From ancp-bounces@ietf.org  Mon Aug  4 09:06:28 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id AEC223A6AA2;
	Mon,  4 Aug 2008 09:06:28 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id ED7F23A6A60
	for <ancp@core3.amsl.com>; Mon,  4 Aug 2008 09:06:27 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.719
X-Spam-Level: 
X-Spam-Status: No, score=-0.719 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id rb0JJkDyeQ1t for <ancp@core3.amsl.com>;
	Mon,  4 Aug 2008 09:06:27 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id 7154D3A68E4
	for <ancp@ietf.org>; Mon,  4 Aug 2008 09:06:26 -0700 (PDT)
Received: from ptpxch007rm001.idc.cww.telecomitalia.it ([156.54.205.169]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Mon, 4 Aug 2008 18:06:52 +0200
Received: from grfhub702rm001.griffon.local ([10.19.3.9]) by
	ptpxch007rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Mon, 4 Aug 2008 18:06:52 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub702rm001.griffon.local ([10.19.3.9]) with mapi;
	Mon, 4 Aug 2008 18:06:52 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: <ancp@ietf.org>
Date: Mon, 4 Aug 2008 18:06:52 +0200
Thread-Topic: Result field use and ANCP messaging principles 
thread-index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4g==
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 04 Aug 2008 16:06:52.0589 (UTC)
	FILETIME=[1BE355D0:01C8F64C]
Subject: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header fields that may deserve some additional clarifications in ANCP protocol spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result = 1
AckAll:             Result = 2
Success:            Result = 3
Failure:            Result = 4
More:               Result = 5
ReturnReceipt       Result = 6

"NoSuccessAck" and  "AckAll" used in Request message;
"Success," "More," and "Failure" used in Response message;
and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP design, it could be worth to rationalize the use of Result field in case of ANCP. In particular I'd like to identify/clarify which are the most useful codes for this field.

Following the proposal presented by the design team for Multicast extension the most useful codes for Result would seem to be Nack (NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure 0x4 for response. Considering that WG selected to carry ANCP over a reliable transport protocol NAck appears to be the most suitable candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify the code for Result. Does anybody know the Result code for this scenario?

The use of Result field 0 looks quite strange to me as this code does not belong to the range of values defined for Result field in RFC3292 thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result field has impacts on the messaging interaction mode in terms of request/response (Nack versus AckAll.) Regarding messaging principles RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type
2. A request message type with a response sent in the same message type but a different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and having in mind also protocol flexibility for future extensions it would seem reasonable to use the 1st alternative as it is much more easier to understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of Result codes for ANCP (deviations from GSMP), in particular if nobody has a use-case that requires AckAll I would propose to remove the Ackall behavior for Result field from the Request message and to use Nack for all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response message type) as the only mode for messaging interaction in ANCP and discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 01:22:30 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 8FE2228C21E;
	Tue,  5 Aug 2008 01:22:30 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id E70B728C21E
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 01:22:28 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.249
X-Spam-Level: 
X-Spam-Status: No, score=-2.249 tagged_above=-999 required=5 tests=[AWL=0.000, 
	BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id FSKi4ifVL4tq for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 01:22:28 -0700 (PDT)
Received: from smail6.alcatel.fr (gc-na5.alcatel.fr [64.208.49.5])
	by core3.amsl.com (Postfix) with ESMTP id EE36528C223
	for <ancp@ietf.org>; Tue,  5 Aug 2008 01:22:27 -0700 (PDT)
Received: from FRVELSBHS02.ad2.ad.alcatel.com (frvelsbhs02.ad2.ad.alcatel.com
	[155.132.6.74])
	by smail6.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m758Muat030519
	for <ancp@ietf.org>; Tue, 5 Aug 2008 10:22:56 +0200
Received: from FRVELSMBS11.ad2.ad.alcatel.com ([155.132.6.37]) by
	FRVELSBHS02.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Tue, 5 Aug 2008 10:22:57 +0200
X-MIMEOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 10:22:55 +0200
Message-ID: <0458D2EE0C36744BABB36BE37805C29A025C476C@FRVELSMBS11.ad2.ad.alcatel.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Please ask your WG... 
Thread-Index: Acj2TD7+jpdvtlJaQp2IbO/irnzGbwAiCHDA
From: "BOCCI Matthew" <Matthew.Bocci@alcatel-lucent.co.uk>
To: <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 08:22:57.0098 (UTC)
	FILETIME=[770E3AA0:01C8F6D4]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.84
Subject: [ANCP] FW: Please ask your WG...
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Forwarded from the nomcom chair.

-----Original Message-----
From: wgchairs-bounces@ietf.org [mailto:wgchairs-bounces@ietf.org] On
Behalf Of NomCom Chair
Sent: 03 August 2008 15:43
To: Working Group Chairs
Subject: Please ask your WG... 

To volunteer for the nomcom.
The nomcom process is (in my opinion) better served by a large pool of
volunteers drawn from a wide spectrum of IETF attendees.
As such, please ask on your individual mailing lists for folks to
volunteer.

Obviously, the exact method for doing so is up to you.
The most recent call for volunteers can be reference here:
    https://datatracker.ietf.org/public/show_nomcom_message.cgi?id=1617
Whether you copy that message, or reference is probably up to you and
the habits of your working group mailing list.
If you want to reference or copy the status message I sent out, that can
be found at:
    https://datatracker.ietf.org/public/show_nomcom_message.cgi?id=1618

Thank you,
Joel M. Halpern
Nomcom Chair
jmh@joelahlpern.com
nomcom-chair@ietf.org
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 01:58:54 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 96C933A6BD6;
	Tue,  5 Aug 2008 01:58:54 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 493613A6BD6
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 01:58:53 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.249
X-Spam-Level: 
X-Spam-Status: No, score=-2.249 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id xdTGS+WeLDWn for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 01:58:52 -0700 (PDT)
Received: from smail5.alcatel.fr (smail5.alcatel.fr [64.208.49.27])
	by core3.amsl.com (Postfix) with ESMTP id D3B7C3A6B58
	for <ancp@ietf.org>; Tue,  5 Aug 2008 01:58:51 -0700 (PDT)
Received: from FRVELSBHS03.ad2.ad.alcatel.com (frvelsbhs03.ad2.ad.alcatel.com
	[155.132.6.75])
	by smail5.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m758xKHa009035; 
	Tue, 5 Aug 2008 10:59:20 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.51]) by
	FRVELSBHS03.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Tue, 5 Aug 2008 10:59:20 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 10:59:19 +0200
Message-ID: <7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
Thread-Index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23w
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local>
From: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>
To: "Maglione Roberta" <roberta.maglione@telecomitalia.it>, <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 08:59:20.0657 (UTC)
	FILETIME=[8C8EE410:01C8F6D9]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.13
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Hi,

Would it not be possible to have a mix of both modes: for some messages
the response has the same message type and for others the response is a
different message type?  the meaning of ackall in oam looks quite clear
to me as it is explained in the ancp spec.  The value NoSuccessAck is
not good for OAM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292
and see that value 6 is used in case an event message expects a response
but the rfc does not mention what value has to be used in case the event
message does not expect a response.  The value 0 looks a reasonable
choice to me. it means that the field is not used.  Actually, RFC3292
seems to suggest that for port up messages no response is sent unless
the result field is set to 6 in the port up.  This means that when no
response is expected like in ancp, any value is allowed besides from 6
so value 0 is good as well.  ANCP spec is further restricting that it
has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header
fields that may deserve some additional clarifications in ANCP protocol
spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result = 1
AckAll:             Result = 2
Success:            Result = 3
Failure:            Result = 4
More:               Result = 5
ReturnReceipt       Result = 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP
design, it could be worth to rationalize the use of Result field in case
of ANCP. In particular I'd like to identify/clarify which are the most
useful codes for this field.

Following the proposal presented by the design team for Multicast
extension the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a
reliable transport protocol NAck appears to be the most suitable
candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of
implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case
of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use
case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify
the code for Result. Does anybody know the Result code for this
scenario?

The use of Result field 0 looks quite strange to me as this code does
not belong to the range of values defined for Result field in RFC3292
thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result
field has impacts on the messaging interaction mode in terms of
request/response (Nack versus AckAll.) Regarding messaging principles
RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type 2. A request
message type with a response sent in the same message type but a
different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and
having in mind also protocol flexibility for future extensions it would
seem reasonable to use the 1st alternative as it is much more easier to
understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback
on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of
Result codes for ANCP (deviations from GSMP), in particular if nobody
has a use-case that requires AckAll I would propose to remove the Ackall
behavior for Result field from the Request message and to use Nack for
all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response
message type) as the only mode for messaging interaction in ANCP and
discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 04:44:00 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 67BEA3A6CEA;
	Tue,  5 Aug 2008 04:44:00 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 9E3623A6CE2
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 04:43:59 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.719
X-Spam-Level: 
X-Spam-Status: No, score=-0.719 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id zhU4LY6ExZbU for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 04:43:58 -0700 (PDT)
Received: from mailm.telecomitalia.it (mailm.telecomitalia.it [217.169.121.53])
	by core3.amsl.com (Postfix) with ESMTP id B34CD3A6CE9
	for <ancp@ietf.org>; Tue,  5 Aug 2008 04:43:57 -0700 (PDT)
Received: from ptpxch009rm001.idc.cww.telecomitalia.it ([156.54.205.171]) by
	mailm.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Tue, 5 Aug 2008 13:44:26 +0200
Received: from GRFHUB704RM001.griffon.local ([10.19.3.67]) by
	ptpxch009rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Tue, 5 Aug 2008 13:44:26 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	GRFHUB704RM001.griffon.local ([10.19.3.67]) with mapi;
	Tue, 5 Aug 2008 13:44:25 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "'DE CNODDER Stefaan'" <stefaan.de_cnodder@alcatel-lucent.be>,
	<ancp@ietf.org>
Date: Tue, 5 Aug 2008 13:44:25 +0200
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
thread-index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSA=
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local>
	<7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 05 Aug 2008 11:44:26.0062 (UTC)
	FILETIME=[9CA3BEE0:01C8F6F0]
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Hello,

> Would it not be possible to have a mix of both modes: for some messages
> the response has the same message type and for others the response is a
> different message type?

I don't say it is not possible to have a mix of both modes for different use-cases, but also thinking about future message designer it would seem cleaner to have one single approach for message principles for all use cases rather having to deal with different methods for different scenarios. In addition there could be cases where the same response message could be re-used across different scenarios.

> the meaning of ackall in oam looks quite clear
> to me as it is explained in the ancp spec.  The value NoSuccessAck is
> not good for OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is there a different way to convey that in case of OAM response is required? For example can OAM request message request/application be made to mean "always reply"?

> value 0 is reserved but is nowhere explained.  I was looking to RFC3292
> and see that value 6 is used in case an event message expects a response
> but the rfc does not mention what value has to be used in case the event
> message does not expect a response.  The value 0 looks a reasonable
> choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no response is sent unless result is set to 6, but as you mentioned RFC 3292 says that ReturnReceipt can be used in Events to indicate that an acknowledgement is required for the message.   It is not clear to me if RetunrnReceipt means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set to zero and implementations interpret this "0" as  "no response" this behavior does not allow to identify error conditions. Let's consider for example a scenario where NAS receives from the AN a Port-UP message where a mandatory TLV is somehow missing or wrong; the ANCP protocol spec says that the NAS is supposed to discard this message ("if a TLV is not understood by the NAS, it is silently ignored".)
In this case the NAS just discards the message but the AN does not have any way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if we use result equal to 6 what do we get ackall or Nack? Don't you think it could be useful to have a sort of "Nack" behavior both for Port-UP and Port-Management (for line configuration) in order to be able to catch error conditions?
In addition in case of the error condition the same response message could be re-used for different scenarios if we choose the philosophy "A request message type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6
> so value 0 is good as well.  ANCP spec is further restricting that it
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages
the response has the same message type and for others the response is a
different message type?  the meaning of ackall in oam looks quite clear
to me as it is explained in the ancp spec.  The value NoSuccessAck is
not good for OAM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292
and see that value 6 is used in case an event message expects a response
but the rfc does not mention what value has to be used in case the event
message does not expect a response.  The value 0 looks a reasonable
choice to me. it means that the field is not used.  Actually, RFC3292
seems to suggest that for port up messages no response is sent unless
the result field is set to 6 in the port up.  This means that when no
response is expected like in ancp, any value is allowed besides from 6
so value 0 is good as well.  ANCP spec is further restricting that it
has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header
fields that may deserve some additional clarifications in ANCP protocol
spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result = 1
AckAll:             Result = 2
Success:            Result = 3
Failure:            Result = 4
More:               Result = 5
ReturnReceipt       Result = 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP
design, it could be worth to rationalize the use of Result field in case
of ANCP. In particular I'd like to identify/clarify which are the most
useful codes for this field.

Following the proposal presented by the design team for Multicast
extension the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a
reliable transport protocol NAck appears to be the most suitable
candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of
implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case
of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use
case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify
the code for Result. Does anybody know the Result code for this
scenario?

The use of Result field 0 looks quite strange to me as this code does
not belong to the range of values defined for Result field in RFC3292
thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result
field has impacts on the messaging interaction mode in terms of
request/response (Nack versus AckAll.) Regarding messaging principles
RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type 2. A request
message type with a response sent in the same message type but a
different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and
having in mind also protocol flexibility for future extensions it would
seem reasonable to use the 1st alternative as it is much more easier to
understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback
on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of
Result codes for ANCP (deviations from GSMP), in particular if nobody
has a use-case that requires AckAll I would propose to remove the Ackall
behavior for Result field from the Request message and to use Nack for
all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response
message type) as the only mode for messaging interaction in ANCP and
discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 06:11:06 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 75F983A691C;
	Tue,  5 Aug 2008 06:11:06 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id D43A23A691C
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 06:11:04 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.249
X-Spam-Level: 
X-Spam-Status: No, score=-2.249 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id OvNCxeiO3V+E for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 06:11:03 -0700 (PDT)
Received: from smail5.alcatel.fr (smail5.alcatel.fr [62.23.212.27])
	by core3.amsl.com (Postfix) with ESMTP id 73CB73A67B6
	for <ancp@ietf.org>; Tue,  5 Aug 2008 06:11:02 -0700 (PDT)
Received: from FRVELSBHS03.ad2.ad.alcatel.com (frvelsbhs03.ad2.ad.alcatel.com
	[155.132.6.75])
	by smail5.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m75DBVUI000943; 
	Tue, 5 Aug 2008 15:11:31 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.51]) by
	FRVELSBHS03.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Tue, 5 Aug 2008 15:11:31 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 15:11:29 +0200
Message-ID: <7A6D4815C5E8F74988784FEBD50F2F1E02158EE2@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
Thread-Index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAABlLJgA==
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local>
	<7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>
From: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>
To: "Maglione Roberta" <roberta.maglione@telecomitalia.it>, <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 13:11:31.0494 (UTC)
	FILETIME=[C73D6460:01C8F6FC]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.13
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Hi Roberta,

It makes no sense to define a response for port-up because it is not
clear to me what the AN should do with that information.  And if their
would be a good case where it is useful, I would prefer to use the same
message type for the response.    The meaning of returnreceipt looks
like an ackall, but if you want a behavior that is like Nack then it is
still possible to define new values for the Result field.

