RE: [Diffserv] Dscp and DscpOrAny TCs

"Andrew Smith" <> Fri, 07 March 2003 17:08 UTC

Received: from ( [] (may be forged)) by (8.9.1a/8.9.1a) with ESMTP id MAA17073 for <>; Fri, 7 Mar 2003 12:08:47 -0500 (EST)
Received: (from mailnull@localhost) by (8.11.6/8.11.6) id h27HKMg14825 for; Fri, 7 Mar 2003 12:20:22 -0500
Received: from (localhost.localdomain []) by (8.11.6/8.11.6) with ESMTP id h27H6hO12263; Fri, 7 Mar 2003 12:06:43 -0500
Received: from ( []) by (8.11.6/8.11.6) with ESMTP id h27Gr0O11466 for <>; Fri, 7 Mar 2003 11:53:00 -0500
Received: from ( []) by (8.9.1a/8.9.1a) with ESMTP id LAA15408 for <>; Fri, 7 Mar 2003 11:40:52 -0500 (EST)
Received: from ([] helo=ANDREWLAPTOP) by with esmtp (Exim 3.33 #1) id 18rKvq-0006IE-00; Fri, 07 Mar 2003 08:42:50 -0800
From: "Andrew Smith" <>
To: "'Wijnen, Bert \(Bert\)'" <>
Cc: <>, <>, <>, <>, <>
Subject: RE: [Diffserv] Dscp and DscpOrAny TCs
Date: Fri, 7 Mar 2003 08:44:41 -0800
Message-ID: <08c301c2e4c8$dd032c40$1500000a@ANDREWLAPTOP>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook, Build 10.0.2627
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000
In-Reply-To: <>
Importance: Normal
Content-Transfer-Encoding: 7bit
X-Mailman-Version: 2.0.12
Precedence: bulk
List-Unsubscribe: <>, <>
List-Id: Diffserv Discussion List <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
Content-Transfer-Encoding: 7bit


When you say "bring their stuff under the IETF umbrella", are you saying
that the original MIB in question was developed outside of IETF and is
being brought in, or are you talking about the link technology itself? 

If the latter, then I think the older IEEE802/IETF cooperation model on
MIBs works OK - there you had an IETF WG (bridge, maumib, ifmib etc.)
defining SMI MIBs under IETF root OID and having WG contributors make
sure that adequate coordination happened; a newer model where IEEE802
has defined some of its own SMI MIBs has also been working OK (e.g. Link
Aggregation MIB from IEEE802, defined under IEEE802's own root OID). 

For the former case, what I don't think works is to have a MIB defined
outside of IETF and then brought to the IETF for some reason. What would
be the reason to do that? for IETF endorsement, for having IETF fixing
things that are broken, just to use the IETF root OID? I don't know the
specific reasons in this case for how this work got here. But it's
obvious to me that, if you bring existing work to IETF for
standardisation, be it the work of another standards' group, an industry
group or individual contributions, *you give up change control* and you
do not get any guarantees of backwards-compatibility: naturally you
don't expect gratuitous non-compatibility with earlier work - the WG
needs a valid reason to change something and one such reason would be
alignment with other IETF MIBs.

[Note that, for this particular example of filter/pattern-matching on IP
packets with DSCPs, I'd originally proposed a separate generic MIB
module (included a "sixTupleClassifier" table) for the reason of making
it clear that this was general purpose and should be used by other MIBs,
including the Diffserv MIB. However, in the end, it was decided to merge
it into the main Diffserv MIB module (it became the
"diffServMultiFieldClfrTable" in RFC3289 - maybe it is less visible
there but maybe ADs can do something to raise awareness of its presence
amongst other WGs].

Andrew Smith

-----Original Message-----
From: Wijnen, Bert (Bert) [] 
Sent: Friday, March 07, 2003 2:57 AM
To: Andrew Smith; Bert Wijnen
Subject: RE: [Diffserv] Dscp and DscpOrAny TCs


Sure.... and what I am doing here is not the ideal solution,
but clearly I am trying to get various solutions aligned
as much as possible (you have seen my issues raised
for common definitions for Dscp, DscpOrAny, FlowLabel,
FlowLabelOrAny, VlanId, VlanIdOrAny ... etc etc.

