Return-Path: william@bourbon.usc.edu
Delivery-Date: Sun Apr 29 18:50:55 2007
X-Spam-Checker-Version: SpamAssassin 3.1.3 (2006-06-01) on merlot.usc.edu
X-Spam-Level: 
X-Spam-Status: No, score=-3.3 required=5.0 tests=ALL_TRUSTED,AWL,BAYES_00,
	NO_REAL_NAME autolearn=ham version=3.1.3
Received: from bourbon.usc.edu (bourbon.usc.edu [128.125.9.75])
	by merlot.usc.edu (8.13.5/8.13.5) with ESMTP id l3U1otiW006345
	for <cs551@merlot.usc.edu>; Sun, 29 Apr 2007 18:50:55 -0700
Received: from bourbon.usc.edu (localhost.localdomain [127.0.0.1])
	by bourbon.usc.edu (8.13.5/8.13.5) with ESMTP id l3U1oEnb013994
	for <cs551@merlot>; Sun, 29 Apr 2007 18:50:14 -0700
Message-Id: <200704300150.l3U1oEnb013994@bourbon.usc.edu>
To: cs551@merlot.usc.edu
Subject: Re: First-hop router in DiffServ 
Date: Sun, 29 Apr 2007 18:50:14 -0700
From: william@bourbon.usc.edu

Someone wrote:

  > On p. 12 of the 2-bit DiffServ slides, we talk about the
  > way first-hop routers treat customer's traffic.
  > 
  > It seems that out-of-profile premium packets are
  > penalized more harshly (delayed or dropped) than
  > out-of-profile assured packets (just mark them
  > as out-of-profile). The same can be said with
  > what happens in border router (drop extra P packet
  > but sends, not drop, extra A packet to forwarding engine,
  > as on p. 14).
  > 
  > Why so? Should premium service get "better"
  > treatment than assured service?

I assume you meant "worse" instead of "better"...  The reason
is that premium service do get treated better in the network
(has high priority over others -- see bottom of slide 15 of
lecture 22).  If you downgrade permium packets when they
arrive without tokens, then you will end up with packet
received out of order, and that's really bad.

When an assured packet get downgraded, it stays in the same
queue but can get dropped with higher probability (but no
packet out of order).
--
Bill Cheng // bill.cheng@usc.edu <URL:http://merlot.usc.edu/william/usc/>