What is the impact of this proposal exactly on the existing port-up and
OAM messages?  Is the proposal to use other message types for OAM
messages?

regards,
Stefaan


-----Original Message-----
From: Maglione Roberta [mailto:roberta.maglione@telecomitalia.it] 
Sent: dinsdag 5 augustus 2008 13:44
To: DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some 
> messages the response has the same message type and for others the 
> response is a different message type?

I don't say it is not possible to have a mix of both modes for different
use-cases, but also thinking about future message designer it would seem
cleaner to have one single approach for message principles for all use
cases rather having to deal with different methods for different
scenarios. In addition there could be cases where the same response
message could be re-used across different scenarios.

> the meaning of ackall in oam looks quite clear to me as it is 
> explained in the ancp spec.  The value NoSuccessAck is not good for 
> OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is
there a different way to convey that in case of OAM response is
required? For example can OAM request message request/application be
made to mean "always reply"?

> value 0 is reserved but is nowhere explained.  I was looking to 
> RFC3292 and see that value 6 is used in case an event message expects 
> a response but the rfc does not mention what value has to be used in 
> case the event message does not expect a response.  The value 0 looks 
> a reasonable choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless 
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no
response is sent unless result is set to 6, but as you mentioned RFC
3292 says that ReturnReceipt can be used in Events to indicate that an
acknowledgement is required for the message.   It is not clear to me if
RetunrnReceipt means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result
set to zero and implementations interpret this "0" as  "no response"
this behavior does not allow to identify error conditions. Let's
consider for example a scenario where NAS receives from the AN a Port-UP
message where a mandatory TLV is somehow missing or wrong; the ANCP
protocol spec says that the NAS is supposed to discard this message ("if
a TLV is not understood by the NAS, it is silently ignored".) In this
case the NAS just discards the message but the AN does not have any way
to know the something wrong happened.
Now back to the original question today we have 0 meaning no response,
if we use result equal to 6 what do we get ackall or Nack? Don't you
think it could be useful to have a sort of "Nack" behavior both for
Port-UP and Port-Management (for line configuration) in order to be able
to catch error conditions?
In addition in case of the error condition the same response message
could be re-used for different scenarios if we choose the philosophy "A
request message type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6

> so value 0 is good as well.  ANCP spec is further restricting that it 
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages
the response has the same message type and for others the response is a
different message type?  the meaning of ackall in oam looks quite clear
to me as it is explained in the ancp spec.  The value NoSuccessAck is
not good for OAM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292
and see that value 6 is used in case an event message expects a response
but the rfc does not mention what value has to be used in case the event
message does not expect a response.  The value 0 looks a reasonable
choice to me. it means that the field is not used.  Actually, RFC3292
seems to suggest that for port up messages no response is sent unless
the result field is set to 6 in the port up.  This means that when no
response is expected like in ancp, any value is allowed besides from 6
so value 0 is good as well.  ANCP spec is further restricting that it
has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header
fields that may deserve some additional clarifications in ANCP protocol
spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result = 1
AckAll:             Result = 2
Success:            Result = 3
Failure:            Result = 4
More:               Result = 5
ReturnReceipt       Result = 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP
design, it could be worth to rationalize the use of Result field in case
of ANCP. In particular I'd like to identify/clarify which are the most
useful codes for this field.

Following the proposal presented by the design team for Multicast
extension the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a
reliable transport protocol NAck appears to be the most suitable
candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of
implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case
of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use
case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify
the code for Result. Does anybody know the Result code for this
scenario?

The use of Result field 0 looks quite strange to me as this code does
not belong to the range of values defined for Result field in RFC3292
thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result
field has impacts on the messaging interaction mode in terms of
request/response (Nack versus AckAll.) Regarding messaging principles
RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type 2. A request
message type with a response sent in the same message type but a
different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and
having in mind also protocol flexibility for future extensions it would
seem reasonable to use the 1st alternative as it is much more easier to
understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback
on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of
Result codes for ANCP (deviations from GSMP), in particular if nobody
has a use-case that requires AckAll I would propose to remove the Ackall
behavior for Result field from the Request message and to use Nack for
all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response
message type) as the only mode for messaging interaction in ANCP and
discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 06:52:32 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 5FA2D3A6B1A;
	Tue,  5 Aug 2008 06:52:32 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 46B3E3A6B1A
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 06:52:31 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.249
X-Spam-Level: 
X-Spam-Status: No, score=-3.249 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_DE=0.35, RCVD_IN_DNSWL_LOW=-1]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id nbAa9ILDVvuI for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 06:52:29 -0700 (PDT)
Received: from tcmail23.telekom.de (tcmail23.telekom.de [217.6.95.237])
	by core3.amsl.com (Postfix) with ESMTP id D985D3A68EF
	for <ancp@ietf.org>; Tue,  5 Aug 2008 06:52:28 -0700 (PDT)
Received: from s4de8psaans.mitte.t-com.de (s4de8psaans.mitte.t-com.de
	[10.151.180.168]) by tcmail21.telekom.de with ESMTP;
	Tue, 5 Aug 2008 15:52:46 +0200
Received: from S4DE8PSAANK.mitte.t-com.de ([10.151.229.10]) by
	s4de8psaans.mitte.t-com.de with Microsoft SMTPSVC(6.0.3790.3959); 
	Tue, 5 Aug 2008 15:52:45 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 15:56:10 +0200
Message-Id: <1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
thread-index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0A==
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>
From: "Haag, T" <Thomas.Haag@t-systems.com>
To: <roberta.maglione@telecomitalia.it>,
	<stefaan.de_cnodder@alcatel-lucent.be>, <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 13:52:45.0919 (UTC)
	FILETIME=[8A1CAEF0:01C8F702]
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more. So the benefit is not clear to me for use case topology d=
iscovery.

Mechanisms for connectivity control:  =

If we run adjacency we have keep alive between AN and BRAS. If we run TCP/I=
P we have flow control. I don't see an added value for using response messa=
ges because in case of failure the mechanism inside AN and / or BRAS failed.

Considering the OAM use case error conditions for OAM use case are defined =
too.

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

Regards
 =

Thomas
 =

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some messages
> the response has the same message type and for others the response is a
> different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear
> to me as it is explained in the ancp spec.  The value NoSuccessAck is
> not good for OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to RFC3292
> and see that value 6 is used in case an event message expects a response
> but the rfc does not mention what value has to be used in case the event
> message does not expect a response.  The value 0 looks a reasonable
> choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".)
In this case the NAS just discards the message but the AN does not have any=
 way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6
> so value 0 is good as well.  ANCP spec is further restricting that it
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages
the response has the same message type and for others the response is a
different message type?  the meaning of ackall in oam looks quite clear
to me as it is explained in the ancp spec.  The value NoSuccessAck is
not good for OAM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292
and see that value 6 is used in case an event message expects a response
but the rfc does not mention what value has to be used in case the event
message does not expect a response.  The value 0 looks a reasonable
choice to me. it means that the field is not used.  Actually, RFC3292
seems to suggest that for port up messages no response is sent unless
the result field is set to 6 in the port up.  This means that when no
response is expected like in ancp, any value is allowed besides from 6
so value 0 is good as well.  ANCP spec is further restricting that it
has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header
fields that may deserve some additional clarifications in ANCP protocol
spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP
design, it could be worth to rationalize the use of Result field in case
of ANCP. In particular I'd like to identify/clarify which are the most
useful codes for this field.

Following the proposal presented by the design team for Multicast
extension the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a
reliable transport protocol NAck appears to be the most suitable
candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of
implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case
of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use
case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify
the code for Result. Does anybody know the Result code for this
scenario?

The use of Result field 0 looks quite strange to me as this code does
not belong to the range of values defined for Result field in RFC3292
thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result
field has impacts on the messaging interaction mode in terms of
request/response (Nack versus AckAll.) Regarding messaging principles
RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type 2. A request
message type with a response sent in the same message type but a
different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and
having in mind also protocol flexibility for future extensions it would
seem reasonable to use the 1st alternative as it is much more easier to
understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback
on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of
Result codes for ANCP (deviations from GSMP), in particular if nobody
has a use-case that requires AckAll I would propose to remove the Ackall
behavior for Result field from the Request message and to use Nack for
all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response
message type) as the only mode for messaging interaction in ANCP and
discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        =

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 07:27:11 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 9EA1C3A69B6;
	Tue,  5 Aug 2008 07:27:11 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 0B3E73A69B6
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 07:27:11 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.719
X-Spam-Level: 
X-Spam-Status: No, score=-0.719 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id xD7xtVNEJSRx for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 07:27:09 -0700 (PDT)
Received: from mailm.telecomitalia.it (mailm.telecomitalia.it [217.169.121.53])
	by core3.amsl.com (Postfix) with ESMTP id 806D23A6917
	for <ancp@ietf.org>; Tue,  5 Aug 2008 07:27:08 -0700 (PDT)
Received: from ptpxch008rm001.idc.cww.telecomitalia.it ([156.54.205.170]) by
	mailm.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Tue, 5 Aug 2008 16:27:16 +0200
Received: from GRFHUB706RM001.griffon.local ([10.19.3.71]) by
	ptpxch008rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Tue, 5 Aug 2008 16:27:15 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	GRFHUB706RM001.griffon.local ([10.19.3.71]) with mapi;
	Tue, 5 Aug 2008 16:27:15 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>,
	<ancp@ietf.org>
Date: Tue, 5 Aug 2008 16:27:14 +0200
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
thread-index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAABlLJgAACOcQ1
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A201294A2BE8@GRFMBX702RM001.griffon.local>
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local>
	<7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>,
	<7A6D4815C5E8F74988784FEBD50F2F1E02158EE2@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <7A6D4815C5E8F74988784FEBD50F2F1E02158EE2@FRVELSMBS22.ad2.ad.alcatel.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 05 Aug 2008 14:27:15.0856 (UTC)
	FILETIME=[5BE40100:01C8F707]
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Stefaan,
 thanks for your comments please see inline.

Regards,
Roberta
________________________________________
From: DE CNODDER Stefaan [stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 3:11 PM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles

Hi Roberta,

It makes no sense to define a response for port-up because it is not
clear to me what the AN should do with that information.  And if their
would be a good case where it is useful,

[RM] Don't you think it would be useful for troubleshooting purposes to be able to detect an error condition like for example malformed TLV in Port-UP?

 I would prefer to use the same message type for the response.
[RM] if we would like to detect an error condition for topology discovery you are suggesting the same message type for the response thus a port up with code error
For line configuration: NAS uses Port management to configure some parameters on the AN and something goes wrong also in this case do you think is better to  use the same message type for the response?


The meaning of returnreceipt looks like an ackall,

[RM] Ackall does not seem useful with TCP and it does not scale

but if you want a behavior that is like Nack then it is
still possible to define new values for the Result field.

[RM] Why do we need to re-define the Nack behavior if there is already a Result code for Nack?
ReturnReceipt is already a duplicate of Ackall, do you find useful having two different codes with same meaning but to be used for different message types?


What is the impact of this proposal exactly on the existing port-up and
OAM messages?  Is the proposal to use other message types for OAM
messages?

[RM] I am just trying to understand if it is possible to use a single mode of interaction ("A
request message type and a response message type") adding additional functionalities (like detecting error conditions) but without or minimizing impacts on current implementations in order to maintain backward compatibility with current implementations.

Regards,
Roberta

regards,
Stefaan


-----Original Message-----
From: Maglione Roberta [mailto:roberta.maglione@telecomitalia.it]
Sent: dinsdag 5 augustus 2008 13:44
To: DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some
> messages the response has the same message type and for others the
> response is a different message type?

I don't say it is not possible to have a mix of both modes for different
use-cases, but also thinking about future message designer it would seem
cleaner to have one single approach for message principles for all use
cases rather having to deal with different methods for different
scenarios. In addition there could be cases where the same response
message could be re-used across different scenarios.

> the meaning of ackall in oam looks quite clear to me as it is
> explained in the ancp spec.  The value NoSuccessAck is not good for
> OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is
there a different way to convey that in case of OAM response is
required? For example can OAM request message request/application be
made to mean "always reply"?

> value 0 is reserved but is nowhere explained.  I was looking to
> RFC3292 and see that value 6 is used in case an event message expects
> a response but the rfc does not mention what value has to be used in
> case the event message does not expect a response.  The value 0 looks
> a reasonable choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no
response is sent unless result is set to 6, but as you mentioned RFC
3292 says that ReturnReceipt can be used in Events to indicate that an
acknowledgement is required for the message.   It is not clear to me if
RetunrnReceipt means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result
set to zero and implementations interpret this "0" as  "no response"
this behavior does not allow to identify error conditions. Let's
consider for example a scenario where NAS receives from the AN a Port-UP
message where a mandatory TLV is somehow missing or wrong; the ANCP
protocol spec says that the NAS is supposed to discard this message ("if
a TLV is not understood by the NAS, it is silently ignored".) In this
case the NAS just discards the message but the AN does not have any way
to know the something wrong happened.
Now back to the original question today we have 0 meaning no response,
if we use result equal to 6 what do we get ackall or Nack? Don't you
think it could be useful to have a sort of "Nack" behavior both for
Port-UP and Port-Management (for line configuration) in order to be able
to catch error conditions?
In addition in case of the error condition the same response message
could be re-used for different scenarios if we choose the philosophy "A
request message type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6

> so value 0 is good as well.  ANCP spec is further restricting that it
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages
the response has the same message type and for others the response is a
different message type?  the meaning of ackall in oam looks quite clear
to me as it is explained in the ancp spec.  The value NoSuccessAck is
not good for OAM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292
and see that value 6 is used in case an event message expects a response
but the rfc does not mention what value has to be used in case the event
message does not expect a response.  The value 0 looks a reasonable
choice to me. it means that the field is not used.  Actually, RFC3292
seems to suggest that for port up messages no response is sent unless
the result field is set to 6 in the port up.  This means that when no
response is expected like in ancp, any value is allowed besides from 6
so value 0 is good as well.  ANCP spec is further restricting that it
has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header
fields that may deserve some additional clarifications in ANCP protocol
spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result = 1
AckAll:             Result = 2
Success:            Result = 3
Failure:            Result = 4
More:               Result = 5
ReturnReceipt       Result = 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP
design, it could be worth to rationalize the use of Result field in case
of ANCP. In particular I'd like to identify/clarify which are the most
useful codes for this field.

Following the proposal presented by the design team for Multicast
extension the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a
reliable transport protocol NAck appears to be the most suitable
candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of
implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case
of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use
case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify
the code for Result. Does anybody know the Result code for this
scenario?

