RE: use of CS in qos?

yes I know thats why i dont understand the usage of the CS.

if you send a device dscp AF 43 and it going to read only the first 4 bits (ie
4) as the IPP anyway, whats the point of sending CS 4 instead for the sake of
compatibility?

Abderrahim

Date: Fri, 3 Jul 2009 20:52:33 -0400
Subject: Re: use of CS in qos?
From: jastorino_at_ipexpert.com
To: a_sadki1_at_hotmail.com
CC: ccielab_at_groupstudy.com

I think you are misunderstanding Abderrahim. The bits are exactly the same,
it is just a matter of what you call them. A device that only looks at IPP
will look at the 3 most significant bits in the ToS field of the IP header.
DSCP CS values use the 6 most significant bits of the same field, but only the
3 most significant bits really matter in CS since the lower 3 bits of every CS
will always be 0.

So check it out for example:

[ToS Field]

       ^
001 00000 = IPP 1 = CS1

It's all the same thing.

On Fri, Jul 3, 2009 at 6:18 PM, abderrahim sadki <a_sadki1_at_hotmail.com>
wrote:

thanks for your answer.

but i was wondering does a device (understading only IPP) gets a DSCP marking
isnt it going to ignore the last 3 bits anyway?
if yes what the point of the CS?

Abderrahim

Date: Fri, 3 Jul 2009 18:12:33 -0400
Subject: Re: use of CS in qos?
From: jastorino_at_ipexpert.com
To: a_sadki1_at_hotmail.com

CC: ccielab_at_groupstudy.com

Hey,

IPP uses the most significant 3 bits in the ToS field in the IP header. With
DSCP the CS values essentially do the same thing. In order to be backwards
compatible with IPP it only uses the 3 most significant bits. Maybe this
will help you visualize:

CSx Dec. IPP Binary

CS1 8 1 001 000
CS2 16 2 010 000
CS3 24 3 011 000
CS4 32 4 100 000
CS5 40 5 101 000
CS6 48 6 110 000
CS7 56 7 111 000

HTH

On Fri, Jul 3, 2009 at 5:04 PM, abderrahim sadki <a_sadki1_at_hotmail.com>
wrote:

Hi all.

I understand that CS is used for backward compatibility with IP precedence.

My question is :

if a device only understands the IP precedence and it receives a packet with
the dscp marking isnt it going to ignore the last 3 bits anyway?

Thanks,
Abderrahim
Received on Sat Jul 04 2009 - 00:55:35 ART