In the case of DOCSIS, the original stuff got developed
outside IETF. They now try to bring their stuff under the
IETF umbrella. We could tell them to drop/ditch all their own
stuff and start from scratch... which I suspect will not work,
or we can try to get them aligned as much as possible without
making them start from scratch.

I think I am on the latter path...
Are you suggesting we (IETF) should be on the former path?

Just trying to understand what you are urging us to do.


> -----Original Message-----
> From: Andrew Smith []
> Sent: woensdag 5 maart 2003 21:42
> To: Bert Wijnen
> Cc:;;
> Subject: RE: [Diffserv] Dscp and DscpOrAny TCs
> I've not been following the DOCSIS MIB work closely and don't know the
> context in which this MIB was developed. That said, shouldn't IETF be
> promoting the use of generic, already-standards-track 
> solutions, instead
> of trying to police each new reincarnation of the same thing? It would
> have been nice if, given that this MIB is a work item of an 
> IETF WG, the
> IESG had put in some provision in the WG charter about re-use of
> existing IETF work where appropriate, rather than reinventing 
> the wheel
> (which this looks like - see my first sentence above though). Granted,
> there is a choice of almost-equivalent wheels for packet
> matching/filtering in the IETF standards-track repertoire but 
> that's not
> an argument for creating yet another one.
> This is one aspect of the general IETF process issue of how to ensure
> that appropriate "subject matter experts" are available to a WG that
> needs them, even when that subject matter's own WG is no longer
> existent: the more generic the solutions that IESG steers us to, the
> less this maintenance problem will be.
> Reuse of already-defined MIB ojects/tables is also one of the issues
> that an SMIng should try to make easier (and people have been 
> looking at
> doing this, I believe).
> Andrew Smith
> -----Original Message-----
> From: 
> [] On Behalf
> Of John Schnizlein
> Sent: Wednesday, March 05, 2003 9:29 AM
> To: Wijnen, Bert (Bert)
> Cc:
> Subject: Re: [Diffserv] Dscp and DscpOrAny TCs
> Good catch, Bert.
> The idea of a mask for the DSCP was killed some time ago in Policy
> Framework based on its incompatibility with the following explicit
> prohibition in the definition of the Differentiated Services Field:
> RFC 2474 Section 3, page 7
>    Implementors should note that the DSCP field is six bits wide.  DS-
>    compliant nodes MUST select PHBs by matching against the 
> entire 6-bit
>    DSCP field, e.g., by treating the value of the field as a 
> table index
>    which is used to select a particular packet handling 
> mechanism which
>    has been implemented in that device.  The value of the CU 
> field MUST
>    be ignored by PHB selection.  The DSCP field is defined as an
>    unstructured field to facilitate the definition of future per-hop
>    behaviors.
> John
> At 11:36 AM 3/5/2003, Wijnen, Bert (Bert) wrote:
> >Do Diffservers think that the following is wise and makes sense?
> >
> >   docsSubMgtPktFilterDscpValue OBJECT-TYPE
> >       SYNTAX      Dscp
> >       MAX-ACCESS  read-create
> >       STATUS      current
> >           "The Differentiated Services Code Point (DSCP) value to
> match
> >            in the IP packet."
> >       DEFVAL { 0 }
> >       ::= { docsSubMgtPktFilterEntry 9 }
> >
> >   docsSubMgtPktFilterDscpMask OBJECT-TYPE
> >       SYNTAX      Dscp
> >       MAX-ACCESS  read-create
> >       STATUS      current
> >           "The mask to apply against the DSCP value to be matched in
> >       the IP packet.  The default for both these objects taken
> together
> >       matches all DSCP values. A packet matches this filter if the
> >       following is true:
> >           AND (FilterDscpValue, FilterDscpMask) ==
> >           AND (Packet DSCP Value, FilterDscpMask)."
> >       DEFVAL { 0 }
> >       ::= { docsSubMgtPktFilterEntry 10 }
> >
> >It is defined in draft-ietf-ipcdn-subscriber-mib-10.txt and I'd like
> >to hear comments from Diffserv experts.
> >
> >Thanks,
> >Bert 
> _______________________________________________
> diffserv mailing list
> Archive:

diffserv mailing list