The use of Result field 0 looks quite strange to me as this code does
not belong to the range of values defined for Result field in RFC3292
thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result
field has impacts on the messaging interaction mode in terms of
request/response (Nack versus AckAll.) Regarding messaging principles
RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type 2. A request
message type with a response sent in the same message type but a
different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and
having in mind also protocol flexibility for future extensions it would
seem reasonable to use the 1st alternative as it is much more easier to
understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback
on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of
Result codes for ANCP (deviations from GSMP), in particular if nobody
has a use-case that requires AckAll I would propose to remove the Ackall
behavior for Result field from the Request message and to use Nack for
all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response
message type) as the only mode for messaging interaction in ANCP and
discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 08:02:57 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 1348428C186;
	Tue,  5 Aug 2008 08:02:57 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id C086A28C2B2
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 08:02:55 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.719
X-Spam-Level: 
X-Spam-Status: No, score=-0.719 tagged_above=-999 required=5
	tests=[AWL=-0.000, BAYES_00=-2.599, HELO_EQ_IT=0.635,
	HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id wVIvcutDyzKg for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 08:02:54 -0700 (PDT)
Received: from mailm.telecomitalia.it (mailm.telecomitalia.it [217.169.121.53])
	by core3.amsl.com (Postfix) with ESMTP id 8DE8728C124
	for <ancp@ietf.org>; Tue,  5 Aug 2008 08:02:53 -0700 (PDT)
Received: from ptpxch008rm001.idc.cww.telecomitalia.it ([156.54.205.170]) by
	mailm.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Tue, 5 Aug 2008 17:03:18 +0200
Received: from grfhub701rm001.griffon.local ([10.19.3.8]) by
	ptpxch008rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Tue, 5 Aug 2008 17:03:18 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub701rm001.griffon.local ([10.19.3.8]) with mapi;
	Tue, 5 Aug 2008 17:03:18 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "Haag, T" <Thomas.Haag@t-systems.com>,
	<stefaan.de_cnodder@alcatel-lucent.be>, <ancp@ietf.org>
Date: Tue, 5 Aug 2008 17:03:17 +0200
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
thread-index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0AAB02dH
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A201294A2BE9@GRFMBX702RM001.griffon.local>
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>,
	<1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de>
In-Reply-To: <1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de>
Accept-Language: en-US, it-IT
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US, it-IT
MIME-Version: 1.0
X-OriginalArrivalTime: 05 Aug 2008 15:03:18.0634 (UTC)
	FILETIME=[6501BCA0:01C8F70C]
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Thomas,
   please see inline.
Thanks,
Regards,
Roberta
________________________________________
From: Haag, T [Thomas.Haag@t-systems.com]
Sent: Tuesday, August 05, 2008 3:56 PM
To: Maglione Roberta; stefaan.de_cnodder@alcatel-lucent.be; ancp@ietf.org
Subject: AW: [ANCP] Result field use and ANCP messaging principles

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more.

[RM] Agree with you on this point that's why I'm saying that AckAll for Res=
ult does not scale, we should have a Nack behavior.

So the benefit is not clear to me for use case topology discovery.

[RM] I have seen the following scenarios:
1. AN and NAS implementing two different flavors of ANCP: AN was sending Po=
rt-UP with one label format, while NAS was expecting a label with another f=
ormat.
2. Other case: problems caused by deprecated TLVs same behavior
AN was sending PORT-UP with TLV 0x04 sub-TLV 0x91 while NAS was expecting P=
ORT-UP TLV 0x04 sub-TLV 0x90
in both cases NAS just silently discarded the packet.
I'm just asking if there is interest in trying to improve the protocol to a=
ddress these kind of corner cases or if you think it is not useful or it's =
too complex

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

[RM] Agree.

Regards,
Roberta

Regards

Thomas

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some messages
> the response has the same message type and for others the response is a
> different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear
> to me as it is explained in the ancp spec.  The value NoSuccessAck is
> not good for OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to RFC3292
> and see that value 6 is used in case an event message expects a response
> but the rfc does not mention what value has to be used in case the event
> message does not expect a response.  The value 0 looks a reasonable
> choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".)
In this case the NAS just discards the message but the AN does not have any=
 way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6
> so value 0 is good as well.  ANCP spec is further restricting that it
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages
the response has the same message type and for others the response is a
different message type?  the meaning of ackall in oam looks quite clear
to me as it is explained in the ancp spec.  The value NoSuccessAck is
not good for OAM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292
and see that value 6 is used in case an event message expects a response
but the rfc does not mention what value has to be used in case the event
message does not expect a response.  The value 0 looks a reasonable
choice to me. it means that the field is not used.  Actually, RFC3292
seems to suggest that for port up messages no response is sent unless
the result field is set to 6 in the port up.  This means that when no
response is expected like in ancp, any value is allowed besides from 6
so value 0 is good as well.  ANCP spec is further restricting that it
has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header
fields that may deserve some additional clarifications in ANCP protocol
spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP
design, it could be worth to rationalize the use of Result field in case
of ANCP. In particular I'd like to identify/clarify which are the most
useful codes for this field.

Following the proposal presented by the design team for Multicast
extension the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a
reliable transport protocol NAck appears to be the most suitable
candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of
implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case
of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use
case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify
the code for Result. Does anybody know the Result code for this
scenario?

The use of Result field 0 looks quite strange to me as this code does
not belong to the range of values defined for Result field in RFC3292
thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result
field has impacts on the messaging interaction mode in terms of
request/response (Nack versus AckAll.) Regarding messaging principles
RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type 2. A request
message type with a response sent in the same message type but a
different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and
having in mind also protocol flexibility for future extensions it would
seem reasonable to use the 1st alternative as it is much more easier to
understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback
on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of
Result codes for ANCP (deviations from GSMP), in particular if nobody
has a use-case that requires AckAll I would propose to remove the Ackall
behavior for Result field from the Request message and to use Nack for
all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response
message type) as the only mode for messaging interaction in ANCP and
discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        =

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 08:16:03 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 04F293A6CC1;
	Tue,  5 Aug 2008 08:16:03 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 77C2F3A6B7A
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 08:16:02 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.249
X-Spam-Level: 
X-Spam-Status: No, score=-2.249 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id Z2i2qq5J8RW7 for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 08:16:00 -0700 (PDT)
Received: from smail5.alcatel.fr (smail5.alcatel.fr [62.23.212.27])
	by core3.amsl.com (Postfix) with ESMTP id 2C27F3A689A
	for <ancp@ietf.org>; Tue,  5 Aug 2008 08:15:59 -0700 (PDT)
Received: from FRVELSBHS06.ad2.ad.alcatel.com (frvelsbhs06.ad2.ad.alcatel.com
	[155.132.6.78])
	by smail5.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m75FGTcp001566; 
	Tue, 5 Aug 2008 17:16:29 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.51]) by
	FRVELSBHS06.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Tue, 5 Aug 2008 17:16:29 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 17:16:28 +0200
Message-ID: <7A6D4815C5E8F74988784FEBD50F2F1E02158FC9@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A201294A2BE9@GRFMBX702RM001.griffon.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
Thread-Index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0AAB02dHAAFHqIA=
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>,
	<1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de>
	<A1F769BC58A8B146B2EEA818EAE052A201294A2BE9@GRFMBX702RM001.griffon.local>
From: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>
To: "Maglione Roberta" <roberta.maglione@telecomitalia.it>, <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 15:16:29.0796 (UTC)
	FILETIME=[3C937A40:01C8F70E]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.13
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Roberta,

What if the response is malformed?  Introducing a new message to detect whe=
ther something wrong is another message is not going to help if it is just =
for troubleshooting.  We should not introduce new messages just to detect b=
ugs because they might only introduce bugs.  Secondly, it is still not clea=
r to me what the AN should do with the response.  You can see on the NAS th=
at it received a malformed packet.  Actually, in the NAS MIB there is an ob=
ject showing the number of malformed packets that a NAS has received from a=
 particular access node (ancpNasCurrentSessionStatDiscardedMessages).  If y=
ou see this value incrementing, you know that the AN is misbehaving and you=
 can enable whatever tracing tools that are available on the NAS to further=
 pin down what is exactly going wrong.

If there is an existing value defined for the Result field that suites your=
 needs, then use that value.  If there is no such value, then define a new =
value.

regards,
Stefaan

-----Original Message-----
From: Maglione Roberta [mailto:roberta.maglione@telecomitalia.it] =

Sent: dinsdag 5 augustus 2008 17:03
To: Haag, T; DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles

Hi Thomas,
   please see inline.
Thanks,
Regards,
Roberta
________________________________________
From: Haag, T [Thomas.Haag@t-systems.com]
Sent: Tuesday, August 05, 2008 3:56 PM
To: Maglione Roberta; stefaan.de_cnodder@alcatel-lucent.be; ancp@ietf.org
Subject: AW: [ANCP] Result field use and ANCP messaging principles

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more.

[RM] Agree with you on this point that's why I'm saying that AckAll for Res=
ult does not scale, we should have a Nack behavior.

So the benefit is not clear to me for use case topology discovery.

[RM] I have seen the following scenarios:
1. AN and NAS implementing two different flavors of ANCP: AN was sending Po=
rt-UP with one label format, while NAS was expecting a label with another f=
ormat.
2. Other case: problems caused by deprecated TLVs same behavior AN was send=
ing PORT-UP with TLV 0x04 sub-TLV 0x91 while NAS was expecting PORT-UP TLV =
0x04 sub-TLV 0x90 in both cases NAS just silently discarded the packet.
I'm just asking if there is interest in trying to improve the protocol to a=
ddress these kind of corner cases or if you think it is not useful or it's =
too complex

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

[RM] Agree.

Regards,
Roberta

Regards

Thomas

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some =

> messages the response has the same message type and for others the =

> response is a different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear to me as it is =

> explained in the ancp spec.  The value NoSuccessAck is not good for =

> OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to =

> RFC3292 and see that value 6 is used in case an event message expects =

> a response but the rfc does not mention what value has to be used in =

> case the event message does not expect a response.  The value 0 looks =

> a reasonable choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless =

> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".) In this case the NAS just discards the messa=
ge but the AN does not have any way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6 =

> so value 0 is good as well.  ANCP spec is further restricting that it =

> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages the=
 response has the same message type and for others the response is a differ=
ent message type?  the meaning of ackall in oam looks quite clear to me as =
it is explained in the ancp spec.  The value NoSuccessAck is not good for O=
AM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292 and=
 see that value 6 is used in case an event message expects a response but t=
he rfc does not mention what value has to be used in case the event message=
 does not expect a response.  The value 0 looks a reasonable choice to me. =
it means that the field is not used.  Actually, RFC3292 seems to suggest th=
at for port up messages no response is sent unless the result field is set =
to 6 in the port up.  This means that when no response is expected like in =
ancp, any value is allowed besides from 6 so value 0 is good as well.  ANCP=
 spec is further restricting that it has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Mag=
lione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header fi=
elds that may deserve some additional clarifications in ANCP protocol spec;=
 an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP de=
sign, it could be worth to rationalize the use of Result field in case of A=
NCP. In particular I'd like to identify/clarify which are the most useful c=
odes for this field.

Following the proposal presented by the design team for Multicast extension=
 the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a reliabl=
e transport protocol NAck appears to be the most suitable candidate for def=
ault Result code.

Now according to both current ANCP protocol spec and a couple of implementa=
tions I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case of T=
opology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use ca=
se (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify the c=
ode for Result. Does anybody know the Result code for this scenario?

The use of Result field 0 looks quite strange to me as this code does not b=
elong to the range of values defined for Result field in RFC3292 thus I hav=
e a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result fi=
eld has impacts on the messaging interaction mode in terms of request/respo=
nse (Nack versus AckAll.) Regarding messaging principles RFC 3292 allowed f=
or two modes of messaging interaction:
1. A request message type and a response message type 2. A request message =
type with a response sent in the same message type but a different result a=
nd code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and hav=
ing in mind also protocol flexibility for future extensions it would seem r=
easonable to use the 1st alternative as it is much more easier to understan=
d and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback on =
the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of Res=
ult codes for ANCP (deviations from GSMP), in particular if nobody has a us=
e-case that requires AckAll I would propose to remove the Ackall behavior f=
or Result field from the Request message and to use Nack for all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response mess=
age type) as the only mode for messaging interaction in ANCP and discard mo=
de 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        =

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 08:31:17 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id C6A7A3A6CE7;
	Tue,  5 Aug 2008 08:31:17 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 4FB363A6CB9
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 08:31:16 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.599
X-Spam-Level: 
X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id ZdnVV87zOhI1 for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 08:31:14 -0700 (PDT)
Received: from exprod7og113.obsmtp.com (exprod7og113.obsmtp.com [64.18.2.179])
	by core3.amsl.com (Postfix) with ESMTP id 26EC43A6CF4
	for <ancp@ietf.org>; Tue,  5 Aug 2008 08:31:08 -0700 (PDT)
Received: from source ([66.129.228.6]) by exprod7ob113.postini.com
	([64.18.6.12]) with SMTP; Tue, 05 Aug 2008 08:31:27 PDT
Received: from p-emlb02-sac.jnpr.net ([66.129.254.47]) by p-emsmtp01.jnpr.net
	with Microsoft SMTPSVC(6.0.3790.3959); 
	Tue, 5 Aug 2008 08:31:38 -0700
Received: from antipi.jnpr.net ([10.10.2.34]) by p-emlb02-sac.jnpr.net with
	Microsoft SMTPSVC(6.0.3790.3959); Tue, 5 Aug 2008 08:31:38 -0700
Received: from proton.jnpr.net ([10.10.2.37]) by antipi.jnpr.net with
	Microsoft SMTPSVC(6.0.3790.1830); Tue, 5 Aug 2008 11:31:35 -0400
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 11:30:15 -0400
Message-ID: <9BD5D7887235424FA97DFC223CAE3C2813889DB3@proton.jnpr.net>
In-Reply-To: <7A6D4815C5E8F74988784FEBD50F2F1E02158FC9@FRVELSMBS22.ad2.ad.alcatel.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
Thread-Index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0AAB02dHAAFHqIAAAJ3KMA==
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com><A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>,
	<1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de><A1F769BC58A8B146B2EEA818EAE052A201294A2BE9@GRFMBX702RM001.griffon.local>
	<7A6D4815C5E8F74988784FEBD50F2F1E02158FC9@FRVELSMBS22.ad2.ad.alcatel.com>
From: "Sanjay Wadhwa" <swadhwa@juniper.net>
To: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>,
	"Maglione Roberta" <roberta.maglione@telecomitalia.it>, <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 15:31:35.0286 (UTC)
	FILETIME=[584A4960:01C8F710]
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Stefaan
  For certain error conditions it is useful for the side that detects it to=
 let the other end know of the exact error and automatically bring down the=
 adjacency. It is not necessary that the same operator/organization is mana=
ging both NAS and AN. Adding hooks to the protocol for suitable error notif=
ications and a well defined behavior when encountering errors seems useful.

