv4 TOS bits are depreciated as proposed in RFC1349; RFC1349
was obsoleted by RFC2474.
The DiffServ interpretation of the same bits used by RFC1349 (called
the DS bits in RFC2474) does not impose end-to-end semantics on their
use; it specifies "per-hop behaviours and mechanisms".
Joe