-Sanjay

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of DE =
CNODDER Stefaan
Sent: Tuesday, August 05, 2008 11:16 AM
To: Maglione Roberta; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles


Roberta,

What if the response is malformed?  Introducing a new message to detect whe=
ther something wrong is another message is not going to help if it is just =
for troubleshooting.  We should not introduce new messages just to detect b=
ugs because they might only introduce bugs.  Secondly, it is still not clea=
r to me what the AN should do with the response.  You can see on the NAS th=
at it received a malformed packet.  Actually, in the NAS MIB there is an ob=
ject showing the number of malformed packets that a NAS has received from a=
 particular access node (ancpNasCurrentSessionStatDiscardedMessages).  If y=
ou see this value incrementing, you know that the AN is misbehaving and you=
 can enable whatever tracing tools that are available on the NAS to further=
 pin down what is exactly going wrong.

If there is an existing value defined for the Result field that suites your=
 needs, then use that value.  If there is no such value, then define a new =
value.

regards,
Stefaan

-----Original Message-----
From: Maglione Roberta [mailto:roberta.maglione@telecomitalia.it] =

Sent: dinsdag 5 augustus 2008 17:03
To: Haag, T; DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles

Hi Thomas,
   please see inline.
Thanks,
Regards,
Roberta
________________________________________
From: Haag, T [Thomas.Haag@t-systems.com]
Sent: Tuesday, August 05, 2008 3:56 PM
To: Maglione Roberta; stefaan.de_cnodder@alcatel-lucent.be; ancp@ietf.org
Subject: AW: [ANCP] Result field use and ANCP messaging principles

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more.

[RM] Agree with you on this point that's why I'm saying that AckAll for Res=
ult does not scale, we should have a Nack behavior.

So the benefit is not clear to me for use case topology discovery.

[RM] I have seen the following scenarios:
1. AN and NAS implementing two different flavors of ANCP: AN was sending Po=
rt-UP with one label format, while NAS was expecting a label with another f=
ormat.
2. Other case: problems caused by deprecated TLVs same behavior AN was send=
ing PORT-UP with TLV 0x04 sub-TLV 0x91 while NAS was expecting PORT-UP TLV =
0x04 sub-TLV 0x90 in both cases NAS just silently discarded the packet.
I'm just asking if there is interest in trying to improve the protocol to a=
ddress these kind of corner cases or if you think it is not useful or it's =
too complex

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

[RM] Agree.

Regards,
Roberta

Regards

Thomas

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some =

> messages the response has the same message type and for others the =

> response is a different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear to me as it is =

> explained in the ancp spec.  The value NoSuccessAck is not good for =

> OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to =

> RFC3292 and see that value 6 is used in case an event message expects =

> a response but the rfc does not mention what value has to be used in =

> case the event message does not expect a response.  The value 0 looks =

> a reasonable choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless =

> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".) In this case the NAS just discards the messa=
ge but the AN does not have any way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6 =

> so value 0 is good as well.  ANCP spec is further restricting that it =

> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages the=
 response has the same message type and for others the response is a differ=
ent message type?  the meaning of ackall in oam looks quite clear to me as =
it is explained in the ancp spec.  The value NoSuccessAck is not good for O=
AM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292 and=
 see that value 6 is used in case an event message expects a response but t=
he rfc does not mention what value has to be used in case the event message=
 does not expect a response.  The value 0 looks a reasonable choice to me. =
it means that the field is not used.  Actually, RFC3292 seems to suggest th=
at for port up messages no response is sent unless the result field is set =
to 6 in the port up.  This means that when no response is expected like in =
ancp, any value is allowed besides from 6 so value 0 is good as well.  ANCP=
 spec is further restricting that it has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Mag=
lione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header fi=
elds that may deserve some additional clarifications in ANCP protocol spec;=
 an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP de=
sign, it could be worth to rationalize the use of Result field in case of A=
NCP. In particular I'd like to identify/clarify which are the most useful c=
odes for this field.

Following the proposal presented by the design team for Multicast extension=
 the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a reliabl=
e transport protocol NAck appears to be the most suitable candidate for def=
ault Result code.

Now according to both current ANCP protocol spec and a couple of implementa=
tions I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case of T=
opology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use ca=
se (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify the c=
ode for Result. Does anybody know the Result code for this scenario?

The use of Result field 0 looks quite strange to me as this code does not b=
elong to the range of values defined for Result field in RFC3292 thus I hav=
e a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result fi=
eld has impacts on the messaging interaction mode in terms of request/respo=
nse (Nack versus AckAll.) Regarding messaging principles RFC 3292 allowed f=
or two modes of messaging interaction:
1. A request message type and a response message type 2. A request message =
type with a response sent in the same message type but a different result a=
nd code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and hav=
ing in mind also protocol flexibility for future extensions it would seem r=
easonable to use the 1st alternative as it is much more easier to understan=
d and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback on =
the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of Res=
ult codes for ANCP (deviations from GSMP), in particular if nobody has a us=
e-case that requires AckAll I would propose to remove the Ackall behavior f=
or Result field from the Request message and to use Nack for all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response mess=
age type) as the only mode for messaging interaction in ANCP and discard mo=
de 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        =

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 08:31:17 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id EC7FD3A6D18;
	Tue,  5 Aug 2008 08:31:17 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 9143E3A6CB9
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 08:31:16 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.599
X-Spam-Level: 
X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id rjDQNERv3RLJ for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 08:31:15 -0700 (PDT)
Received: from exprod7og112.obsmtp.com (exprod7og112.obsmtp.com [64.18.2.177])
	by core3.amsl.com (Postfix) with ESMTP id 232563A6CE7
	for <ancp@ietf.org>; Tue,  5 Aug 2008 08:31:11 -0700 (PDT)
Received: from source ([66.129.228.6]) by exprod7ob112.postini.com
	([64.18.6.12]) with SMTP; Tue, 05 Aug 2008 08:31:37 PDT
Received: from p-emlb01-sac.jnpr.net ([66.129.254.46]) by p-emsmtp01.jnpr.net
	with Microsoft SMTPSVC(6.0.3790.3959); 
	Tue, 5 Aug 2008 08:31:36 -0700
Received: from antipi.jnpr.net ([10.10.2.34]) by p-emlb01-sac.jnpr.net with
	Microsoft SMTPSVC(6.0.3790.3959); Tue, 5 Aug 2008 08:31:35 -0700
Received: from proton.jnpr.net ([10.10.2.37]) by antipi.jnpr.net with
	Microsoft SMTPSVC(6.0.3790.1830); Tue, 5 Aug 2008 11:31:34 -0400
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 11:24:29 -0400
Message-ID: <9BD5D7887235424FA97DFC223CAE3C2813889DA4@proton.jnpr.net>
In-Reply-To: <1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
Thread-Index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0AACZghA
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com><A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>
	<1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de>
From: "Sanjay Wadhwa" <swadhwa@juniper.net>
To: "Haag, T" <Thomas.Haag@t-systems.com>, <roberta.maglione@telecomitalia.it>,
	<stefaan.de_cnodder@alcatel-lucent.be>, <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 15:31:34.0411 (UTC)
	FILETIME=[57C4C5B0:01C8F710]
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Thomas
  Any change to existing spec will be with the constraint that we remain ba=
ckwards compatible. I don't believe the proposal is an "AckAll" for PORT-UP=
 messages that convey DSL line attributes, which as you rightly said will i=
mpose additional messaging and processing load on both BRAS and AN, and is =
not needed. Also, transaction-id is always set to 0 in PORT-UP message, whi=
ch means the sender doesn't have an easy way to relate the response to the =
request. =


However, it is useful to handle general "fatal" errors such as missing mand=
atory TLVs and sub-TLVs. We can handle these via a "generic" TBD ERROR mess=
age which results in tearing the adjacency down. =


In general, defining "AckAll" (always send a response), "NAck" (generate a =
response only in case of an error condition), and "No Response required" (w=
hich could be 0) as valid response types for ANCP messages is useful. These=
 values seem sufficient for current known use-cases.
OAM requests and "admission control" requests must always expect a response=
 back (could be implicit, but explicitly requiring AckAll to be set in the =
request doesn't cause any issue). For messages such as "replication control=
", "access control" or line configuration in general, having result field o=
ptionally set as "AckAll" or "NAck" allows for some flexibility. We can spe=
cify the default value for these use-cases.

Thanks
-Sanjay

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Haa=
g, T
Sent: Tuesday, August 05, 2008 9:56 AM
To: roberta.maglione@telecomitalia.it; stefaan.de_cnodder@alcatel-lucent.be=
; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more. So the benefit is not clear to me for use case topology d=
iscovery.

Mechanisms for connectivity control:  =

If we run adjacency we have keep alive between AN and BRAS. If we run TCP/I=
P we have flow control. I don't see an added value for using response messa=
ges because in case of failure the mechanism inside AN and / or BRAS failed.

Considering the OAM use case error conditions for OAM use case are defined =
too.

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

Regards
 =

Thomas
 =

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some messages
> the response has the same message type and for others the response is a
> different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear
> to me as it is explained in the ancp spec.  The value NoSuccessAck is
> not good for OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to RFC3292
> and see that value 6 is used in case an event message expects a response
> but the rfc does not mention what value has to be used in case the event
> message does not expect a response.  The value 0 looks a reasonable
> choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".)
In this case the NAS just discards the message but the AN does not have any=
 way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6
> so value 0 is good as well.  ANCP spec is further restricting that it
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages
the response has the same message type and for others the response is a
different message type?  the meaning of ackall in oam looks quite clear
to me as it is explained in the ancp spec.  The value NoSuccessAck is
not good for OAM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292
and see that value 6 is used in case an event message expects a response
but the rfc does not mention what value has to be used in case the event
message does not expect a response.  The value 0 looks a reasonable
choice to me. it means that the field is not used.  Actually, RFC3292
seems to suggest that for port up messages no response is sent unless
the result field is set to 6 in the port up.  This means that when no
response is expected like in ancp, any value is allowed besides from 6
so value 0 is good as well.  ANCP spec is further restricting that it
has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header
fields that may deserve some additional clarifications in ANCP protocol
spec; an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP
design, it could be worth to rationalize the use of Result field in case
of ANCP. In particular I'd like to identify/clarify which are the most
useful codes for this field.

Following the proposal presented by the design team for Multicast
extension the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a
reliable transport protocol NAck appears to be the most suitable
candidate for default Result code.

Now according to both current ANCP protocol spec and a couple of
implementations I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case
of Topology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use
case (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify
the code for Result. Does anybody know the Result code for this
scenario?

The use of Result field 0 looks quite strange to me as this code does
not belong to the range of values defined for Result field in RFC3292
thus I have a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result
field has impacts on the messaging interaction mode in terms of
request/response (Nack versus AckAll.) Regarding messaging principles
RFC 3292 allowed for two modes of messaging interaction:
1. A request message type and a response message type 2. A request
message type with a response sent in the same message type but a
different result and code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and
having in mind also protocol flexibility for future extensions it would
seem reasonable to use the 1st alternative as it is much more easier to
understand and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback
on the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of
Result codes for ANCP (deviations from GSMP), in particular if nobody
has a use-case that requires AckAll I would propose to remove the Ackall
behavior for Result field from the Request message and to use Nack for
all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response
message type) as the only mode for messaging interaction in ANCP and
discard mode 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        =

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Tue Aug  5 09:39:38 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 074AA3A6960;
	Tue,  5 Aug 2008 09:39:38 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 740373A6960
	for <ancp@core3.amsl.com>; Tue,  5 Aug 2008 09:39:37 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.249
X-Spam-Level: 
X-Spam-Status: No, score=-2.249 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id H2l43932HqoH for <ancp@core3.amsl.com>;
	Tue,  5 Aug 2008 09:39:35 -0700 (PDT)
Received: from smail5.alcatel.fr (smail5.alcatel.fr [62.23.212.27])
	by core3.amsl.com (Postfix) with ESMTP id 380BE3A6B83
	for <ancp@ietf.org>; Tue,  5 Aug 2008 09:39:30 -0700 (PDT)
Received: from FRVELSBHS06.ad2.ad.alcatel.com (frvelsbhs06.ad2.ad.alcatel.com
	[155.132.6.78])
	by smail5.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m75GdwBd029023; 
	Tue, 5 Aug 2008 18:39:58 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.51]) by
	FRVELSBHS06.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Tue, 5 Aug 2008 18:39:58 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Tue, 5 Aug 2008 18:39:57 +0200
Message-ID: <7A6D4815C5E8F74988784FEBD50F2F1E02159009@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <9BD5D7887235424FA97DFC223CAE3C2813889DA4@proton.jnpr.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
Thread-Index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0AACZghAAAOiXHA=
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com><A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local>
	<1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de>
	<9BD5D7887235424FA97DFC223CAE3C2813889DA4@proton.jnpr.net>
From: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>
To: "Sanjay Wadhwa" <swadhwa@juniper.net>, <ancp@ietf.org>
X-OriginalArrivalTime: 05 Aug 2008 16:39:58.0612 (UTC)
	FILETIME=[E6102540:01C8F719]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.13
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Such an error message looks more useful to me than defining responses to po=
rt up messages.  It can be sent before closing the session as a purely info=
rmative message indicating the reason of the session shutdown.  However, it=
 has nothing to do with AckAll or Nack in the Result field.  It can be sent=
 when it is zero as well, and can even be sent without receiving a message =
before (e.g., when the session goes down because the operator has set it to=
 admin down).  I guess the content of the error message is a purely informa=
tive human readable string.

Defining responses to port up messages will only create problems: the NAS t=
ells the access node that a mandatory TLV was missing but the AN will say t=
hat this is not true, so where is the problem?  Is there a bug in the AN, o=
r is there a bug in the NAS?  You still dont know.  In addition, this is a =
very specific case.  So many things can go wrong, and I hope that no error =
types are defined for each individual case that might go wrong.  A missing =
mandatory TLV is one out of many problems that might happen.

regards,
Stefaan

-----Original Message-----
From: Sanjay Wadhwa [mailto:swadhwa@juniper.net] =

Sent: dinsdag 5 augustus 2008 17:24
To: Haag, T; roberta.maglione@telecomitalia.it; DE CNODDER Stefaan; ancp@ie=
tf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles

Thomas
  Any change to existing spec will be with the constraint that we remain ba=
ckwards compatible. I don't believe the proposal is an "AckAll" for PORT-UP=
 messages that convey DSL line attributes, which as you rightly said will i=
mpose additional messaging and processing load on both BRAS and AN, and is =
not needed. Also, transaction-id is always set to 0 in PORT-UP message, whi=
ch means the sender doesn't have an easy way to relate the response to the =
request. =


However, it is useful to handle general "fatal" errors such as missing mand=
atory TLVs and sub-TLVs. We can handle these via a "generic" TBD ERROR mess=
age which results in tearing the adjacency down. =


In general, defining "AckAll" (always send a response), "NAck" (generate a =
response only in case of an error condition), and "No Response required" (w=
hich could be 0) as valid response types for ANCP messages is useful. These=
 values seem sufficient for current known use-cases.
OAM requests and "admission control" requests must always expect a response=
 back (could be implicit, but explicitly requiring AckAll to be set in the =
request doesn't cause any issue). For messages such as "replication control=
", "access control" or line configuration in general, having result field o=
ptionally set as "AckAll" or "NAck" allows for some flexibility. We can spe=
cify the default value for these use-cases.

Thanks
-Sanjay

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Haa=
g, T
Sent: Tuesday, August 05, 2008 9:56 AM
To: roberta.maglione@telecomitalia.it; stefaan.de_cnodder@alcatel-lucent.be=
; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more. So the benefit is not clear to me for use case topology d=
iscovery.

Mechanisms for connectivity control:  =

If we run adjacency we have keep alive between AN and BRAS. If we run TCP/I=
P we have flow control. I don't see an added value for using response messa=
ges because in case of failure the mechanism inside AN and / or BRAS failed.

Considering the OAM use case error conditions for OAM use case are defined =
too.

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

Regards
 =

Thomas
 =

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some =

> messages the response has the same message type and for others the =

> response is a different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear to me as it is =

> explained in the ancp spec.  The value NoSuccessAck is not good for =

> OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to =

> RFC3292 and see that value 6 is used in case an event message expects =

> a response but the rfc does not mention what value has to be used in =

> case the event message does not expect a response.  The value 0 looks =

> a reasonable choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless =

> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".) In this case the NAS just discards the messa=
ge but the AN does not have any way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6 =

> so value 0 is good as well.  ANCP spec is further restricting that it =

> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages the=
 response has the same message type and for others the response is a differ=
ent message type?  the meaning of ackall in oam looks quite clear to me as =
it is explained in the ancp spec.  The value NoSuccessAck is not good for O=
AM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292 and=
 see that value 6 is used in case an event message expects a response but t=
he rfc does not mention what value has to be used in case the event message=
 does not expect a response.  The value 0 looks a reasonable choice to me. =
it means that the field is not used.  Actually, RFC3292 seems to suggest th=
at for port up messages no response is sent unless the result field is set =
to 6 in the port up.  This means that when no response is expected like in =
ancp, any value is allowed besides from 6 so value 0 is good as well.  ANCP=
 spec is further restricting that it has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Mag=
lione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header fi=
elds that may deserve some additional clarifications in ANCP protocol spec;=
 an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP de=
sign, it could be worth to rationalize the use of Result field in case of A=
NCP. In particular I'd like to identify/clarify which are the most useful c=
odes for this field.

Following the proposal presented by the design team for Multicast extension=
 the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a reliabl=
e transport protocol NAck appears to be the most suitable candidate for def=
ault Result code.

Now according to both current ANCP protocol spec and a couple of implementa=
tions I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case of T=
opology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use ca=
se (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify the c=
ode for Result. Does anybody know the Result code for this scenario?

The use of Result field 0 looks quite strange to me as this code does not b=
elong to the range of values defined for Result field in RFC3292 thus I hav=
e a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result fi=
eld has impacts on the messaging interaction mode in terms of request/respo=
nse (Nack versus AckAll.) Regarding messaging principles RFC 3292 allowed f=
or two modes of messaging interaction:
1. A request message type and a response message type 2. A request message =
type with a response sent in the same message type but a different result a=
nd code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and hav=
ing in mind also protocol flexibility for future extensions it would seem r=
easonable to use the 1st alternative as it is much more easier to understan=
d and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback on =
the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of Res=
ult codes for ANCP (deviations from GSMP), in particular if nobody has a us=
e-case that requires AckAll I would propose to remove the Ackall behavior f=
or Result field from the Request message and to use Nack for all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response mess=
age type) as the only mode for messaging interaction in ANCP and discard mo=
de 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        =

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Wed Aug  6 01:33:53 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 3B95A28C130;
	Wed,  6 Aug 2008 01:33:53 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 6541D28C142
	for <ancp@core3.amsl.com>; Wed,  6 Aug 2008 01:33:52 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.719
X-Spam-Level: 
X-Spam-Status: No, score=-0.719 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id bmY37NrMfVxg for <ancp@core3.amsl.com>;
	Wed,  6 Aug 2008 01:33:50 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id F02D928C117
	for <ancp@ietf.org>; Wed,  6 Aug 2008 01:33:49 -0700 (PDT)
Received: from ptpxch008rm001.idc.cww.telecomitalia.it ([156.54.205.170]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Wed, 6 Aug 2008 10:34:21 +0200
Received: from GRFHUB705RM001.griffon.local ([10.19.3.69]) by
	ptpxch008rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Wed, 6 Aug 2008 10:34:20 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	GRFHUB705RM001.griffon.local ([10.19.3.69]) with mapi;
	Wed, 6 Aug 2008 10:34:19 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>,
	"Sanjay Wadhwa" <swadhwa@juniper.net>, <ancp@ietf.org>
Date: Wed, 6 Aug 2008 10:34:19 +0200
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
thread-index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0AACZghAAAOiXHAAIP1nGw==
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A201294A2BED@GRFMBX702RM001.griffon.local>
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com><A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local><1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de><9BD5D7887235424FA97DFC223CAE3C2813889DA4@proton.jnpr.net>,
	<7A6D4815C5E8F74988784FEBD50F2F1E02159009@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <7A6D4815C5E8F74988784FEBD50F2F1E02159009@FRVELSMBS22.ad2.ad.alcatel.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 06 Aug 2008 08:34:20.0245 (UTC)
	FILETIME=[38A7F050:01C8F79F]
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hello,
        I briefly summarize the initial outcomes of the discussion we had o=
n Result field and error conditions:

1. Result field:
NoResp :    No Response required:  Res=3D 0x0
NAck    :    response  required only in case of an error condition code Res=
=3D  0x1
AckAll   :    always send a response      Res=3D  0x2

Success:    Res=3D0x3
Failure  :    Res=3D0x4

These values seem sufficient for current known use-cases, thus I propose to=
 add a clarification in ANCP protocol I-D to explicitly say that  in case o=
f ANCP, Result field can only take one of these codes.

In particular in order to maintain backward compatibility:
- for topology discovery Result field is set 0  No Response required;
- OAM requests and "admission control" requests must always expect a respon=
se back:
        * this could be implicit,
        * explicitly requiring AckAll to be set in the request
Any preference on this point?

- for "multicast replication control", "access control" or line configurati=
on in general, result field can be optionally set as "AckAll" or "NAck". We=
 can specify the default value for these use-cases.


2. Generic Error Message

It seems to be interest in defining a new message type to be able to handle=
 error conditions.
Action Item for ANCP design team: specify this new Generic Error Message

According to the feedback currently received this message could be sent ind=
ependently of the result field value. Other opinions on this point?


3. As far as modes of messaging interaction:
- A request message type and a response message type
- A request message type with a response sent in the same message type but =
a different result and code fields setting.

We received one preference for a mixed mode: additional feedback/opinion on=
 this point?

Thanks
Best Regards,
Roberta





________________________________________
From: ancp-bounces@ietf.org [ancp-bounces@ietf.org] On Behalf Of DE CNODDER=
 Stefaan [stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 6:39 PM
To: Sanjay Wadhwa; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles

Such an error message looks more useful to me than defining responses to po=
rt up messages.  It can be sent before closing the session as a purely info=
rmative message indicating the reason of the session shutdown.  However, it=
 has nothing to do with AckAll or Nack in the Result field.  It can be sent=
 when it is zero as well, and can even be sent without receiving a message =
before (e.g., when the session goes down because the operator has set it to=
 admin down).  I guess the content of the error message is a purely informa=
tive human readable string.

Defining responses to port up messages will only create problems: the NAS t=
ells the access node that a mandatory TLV was missing but the AN will say t=
hat this is not true, so where is the problem?  Is there a bug in the AN, o=
r is there a bug in the NAS?  You still dont know.  In addition, this is a =
very specific case.  So many things can go wrong, and I hope that no error =
types are defined for each individual case that might go wrong.  A missing =
mandatory TLV is one out of many problems that might happen.

regards,
Stefaan

-----Original Message-----
From: Sanjay Wadhwa [mailto:swadhwa@juniper.net]
Sent: dinsdag 5 augustus 2008 17:24
To: Haag, T; roberta.maglione@telecomitalia.it; DE CNODDER Stefaan; ancp@ie=
tf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles

Thomas
  Any change to existing spec will be with the constraint that we remain ba=
ckwards compatible. I don't believe the proposal is an "AckAll" for PORT-UP=
 messages that convey DSL line attributes, which as you rightly said will i=
mpose additional messaging and processing load on both BRAS and AN, and is =
not needed. Also, transaction-id is always set to 0 in PORT-UP message, whi=
ch means the sender doesn't have an easy way to relate the response to the =
request.

However, it is useful to handle general "fatal" errors such as missing mand=
atory TLVs and sub-TLVs. We can handle these via a "generic" TBD ERROR mess=
age which results in tearing the adjacency down.

In general, defining "AckAll" (always send a response), "NAck" (generate a =
response only in case of an error condition), and "No Response required" (w=
hich could be 0) as valid response types for ANCP messages is useful. These=
 values seem sufficient for current known use-cases.
OAM requests and "admission control" requests must always expect a response=
 back (could be implicit, but explicitly requiring AckAll to be set in the =
request doesn't cause any issue). For messages such as "replication control=
", "access control" or line configuration in general, having result field o=
ptionally set as "AckAll" or "NAck" allows for some flexibility. We can spe=
cify the default value for these use-cases.

Thanks
-Sanjay

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Haa=
g, T
Sent: Tuesday, August 05, 2008 9:56 AM
To: roberta.maglione@telecomitalia.it; stefaan.de_cnodder@alcatel-lucent.be=
; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more. So the benefit is not clear to me for use case topology d=
iscovery.

Mechanisms for connectivity control:
If we run adjacency we have keep alive between AN and BRAS. If we run TCP/I=
P we have flow control. I don't see an added value for using response messa=
ges because in case of failure the mechanism inside AN and / or BRAS failed.

Considering the OAM use case error conditions for OAM use case are defined =
too.

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

Regards

Thomas

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some
> messages the response has the same message type and for others the
> response is a different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear to me as it is
> explained in the ancp spec.  The value NoSuccessAck is not good for
> OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to
> RFC3292 and see that value 6 is used in case an event message expects
> a response but the rfc does not mention what value has to be used in
> case the event message does not expect a response.  The value 0 looks
> a reasonable choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".) In this case the NAS just discards the messa=
ge but the AN does not have any way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6
> so value 0 is good as well.  ANCP spec is further restricting that it
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages the=
 response has the same message type and for others the response is a differ=
ent message type?  the meaning of ackall in oam looks quite clear to me as =
it is explained in the ancp spec.  The value NoSuccessAck is not good for O=
AM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292 and=
 see that value 6 is used in case an event message expects a response but t=
he rfc does not mention what value has to be used in case the event message=
 does not expect a response.  The value 0 looks a reasonable choice to me. =
it means that the field is not used.  Actually, RFC3292 seems to suggest th=
at for port up messages no response is sent unless the result field is set =
to 6 in the port up.  This means that when no response is expected like in =
ancp, any value is allowed besides from 6 so value 0 is good as well.  ANCP=
 spec is further restricting that it has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Mag=
lione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header fi=
elds that may deserve some additional clarifications in ANCP protocol spec;=
 an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP de=
sign, it could be worth to rationalize the use of Result field in case of A=
NCP. In particular I'd like to identify/clarify which are the most useful c=
odes for this field.

Following the proposal presented by the design team for Multicast extension=
 the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a reliabl=
e transport protocol NAck appears to be the most suitable candidate for def=
ault Result code.

Now according to both current ANCP protocol spec and a couple of implementa=
tions I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case of T=
opology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use ca=
se (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify the c=
ode for Result. Does anybody know the Result code for this scenario?

The use of Result field 0 looks quite strange to me as this code does not b=
elong to the range of values defined for Result field in RFC3292 thus I hav=
e a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result fi=
eld has impacts on the messaging interaction mode in terms of request/respo=
nse (Nack versus AckAll.) Regarding messaging principles RFC 3292 allowed f=
or two modes of messaging interaction:
1. A request message type and a response message type 2. A request message =
type with a response sent in the same message type but a different result a=
nd code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and hav=
ing in mind also protocol flexibility for future extensions it would seem r=
easonable to use the 1st alternative as it is much more easier to understan=
d and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback on =
the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of Res=
ult codes for ANCP (deviations from GSMP), in particular if nobody has a us=
e-case that requires AckAll I would propose to remove the Ackall behavior f=
or Result field from the Request message and to use Nack for all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response mess=
age type) as the only mode for messaging interaction in ANCP and discard mo=
de 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        =

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Wed Aug  6 01:42:31 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 11A063A6B24;
	Wed,  6 Aug 2008 01:42:31 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id F35EE3A6B38
	for <ancp@core3.amsl.com>; Wed,  6 Aug 2008 01:42:29 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.719
X-Spam-Level: 
X-Spam-Status: No, score=-0.719 tagged_above=-999 required=5
	tests=[AWL=-0.000, BAYES_00=-2.599, HELO_EQ_IT=0.635,
	HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id IFaNgUbCB5Dg for <ancp@core3.amsl.com>;
	Wed,  6 Aug 2008 01:42:29 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id E3C7B3A6B24
	for <ancp@ietf.org>; Wed,  6 Aug 2008 01:42:28 -0700 (PDT)
Received: from ptpxch007rm001.idc.cww.telecomitalia.it ([156.54.205.169]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Wed, 6 Aug 2008 10:42:29 +0200
Received: from grfhub703rm001.griffon.local ([10.19.3.10]) by
	ptpxch007rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Wed, 6 Aug 2008 10:42:29 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub703rm001.griffon.local ([10.19.3.10]) with mapi;
	Wed, 6 Aug 2008 10:42:28 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: <ancp@ietf.org>
Date: Wed, 6 Aug 2008 10:39:22 +0200
Thread-Topic: Multicast Admission Control with AN Bandwidth Delegation 
thread-index: AQHI96AkriBQdwMat0KR7FcVajDUPg==
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 06 Aug 2008 08:42:29.0002 (UTC)
	FILETIME=[5BFA56A0:01C8F7A0]
Subject: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hello,
    In Dublin we presented some enhancements to multicast use cases and requirements (draft-maglione-ancp-mcast-01) in particular we proposed to extend the concept of multicast flows that share the same admission control rules (aka flow-group) with a mechanism that we called Admission Control with AN bandwidth delegation.
Proposal would be to incorporate in next revision of the ANCP framework I-D the outcomes of the list discussion on this draft, so I'd like to invite those of you that have questions or comments to send us feedbacks on the list.

Thanks.
Best Regards,
Roberta


--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Wed Aug  6 08:12:40 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 6BF6B3A6BAC;
	Wed,  6 Aug 2008 08:12:40 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 6BC213A6A7A
	for <ancp@core3.amsl.com>; Wed,  6 Aug 2008 08:12:39 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.599
X-Spam-Level: 
X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id eNiOt3HZUrPL for <ancp@core3.amsl.com>;
	Wed,  6 Aug 2008 08:12:36 -0700 (PDT)
Received: from eci-iron1.ecitele.com (eci-iron1.ecitele.com [147.234.242.117])
	by core3.amsl.com (Postfix) with ESMTP id 8AD873A6BAC
	for <ancp@ietf.org>; Wed,  6 Aug 2008 08:12:35 -0700 (PDT)
Received: from unknown (HELO ILPTAM01.ecitele.com) ([147.234.244.44])
	by eci-iron1.ecitele.com with ESMTP; 06 Aug 2008 18:21:23 +0300
Received: from ilptexch01.ecitele.com ([172.31.244.40]) by
	ILPTAM01.ecitele.com with Microsoft SMTPSVC(6.0.3790.3959); 
	Wed, 6 Aug 2008 18:12:50 +0300
Received: from ILPTMAIL02.ecitele.com ([147.234.244.212]) by
	ilptexch01.ecitele.com ([172.31.244.40]) with mapi;
	Wed, 6 Aug 2008 18:12:50 +0300
From: Moti Morgenstern <Moti.Morgenstern@ecitele.com>
To: Maglione Roberta <roberta.maglione@telecomitalia.it>
Date: Wed, 6 Aug 2008 18:12:48 +0300
Thread-Topic: [ANCP] Result field use and ANCP messaging principles
Thread-Index: Acj2TBp5KI12dSqjS/ibsfunoZ2l4gAie23wAAMoSSAAB34S0AACZghAAAOiXHAAIP1nGwANFYOg
Message-ID: <741F3E4A466A1D439C616902C7F8A8C6934583E21B@ILPTMAIL02.ecitele.com>
References: <A1F769BC58A8B146B2EEA818EAE052A20129491EE2@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E02158CEC@FRVELSMBS22.ad2.ad.alcatel.com><A1F769BC58A8B146B2EEA818EAE052A20129491EE5@GRFMBX702RM001.griffon.local><1B6169C658325341A3B8066E23919E1C019539CA@S4DE8PSAANK.mitte.t-com.de><9BD5D7887235424FA97DFC223CAE3C2813889DA4@proton.jnpr.net>,
	<7A6D4815C5E8F74988784FEBD50F2F1E02159009@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A201294A2BED@GRFMBX702RM001.griffon.local>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A201294A2BED@GRFMBX702RM001.griffon.local>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 06 Aug 2008 15:12:50.0762 (UTC)
	FILETIME=[E46F4AA0:01C8F7D6]
Cc: "ancp@ietf.org" <ancp@ietf.org>
Subject: Re: [ANCP] Result field use and ANCP messaging principles
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Roberta,

We also prefer mode #2, i.e., use the same message type for both the reques=
t and the response, with different setting in various fields.

Regards,
Moti

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Mag=
lione Roberta
Sent: Wednesday, August 06, 2008 11:34 AM
To: DE CNODDER Stefaan; Sanjay Wadhwa; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles

Hello,
        I briefly summarize the initial outcomes of the discussion we had o=
n Result field and error conditions:

1. Result field:
NoResp :    No Response required:  Res=3D 0x0
NAck    :    response  required only in case of an error condition code Res=
=3D  0x1
AckAll   :    always send a response      Res=3D  0x2

Success:    Res=3D0x3
Failure  :    Res=3D0x4

These values seem sufficient for current known use-cases, thus I propose to=
 add a clarification in ANCP protocol I-D to explicitly say that  in case o=
f ANCP, Result field can only take one of these codes.

In particular in order to maintain backward compatibility:
- for topology discovery Result field is set 0  No Response required;
- OAM requests and "admission control" requests must always expect a respon=
se back:
        * this could be implicit,
        * explicitly requiring AckAll to be set in the request
Any preference on this point?

- for "multicast replication control", "access control" or line configurati=
on in general, result field can be optionally set as "AckAll" or "NAck". We=
 can specify the default value for these use-cases.


2. Generic Error Message

It seems to be interest in defining a new message type to be able to handle=
 error conditions.
Action Item for ANCP design team: specify this new Generic Error Message

According to the feedback currently received this message could be sent ind=
ependently of the result field value. Other opinions on this point?


3. As far as modes of messaging interaction:
- A request message type and a response message type
- A request message type with a response sent in the same message type but =
a different result and code fields setting.

We received one preference for a mixed mode: additional feedback/opinion on=
 this point?

Thanks
Best Regards,
Roberta





________________________________________
From: ancp-bounces@ietf.org [ancp-bounces@ietf.org] On Behalf Of DE CNODDER=
 Stefaan [stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 6:39 PM
To: Sanjay Wadhwa; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles

Such an error message looks more useful to me than defining responses to po=
rt up messages.  It can be sent before closing the session as a purely info=
rmative message indicating the reason of the session shutdown.  However, it=
 has nothing to do with AckAll or Nack in the Result field.  It can be sent=
 when it is zero as well, and can even be sent without receiving a message =
before (e.g., when the session goes down because the operator has set it to=
 admin down).  I guess the content of the error message is a purely informa=
tive human readable string.

Defining responses to port up messages will only create problems: the NAS t=
ells the access node that a mandatory TLV was missing but the AN will say t=
hat this is not true, so where is the problem?  Is there a bug in the AN, o=
r is there a bug in the NAS?  You still dont know.  In addition, this is a =
very specific case.  So many things can go wrong, and I hope that no error =
types are defined for each individual case that might go wrong.  A missing =
mandatory TLV is one out of many problems that might happen.

regards,
Stefaan

-----Original Message-----
From: Sanjay Wadhwa [mailto:swadhwa@juniper.net]
Sent: dinsdag 5 augustus 2008 17:24
To: Haag, T; roberta.maglione@telecomitalia.it; DE CNODDER Stefaan; ancp@ie=
tf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles

Thomas
  Any change to existing spec will be with the constraint that we remain ba=
ckwards compatible. I don't believe the proposal is an "AckAll" for PORT-UP=
 messages that convey DSL line attributes, which as you rightly said will i=
mpose additional messaging and processing load on both BRAS and AN, and is =
not needed. Also, transaction-id is always set to 0 in PORT-UP message, whi=
ch means the sender doesn't have an easy way to relate the response to the =
request.

However, it is useful to handle general "fatal" errors such as missing mand=
atory TLVs and sub-TLVs. We can handle these via a "generic" TBD ERROR mess=
age which results in tearing the adjacency down.

In general, defining "AckAll" (always send a response), "NAck" (generate a =
response only in case of an error condition), and "No Response required" (w=
hich could be 0) as valid response types for ANCP messages is useful. These=
 values seem sufficient for current known use-cases.
OAM requests and "admission control" requests must always expect a response=
 back (could be implicit, but explicitly requiring AckAll to be set in the =
request doesn't cause any issue). For messages such as "replication control=
", "access control" or line configuration in general, having result field o=
ptionally set as "AckAll" or "NAck" allows for some flexibility. We can spe=
cify the default value for these use-cases.

Thanks
-Sanjay

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Haa=
g, T
Sent: Tuesday, August 05, 2008 9:56 AM
To: roberta.maglione@telecomitalia.it; stefaan.de_cnodder@alcatel-lucent.be=
; ancp@ietf.org
Subject: Re: [ANCP] Result field use and ANCP messaging principles

Roberta, Stefaan,

please consider that in case of designing response messages a BRAS will get=
 a heavy burden of performance because in case of loosing adjacency to all =
AN of one domain (eg. due to aggregation failure/interupt)the BRAS has to r=
espond all port status messages of all AN and all ports which can be easily=
 100000 and more. So the benefit is not clear to me for use case topology d=
iscovery.

Mechanisms for connectivity control:
If we run adjacency we have keep alive between AN and BRAS. If we run TCP/I=
P we have flow control. I don't see an added value for using response messa=
ges because in case of failure the mechanism inside AN and / or BRAS failed.

Considering the OAM use case error conditions for OAM use case are defined =
too.

I would highly recommend also considering and keep existing implementations=
 also with respect to backward compatibility of existing solutions.

Regards

Thomas

-----Urspr=FCngliche Nachricht-----
Von: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] Im Auftrag von Ma=
glione Roberta
Gesendet: Dienstag, 5. August 2008 13:44
An: 'DE CNODDER Stefaan'; ancp@ietf.org
Betreff: Re: [ANCP] Result field use and ANCP messaging principles


Hello,

> Would it not be possible to have a mix of both modes: for some
> messages the response has the same message type and for others the
> response is a different message type?

I don't say it is not possible to have a mix of both modes for different us=
e-cases, but also thinking about future message designer it would seem clea=
ner to have one single approach for message principles for all use cases ra=
ther having to deal with different methods for different scenarios. In addi=
tion there could be cases where the same response message could be re-used =
across different scenarios.

> the meaning of ackall in oam looks quite clear to me as it is
> explained in the ancp spec.  The value NoSuccessAck is not good for
> OAM because a response is always expected.

Agree with you that in case of OAM a response is always expected but is the=
re a different way to convey that in case of OAM response is required? For =
example can OAM request message request/application be made to mean "always=
 reply"?

> value 0 is reserved but is nowhere explained.  I was looking to
> RFC3292 and see that value 6 is used in case an event message expects
> a response but the rfc does not mention what value has to be used in
> case the event message does not expect a response.  The value 0 looks
> a reasonable choice to me. it means that the field is not used.
> Actually, RFC3292
> seems to suggest that for port up messages no response is sent unless
> the result field is set to 6 in the port up.

Agree with you that we can guess from RFC3292 that in case of Port-UP no re=
sponse is sent unless result is set to 6, but as you mentioned RFC 3292 say=
s that ReturnReceipt can be used in Events to indicate that an acknowledgem=
ent is required for the message.   It is not clear to me if RetunrnReceipt =
means a behavior like AckAll or Nack

The reason why I'm asking is because today we have Port-Up with Result set =
to zero and implementations interpret this "0" as  "no response" this behav=
ior does not allow to identify error conditions. Let's consider for example=
 a scenario where NAS receives from the AN a Port-UP message where a mandat=
ory TLV is somehow missing or wrong; the ANCP protocol spec says that the N=
AS is supposed to discard this message ("if a TLV is not understood by the =
NAS, it is silently ignored".) In this case the NAS just discards the messa=
ge but the AN does not have any way to know the something wrong happened.
Now back to the original question today we have 0 meaning no response, if w=
e use result equal to 6 what do we get ackall or Nack? Don't you think it c=
ould be useful to have a sort of "Nack" behavior both for Port-UP and Port-=
Management (for line configuration) in order to be able to catch error cond=
itions?
In addition in case of the error condition the same response message could =
be re-used for different scenarios if we choose the philosophy "A request m=
essage type and a response message type".

Thanks,
Regards,
Roberta



> This means that when no
> response is expected like in ancp, any value is allowed besides from 6
> so value 0 is good as well.  ANCP spec is further restricting that it
> has to be zero.


-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Tuesday, August 05, 2008 10:59 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Result field use and ANCP messaging principles


Hi,

Would it not be possible to have a mix of both modes: for some messages the=
 response has the same message type and for others the response is a differ=
ent message type?  the meaning of ackall in oam looks quite clear to me as =
it is explained in the ancp spec.  The value NoSuccessAck is not good for O=
AM because a response is always expected.

Note following assignments by IANA

Result Value  Result Type Name                        Reference
------------  --------------------------------------  ---------
0             Reserved
1             NoSuccessAck                            [RFC3292]
2             AckAll                                  [RFC3292]
3             Success                                 [RFC3292]
4             Failure                                 [RFC3292]
5             More                                    [RFC3292]
6             ReturnReceipt                           [RFC3292]
7-255         Unassigned
value 0 is reserved but is nowhere explained.  I was looking to RFC3292 and=
 see that value 6 is used in case an event message expects a response but t=
he rfc does not mention what value has to be used in case the event message=
 does not expect a response.  The value 0 looks a reasonable choice to me. =
it means that the field is not used.  Actually, RFC3292 seems to suggest th=
at for port up messages no response is sent unless the result field is set =
to 6 in the port up.  This means that when no response is expected like in =
ancp, any value is allowed besides from 6 so value 0 is good as well.  ANCP=
 spec is further restricting that it has to be zero.

regards,
Stefaan


-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Mag=
lione Roberta
Sent: maandag 4 augustus 2008 18:07
To: ancp@ietf.org
Subject: [ANCP] Result field use and ANCP messaging principles


Hello,
    as part of GSMP legacy, ANCP has inherited some base protocol header fi=
elds that may deserve some additional clarifications in ANCP protocol spec;=
 an example of such a field is the Result.
RFC 3292 in section 3.1.1 defines 6 values for Result field:
NoSuccessAck:       Result =3D 1
AckAll:             Result =3D 2
Success:            Result =3D 3
Failure:            Result =3D 4
More:               Result =3D 5
ReturnReceipt       Result =3D 6

"NoSuccessAck" and  "AckAll" used in Request message; "Success," "More,"
and "Failure" used in Response message; and ReturnReceipt used in Events

As current ANCP spec, in some cases, already deviates from original GMSP de=
sign, it could be worth to rationalize the use of Result field in case of A=
NCP. In particular I'd like to identify/clarify which are the most useful c=
odes for this field.

Following the proposal presented by the design team for Multicast extension=
 the most useful codes for Result would seem to be Nack
(NoSuccessAck) 0x1 and AckAll 0x2, for request and Success 0x3 Failure
0x4 for response. Considering that WG selected to carry ANCP over a reliabl=
e transport protocol NAck appears to be the most suitable candidate for def=
ault Result code.

Now according to both current ANCP protocol spec and a couple of implementa=
tions I've seen the Result field is:
- set to 0 in the GSMP Event message with PORT UP message type in case of T=
opology Discovery  use case  (section 5.4.2)
- set to "AckAll"  in GSMP "Port Management" message, in case of OAM use ca=
se (section 5.4.4)
 - ???  in case of Line Configuration protocol draft does not specify the c=
ode for Result. Does anybody know the Result code for this scenario?

The use of Result field 0 looks quite strange to me as this code does not b=
elong to the range of values defined for Result field in RFC3292 thus I hav=
e a couple of questions:
- does your ANCP implementation pay attention to the Result code for
Topology   Discovery and OAM use-cases?
- What does "0" mean in Port-UP?  A Nack behavior perhaps?
- what does an "AckAll" in an OAM scenario mean?

The reason why I'm asking these questions is because the value of Result fi=
eld has impacts on the messaging interaction mode in terms of request/respo=
nse (Nack versus AckAll.) Regarding messaging principles RFC 3292 allowed f=
or two modes of messaging interaction:
1. A request message type and a response message type 2. A request message =
type with a response sent in the same message type but a different result a=
nd code fields setting.

Based on the OAM use case, and on the proposed multicast extensions and hav=
ing in mind also protocol flexibility for future extensions it would seem r=
easonable to use the 1st alternative as it is much more easier to understan=
d and simpler to be generalized.
Going on this direction I'd like to hear working group opinion/feedback on =
the following proposals:

1 - Add a clarification in ANCP protocol draft that explains the use of Res=
ult codes for ANCP (deviations from GSMP), in particular if nobody has a us=
e-case that requires AckAll I would propose to remove the Ackall behavior f=
or Result field from the Request message and to use Nack for all use-cases.

2 - Adopt mode of interaction 1 (A request message type and a response mess=
age type) as the only mode for messaging interaction in ANCP and discard mo=
de 2;

Thanks and best regards,
Roberta

---------------------------------------------
Roberta Maglione  - CCIE #18425
---------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above =
and may contain confidential information. If you have received the message =
in error, be informed that any use of the content hereof is prohibited. Ple=
ase return it immediately to the sender and delete the message. Should you =
have any questions, please contact us by replying to webmaster@telecomitali=
a.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 02:14:20 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 403123A688B;
	Fri, 22 Aug 2008 02:14:20 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id B22123A67A4
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 02:05:22 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.249
X-Spam-Level: 
X-Spam-Status: No, score=-2.249 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id DuTx9Iofs7LX for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 02:05:21 -0700 (PDT)
Received: from smail6.alcatel.fr (gc-na5.alcatel.fr [64.208.49.5])
	by core3.amsl.com (Postfix) with ESMTP id 4A9203A692A
	for <ancp@ietf.org>; Fri, 22 Aug 2008 02:05:20 -0700 (PDT)
Received: from FRVELSBHS03.ad2.ad.alcatel.com (frvelsbhs03.ad2.ad.alcatel.com
	[155.132.6.75])
	by smail6.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m7M8vkTB026535; 
	Fri, 22 Aug 2008 10:57:52 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.54]) by
	FRVELSBHS03.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Fri, 22 Aug 2008 10:57:51 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Fri, 22 Aug 2008 10:57:50 +0200
Message-ID: <7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
Thread-Index: AQHI96AkriBQdwMat0KR7FcVajDUPo52eu7A
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local>
From: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>
To: "Maglione Roberta" <roberta.maglione@telecomitalia.it>, <ancp@ietf.org>
X-OriginalArrivalTime: 22 Aug 2008 08:57:51.0557 (UTC)
	FILETIME=[28790F50:01C90435]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.84
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


 Hi Roberta,

Some comments on the multicast draft:

- Why is there a need to have a flag in the white/grey list to indicate
whether the AN should do admission control or not?  I would think that
in case of grey list, when the AN has sufficient delegated bandwidth,
and the AN has to necessary information to do the admission control,
then the AN does admission control.  Otherwise, the admission control is
part of the query to the NAS, which must contain an indication whether
or not the AN has already done admission control or not.  With the
current proposal, it might be that the AN does not perform admission
control, queries the NAS, but then the NAS has not sufficient BW and has
to go back to the AN to get some delegated BW back.  At the end, it does
not matter who is reserving the bandwidth, as long it gets reserved.
Both solutions have pros and cons, and I am not claiming that one is
better than the other, but I would think it is better to generalize the
proposal and remove the "Bandwidth Delegation" flag.  For white list,
there is also no need to have the flag: if the AN is provisioned with
the necessary information to do admission control, then the AN is doing
admission control, and otherwise there is no admission control.

- the text sometimes mentions "multicast bandwidth" and "unicast
bandwidth".  I would just make it "bandwidth", after all, it does not
matter whether the bandwidth is used for unicast or multicast on the
NAS.  At least, I assume that you are not proposing to split the
bandwidth in 2 parts: unicast and multicast BW?

- In the meeting in Dublin, it was said that this generalizes the
flow-group concept, but I do not see anymore where these flow-groups
are.  Is there still a grouping of flows?  Is there still a difference
between a user zapping from a channel to a channel is the same group,
and zapping from a channel to a channel belonging to a different group?

- 5th requirement (new text) in section 4.2: If the bandwidth delegation
flag is set, the AN first has to do admission control before querying
the NAS.  However, I would not specify the order and make it
implementation dependent.  The disadvantage of first allocating the BW
is that the BW is also reserved while the query is ongoing, and the
query might fail and then the BW has to be released again.  So doing
first admission control and then query can save messages, but it can
result in allocating BW while it was not really needed, and sessions
setup in parallel might fail because of no sufficient BW.  Other
implementations might be more intelligent: check BW, but does not
reserve, then query NAS, and after possitive answer from NAS, check BW
again and reserve it.

- 6th requirement (new text) in section 4.2: what is the difference
between a NAS querying the AN, and a NAS instructing to decrease the
delegated multicast bandwidth on a given port?

regards,
Stefaan

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: woensdag 6 augustus 2008 10:39
To: ancp@ietf.org
Subject: [ANCP] Multicast Admission Control with AN Bandwidth Delegation

Hello,
    In Dublin we presented some enhancements to multicast use cases and
requirements (draft-maglione-ancp-mcast-01) in particular we proposed to
extend the concept of multicast flows that share the same admission
control rules (aka flow-group) with a mechanism that we called Admission
Control with AN bandwidth delegation.
Proposal would be to incorporate in next revision of the ANCP framework
I-D the outcomes of the list discussion on this draft, so I'd like to
invite those of you that have questions or comments to send us feedbacks
on the list.

Thanks.
Best Regards,
Roberta


--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 04:03:06 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 020153A68A2;
	Fri, 22 Aug 2008 04:03:06 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 4BB443A6905
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 03:57:22 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 1.881
X-Spam-Level: *
X-Spam-Status: No, score=1.881 tagged_above=-999 required=5
	tests=[BAYES_50=0.001, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id 644eUz3e6wPl for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 03:57:21 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id 9A2893A67AE
	for <ancp@ietf.org>; Fri, 22 Aug 2008 03:57:20 -0700 (PDT)
Received: from ptpxch009rm001.idc.cww.telecomitalia.it ([156.54.205.171]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Fri, 22 Aug 2008 12:56:25 +0200
Received: from grfhub702rm001.griffon.local ([10.19.3.9]) by
	ptpxch009rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Fri, 22 Aug 2008 12:56:24 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub702rm001.griffon.local ([10.19.3.9]) with mapi;
	Fri, 22 Aug 2008 12:56:24 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "'DE CNODDER Stefaan'" <stefaan.de_cnodder@alcatel-lucent.be>,
	<ancp@ietf.org>
Date: Fri, 22 Aug 2008 12:56:24 +0200
Thread-Topic: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
thread-index: AQHI96AkriBQdwMat0KR7FcVajDUPo52eu7AgAAEAVA=
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local>
	<7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 22 Aug 2008 10:56:24.0423 (UTC)
	FILETIME=[B8124770:01C90445]
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Stefaan,
    thanks for your comments, please see answers inline [RM].

Regards,
Roberta

-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Friday, August 22, 2008 10:58 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth Delegation


 Hi Roberta,

Some comments on the multicast draft:

- Why is there a need to have a flag in the white/grey list to indicate
whether the AN should do admission control or not?  I would think that
in case of grey list, when the AN has sufficient delegated bandwidth,
and the AN has to necessary information to do the admission control,
then the AN does admission control.  Otherwise, the admission control is
part of the query to the NAS, which must contain an indication whether
or not the AN has already done admission control or not.  With the
current proposal, it might be that the AN does not perform admission
control, queries the NAS, but then the NAS has not sufficient BW and has
to go back to the AN to get some delegated BW back.  At the end, it does
not matter who is reserving the bandwidth, as long it gets reserved.
Both solutions have pros and cons, and I am not claiming that one is
better than the other, but I would think it is better to generalize the
proposal and remove the "Bandwidth Delegation" flag.

[RM]  My vision is slightly different; first of all I'd like to clarify that the purpose of this draft is to enhance one CAC approach currently defined in the framework introducing Multicast Admission with AN  Bandwidth Delegation mechanism but still keep the Multicast Admission without AN Bandwidth Delegation approach currently defined in the framework.
Now in grey list I can have three types of flows:
        1. Multicast flows that require some Conditional Access checks to be    performed by the NAS and CAC can be performed on AN with AN Bandwidth   Delegation;
        2. Multicast flows that require both Conditional Access checks and CAC to be    performed by the NAS.
        3. Multicast flows that do not require Conditional Access but that require CAC          to be performed by the NAS.

Do you agree that we need a way to differentiate case 1 from 2/3? The way we thought is to use the "Bandwidth Delegation" flag but if you think there is a better way to achieve to same goal I am interested in discussing it. In addition when CAC is done by NAS (cases 2 and 3) NAS does not delegate any portion of bandwidth to the AN, thus the scenario that you described (NAS does not have enough BW and has to query the AN) does not apply to these cases.

 For white list,
there is also no need to have the flag: if the AN is provisioned with
the necessary information to do admission control, then the AN is doing
admission control, and otherwise there is no admission control.

[RM] I am no not really sure about that, how do you handle multicast flows that do not require CAC?

- the text sometimes mentions "multicast bandwidth" and "unicast
bandwidth".  I would just make it "bandwidth", after all, it does not
matter whether the bandwidth is used for unicast or multicast on the
NAS.  At least, I assume that you are not proposing to split the
bandwidth in 2 parts: unicast and multicast BW?

[RM] In this proposal bandwidth delegated to AN is just to admit multicast flow not unicast VOD requests.

- In the meeting in Dublin, it was said that this generalizes the
flow-group concept, but I do not see anymore where these flow-groups
are.  Is there still a grouping of flows?  Is there still a difference
between a user zapping from a channel to a channel is the same group,
and zapping from a channel to a channel belonging to a different group?

[RM] The word " flow-group" does not appear in the current framework but the concept is there as " multicast flows that share the same rules" the proposal is to remove this concept for conditional access and to extend it with the "AN Bandwidth Delegation" mechanism in case of CAC.


- 5th requirement (new text) in section 4.2: If the bandwidth delegation
flag is set, the AN first has to do admission control before querying
the NAS.  However, I would not specify the order and make it
implementation dependent.  The disadvantage of first allocating the BW
is that the BW is also reserved while the query is ongoing, and the
query might fail and then the BW has to be released again.  So doing
first admission control and then query can save messages, but it can
result in allocating BW while it was not really needed, and sessions
setup in parallel might fail because of no sufficient BW.  Other
implementations might be more intelligent: check BW, but does not
reserve, then query NAS, and after possitive answer from NAS, check BW
again and reserve it.

[RM] I see pros and cons in both approaches, but if nobody else has additional opinions I'm fine with your suggestion to not specify the order and make it implementation dependent.

- 6th requirement (new text) in section 4.2: what is the difference
between a NAS querying the AN, and a NAS instructing to decrease the
delegated multicast bandwidth on a given port?

[RM] The difference is that with "instruct" the NAS can force the AN to release bandwidth, instead of just asking.

Regards,
Roberta

regards,
Stefaan

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: woensdag 6 augustus 2008 10:39
To: ancp@ietf.org
Subject: [ANCP] Multicast Admission Control with AN Bandwidth Delegation

Hello,
    In Dublin we presented some enhancements to multicast use cases and
requirements (draft-maglione-ancp-mcast-01) in particular we proposed to
extend the concept of multicast flows that share the same admission
control rules (aka flow-group) with a mechanism that we called Admission
Control with AN bandwidth delegation.
Proposal would be to incorporate in next revision of the ANCP framework
I-D the outcomes of the list discussion on this draft, so I'd like to
invite those of you that have questions or comments to send us feedbacks
on the list.

Thanks.
Best Regards,
Roberta


--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 04:33:07 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id F12A13A67AE;
	Fri, 22 Aug 2008 04:33:06 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id A05063A67EE
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 04:33:05 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.76
X-Spam-Level: 
X-Spam-Status: No, score=-0.76 tagged_above=-999 required=5
	tests=[BAYES_05=-1.11, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id 2vZLSt8D1TnA for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 04:33:04 -0700 (PDT)
Received: from smail6.alcatel.fr (colt-na5.alcatel.fr [62.23.212.5])
	by core3.amsl.com (Postfix) with ESMTP id B58913A67A5
	for <ancp@ietf.org>; Fri, 22 Aug 2008 04:33:03 -0700 (PDT)
Received: from FRVELSBHS03.ad2.ad.alcatel.com (frvelsbhs03.ad2.ad.alcatel.com
	[155.132.6.75])
	by smail6.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m7MBXAEh032580; 
	Fri, 22 Aug 2008 13:33:10 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.54]) by
	FRVELSBHS03.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Fri, 22 Aug 2008 13:33:10 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Fri, 22 Aug 2008 13:33:09 +0200
Message-ID: <7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
Thread-Index: AQHI96AkriBQdwMat0KR7FcVajDUPo52eu7AgAAEAVCAACNrkA==
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
From: "OOGHE Sven" <Sven.Ooghe@alcatel-lucent.be>
To: "Maglione Roberta" <roberta.maglione@telecomitalia.it>,
	"DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>,
	<ancp@ietf.org>
X-OriginalArrivalTime: 22 Aug 2008 11:33:10.0308 (UTC)
	FILETIME=[DAE19E40:01C9044A]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.84
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hello Roberta,

My understanding of the proposal of bandwidth delegation is to have a
mechanism that allows reducing the signaling load between the Access
Node and the NAS, and to enable fast acceptance or rejection of a
channel zap request in the Access Node. I'm supportive of that.

What puzzles me a bit is that in the case of grey lists, you seem to
decouple "conditional access" from "admission control", putting
"conditional access" in the NAS and "admission control" in the Access
Node. If I understand the grey list concept correctly, the Access Node
will ALWAYS signal to the NAS to know whether or not the request can be
approved. So whether or not admission control is performed on the Access
Node will have very limited effect on the overall message load. I would
argue that for grey lists it would be conceptually much simpler to
perform both "conditional access" as well as "admission control" in the
NAS.

So, in my view, bandwidth delegation makese sense, but I would apply it
only to white lists. 

The general behavior of bandwidth delegation as defined in the draft
looks fine to me.

Regards,
Sven

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: vrijdag 22 augustus 2008 12:56
To: DE CNODDER Stefaan; ancp@ietf.org
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hi Stefaan,
    thanks for your comments, please see answers inline [RM].

Regards,
Roberta

-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Friday, August 22, 2008 10:58 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation


 Hi Roberta,

Some comments on the multicast draft:

- Why is there a need to have a flag in the white/grey list to indicate
whether the AN should do admission control or not?  I would think that
in case of grey list, when the AN has sufficient delegated bandwidth,
and the AN has to necessary information to do the admission control,
then the AN does admission control.  Otherwise, the admission control is
part of the query to the NAS, which must contain an indication whether
or not the AN has already done admission control or not.  With the
current proposal, it might be that the AN does not perform admission
control, queries the NAS, but then the NAS has not sufficient BW and has
to go back to the AN to get some delegated BW back.  At the end, it does
not matter who is reserving the bandwidth, as long it gets reserved.
Both solutions have pros and cons, and I am not claiming that one is
better than the other, but I would think it is better to generalize the
proposal and remove the "Bandwidth Delegation" flag.

[RM]  My vision is slightly different; first of all I'd like to clarify
that the purpose of this draft is to enhance one CAC approach currently
defined in the framework introducing Multicast Admission with AN
Bandwidth Delegation mechanism but still keep the Multicast Admission
without AN Bandwidth Delegation approach currently defined in the
framework.
Now in grey list I can have three types of flows:
        1. Multicast flows that require some Conditional Access checks
to be    performed by the NAS and CAC can be performed on AN with AN
Bandwidth   Delegation;
        2. Multicast flows that require both Conditional Access checks
and CAC to be    performed by the NAS.
        3. Multicast flows that do not require Conditional Access but
that require CAC          to be performed by the NAS.

Do you agree that we need a way to differentiate case 1 from 2/3? The
way we thought is to use the "Bandwidth Delegation" flag but if you
think there is a better way to achieve to same goal I am interested in
discussing it. In addition when CAC is done by NAS (cases 2 and 3) NAS
does not delegate any portion of bandwidth to the AN, thus the scenario
that you described (NAS does not have enough BW and has to query the AN)
does not apply to these cases.

 For white list,
there is also no need to have the flag: if the AN is provisioned with
the necessary information to do admission control, then the AN is doing
admission control, and otherwise there is no admission control.

[RM] I am no not really sure about that, how do you handle multicast
flows that do not require CAC?

- the text sometimes mentions "multicast bandwidth" and "unicast
bandwidth".  I would just make it "bandwidth", after all, it does not
matter whether the bandwidth is used for unicast or multicast on the
NAS.  At least, I assume that you are not proposing to split the
bandwidth in 2 parts: unicast and multicast BW?

[RM] In this proposal bandwidth delegated to AN is just to admit
multicast flow not unicast VOD requests.

- In the meeting in Dublin, it was said that this generalizes the
flow-group concept, but I do not see anymore where these flow-groups
are.  Is there still a grouping of flows?  Is there still a difference
between a user zapping from a channel to a channel is the same group,
and zapping from a channel to a channel belonging to a different group?

[RM] The word " flow-group" does not appear in the current framework but
the concept is there as " multicast flows that share the same rules" the
proposal is to remove this concept for conditional access and to extend
it with the "AN Bandwidth Delegation" mechanism in case of CAC.


- 5th requirement (new text) in section 4.2: If the bandwidth delegation
flag is set, the AN first has to do admission control before querying
the NAS.  However, I would not specify the order and make it
implementation dependent.  The disadvantage of first allocating the BW
is that the BW is also reserved while the query is ongoing, and the
query might fail and then the BW has to be released again.  So doing
first admission control and then query can save messages, but it can
result in allocating BW while it was not really needed, and sessions
setup in parallel might fail because of no sufficient BW.  Other
implementations might be more intelligent: check BW, but does not
reserve, then query NAS, and after possitive answer from NAS, check BW
again and reserve it.

[RM] I see pros and cons in both approaches, but if nobody else has
additional opinions I'm fine with your suggestion to not specify the
order and make it implementation dependent.

- 6th requirement (new text) in section 4.2: what is the difference
between a NAS querying the AN, and a NAS instructing to decrease the
delegated multicast bandwidth on a given port?

[RM] The difference is that with "instruct" the NAS can force the AN to
release bandwidth, instead of just asking.

Regards,
Roberta

regards,
Stefaan

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: woensdag 6 augustus 2008 10:39
To: ancp@ietf.org
Subject: [ANCP] Multicast Admission Control with AN Bandwidth Delegation

Hello,
    In Dublin we presented some enhancements to multicast use cases and
requirements (draft-maglione-ancp-mcast-01) in particular we proposed to
extend the concept of multicast flows that share the same admission
control rules (aka flow-group) with a mechanism that we called Admission
Control with AN bandwidth delegation.
Proposal would be to incorporate in next revision of the ANCP framework
I-D the outcomes of the list discussion on this draft, so I'd like to
invite those of you that have questions or comments to send us feedbacks
on the list.

Thanks.
Best Regards,
Roberta


--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 05:05:06 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id A8FE53A694D;
	Fri, 22 Aug 2008 05:05:06 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id E29933A694D
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 05:00:39 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 1.788
X-Spam-Level: *
X-Spam-Status: No, score=1.788 tagged_above=-999 required=5 tests=[AWL=0.093, 
	BAYES_40=-0.185, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id dciln8yzbQ8K for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 05:00:38 -0700 (PDT)
Received: from mailm.telecomitalia.it (mailm.telecomitalia.it [217.169.121.53])
	by core3.amsl.com (Postfix) with ESMTP id DFE563A692A
	for <ancp@ietf.org>; Fri, 22 Aug 2008 05:00:37 -0700 (PDT)
Received: from ptpxch009rm001.idc.cww.telecomitalia.it ([156.54.205.171]) by
	mailm.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Fri, 22 Aug 2008 13:59:32 +0200
Received: from GRFHUB705RM001.griffon.local ([10.19.3.69]) by
	ptpxch009rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Fri, 22 Aug 2008 13:59:31 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	GRFHUB705RM001.griffon.local ([10.19.3.69]) with mapi;
	Fri, 22 Aug 2008 13:59:31 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "'OOGHE Sven'" <Sven.Ooghe@alcatel-lucent.be>,
	"DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>,
	<ancp@ietf.org>
Date: Fri, 22 Aug 2008 13:59:30 +0200
Thread-Topic: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
thread-index: AQHI96AkriBQdwMat0KR7FcVajDUPo52eu7AgAAEAVCAACNrkIAACQ0w
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491F23@GRFMBX702RM001.griffon.local>
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
	<7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 22 Aug 2008 11:59:31.0381 (UTC)
	FILETIME=[89463650:01C9044E]
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hello Sven,
    I see your point, but if for one user, most of the channels are in white list  and few of them are in grey then it seems reasonable for bandwidth accounting reasons to have CAC handled on the AN, with AN bandwidth delegation mechanism, for all the multicast flows (both in white and grey list.)

Thanks,
Regards
Roberta


-----Original Message-----
From: OOGHE Sven [mailto:Sven.Ooghe@alcatel-lucent.be]
Sent: Friday, August 22, 2008 1:33 PM
To: Maglione Roberta; DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth Delegation

Hello Roberta,

My understanding of the proposal of bandwidth delegation is to have a
mechanism that allows reducing the signaling load between the Access
Node and the NAS, and to enable fast acceptance or rejection of a
channel zap request in the Access Node. I'm supportive of that.

What puzzles me a bit is that in the case of grey lists, you seem to
decouple "conditional access" from "admission control", putting
"conditional access" in the NAS and "admission control" in the Access
Node. If I understand the grey list concept correctly, the Access Node
will ALWAYS signal to the NAS to know whether or not the request can be
approved. So whether or not admission control is performed on the Access
Node will have very limited effect on the overall message load. I would
argue that for grey lists it would be conceptually much simpler to
perform both "conditional access" as well as "admission control" in the
NAS.

So, in my view, bandwidth delegation makese sense, but I would apply it
only to white lists.

The general behavior of bandwidth delegation as defined in the draft
looks fine to me.

Regards,
Sven

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: vrijdag 22 augustus 2008 12:56
To: DE CNODDER Stefaan; ancp@ietf.org
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hi Stefaan,
    thanks for your comments, please see answers inline [RM].

Regards,
Roberta

-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Friday, August 22, 2008 10:58 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation


 Hi Roberta,

Some comments on the multicast draft:

- Why is there a need to have a flag in the white/grey list to indicate
whether the AN should do admission control or not?  I would think that
in case of grey list, when the AN has sufficient delegated bandwidth,
and the AN has to necessary information to do the admission control,
then the AN does admission control.  Otherwise, the admission control is
part of the query to the NAS, which must contain an indication whether
or not the AN has already done admission control or not.  With the
current proposal, it might be that the AN does not perform admission
control, queries the NAS, but then the NAS has not sufficient BW and has
to go back to the AN to get some delegated BW back.  At the end, it does
not matter who is reserving the bandwidth, as long it gets reserved.
Both solutions have pros and cons, and I am not claiming that one is
better than the other, but I would think it is better to generalize the
proposal and remove the "Bandwidth Delegation" flag.

[RM]  My vision is slightly different; first of all I'd like to clarify
that the purpose of this draft is to enhance one CAC approach currently
defined in the framework introducing Multicast Admission with AN
Bandwidth Delegation mechanism but still keep the Multicast Admission
without AN Bandwidth Delegation approach currently defined in the
framework.
Now in grey list I can have three types of flows:
        1. Multicast flows that require some Conditional Access checks
to be    performed by the NAS and CAC can be performed on AN with AN
Bandwidth   Delegation;
        2. Multicast flows that require both Conditional Access checks
and CAC to be    performed by the NAS.
        3. Multicast flows that do not require Conditional Access but
that require CAC          to be performed by the NAS.

Do you agree that we need a way to differentiate case 1 from 2/3? The
way we thought is to use the "Bandwidth Delegation" flag but if you
think there is a better way to achieve to same goal I am interested in
discussing it. In addition when CAC is done by NAS (cases 2 and 3) NAS
does not delegate any portion of bandwidth to the AN, thus the scenario
that you described (NAS does not have enough BW and has to query the AN)
does not apply to these cases.

 For white list,
there is also no need to have the flag: if the AN is provisioned with
the necessary information to do admission control, then the AN is doing
admission control, and otherwise there is no admission control.

[RM] I am no not really sure about that, how do you handle multicast
flows that do not require CAC?

- the text sometimes mentions "multicast bandwidth" and "unicast
bandwidth".  I would just make it "bandwidth", after all, it does not
matter whether the bandwidth is used for unicast or multicast on the
NAS.  At least, I assume that you are not proposing to split the
bandwidth in 2 parts: unicast and multicast BW?

[RM] In this proposal bandwidth delegated to AN is just to admit
multicast flow not unicast VOD requests.

- In the meeting in Dublin, it was said that this generalizes the
flow-group concept, but I do not see anymore where these flow-groups
are.  Is there still a grouping of flows?  Is there still a difference
between a user zapping from a channel to a channel is the same group,
and zapping from a channel to a channel belonging to a different group?

[RM] The word " flow-group" does not appear in the current framework but
the concept is there as " multicast flows that share the same rules" the
proposal is to remove this concept for conditional access and to extend
it with the "AN Bandwidth Delegation" mechanism in case of CAC.


- 5th requirement (new text) in section 4.2: If the bandwidth delegation
flag is set, the AN first has to do admission control before querying
the NAS.  However, I would not specify the order and make it
implementation dependent.  The disadvantage of first allocating the BW
is that the BW is also reserved while the query is ongoing, and the
query might fail and then the BW has to be released again.  So doing
first admission control and then query can save messages, but it can
result in allocating BW while it was not really needed, and sessions
setup in parallel might fail because of no sufficient BW.  Other
implementations might be more intelligent: check BW, but does not
reserve, then query NAS, and after possitive answer from NAS, check BW
again and reserve it.

[RM] I see pros and cons in both approaches, but if nobody else has
additional opinions I'm fine with your suggestion to not specify the
order and make it implementation dependent.

- 6th requirement (new text) in section 4.2: what is the difference
between a NAS querying the AN, and a NAS instructing to decrease the
delegated multicast bandwidth on a given port?

[RM] The difference is that with "instruct" the NAS can force the AN to
release bandwidth, instead of just asking.

Regards,
Roberta

regards,
Stefaan

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: woensdag 6 augustus 2008 10:39
To: ancp@ietf.org
Subject: [ANCP] Multicast Admission Control with AN Bandwidth Delegation

Hello,
    In Dublin we presented some enhancements to multicast use cases and
requirements (draft-maglione-ancp-mcast-01) in particular we proposed to
extend the concept of multicast flows that share the same admission
control rules (aka flow-group) with a mechanism that we called Admission
Control with AN bandwidth delegation.
Proposal would be to incorporate in next revision of the ANCP framework
I-D the outcomes of the list discussion on this draft, so I'd like to
invite those of you that have questions or comments to send us feedbacks
on the list.

Thanks.
Best Regards,
Roberta


--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 05:29:34 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id B43143A6A3F;
	Fri, 22 Aug 2008 05:27:07 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 1557E28C103
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 05:15:23 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.599
X-Spam-Level: 
X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[AWL=0.000, 
	BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id CgCostUp2O+c for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 05:15:21 -0700 (PDT)
Received: from ams-iport-1.cisco.com (ams-iport-1.cisco.com [144.254.224.140])
	by core3.amsl.com (Postfix) with ESMTP id DD73028C254
	for <ancp@ietf.org>; Fri, 22 Aug 2008 05:15:20 -0700 (PDT)
X-IronPort-AV: E=Sophos;i="4.32,251,1217808000"; d="scan'208";a="17941541"
Received: from ams-dkim-1.cisco.com ([144.254.224.138])
	by ams-iport-1.cisco.com with ESMTP; 22 Aug 2008 12:11:49 +0000
Received: from ams-core-1.cisco.com (ams-core-1.cisco.com [144.254.224.150])
	by ams-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m7MCBnDL007654; 
	Fri, 22 Aug 2008 14:11:49 +0200
Received: from xbh-ams-331.emea.cisco.com (xbh-ams-331.cisco.com
	[144.254.231.71])
	by ams-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m7MCBnqu007914;
	Fri, 22 Aug 2008 12:11:49 GMT
Received: from xfe-ams-331.emea.cisco.com ([144.254.231.72]) by
	xbh-ams-331.emea.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 22 Aug 2008 14:11:49 +0200
Received: from [144.254.53.136] ([144.254.53.136]) by
	xfe-ams-331.emea.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 22 Aug 2008 14:11:48 +0200
In-Reply-To: <7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
	<7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
Mime-Version: 1.0 (Apple Message framework v753.1)
Message-Id: <8BD954D6-D65F-4AF8-A4F5-9BA93BA81BBF@cisco.com>
From: Francois Le Faucheur IMAP <flefauch@cisco.com>
Date: Fri, 22 Aug 2008 14:11:46 +0200
To: OOGHE Sven <Sven.Ooghe@alcatel-lucent.be>
X-Mailer: Apple Mail (2.753.1)
X-OriginalArrivalTime: 22 Aug 2008 12:11:49.0023 (UTC)
	FILETIME=[40F172F0:01C90450]
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=10873; t=1219407109;
	x=1220271109; c=relaxed/simple; s=amsdkim1002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=flefauch@cisco.com;
	z=From:=20Francois=20Le=20Faucheur=20IMAP=20<flefauch@cisco.
	com>
	|Subject:=20Re=3A=20[ANCP]=20Multicast=20Admission=20Contro
	l=20with=20AN=20Bandwidth=20Delegation |Sender:=20;
	bh=PX25IOvcjS47VauGGeMnGOeTgLYxhgKeIouc76W5FDM=;
	b=d5Nl+SDDPg0Ax40+4VIL29XtrVlPup9H56PXq/sefkn4622CoBShkbxeTU
	7KNiR5bx/cKdCQ9ZTGUwOkzgngGgH+zgmgxY7IjwtkavDkTBoJkmvT58TV8o
	5P/iaV35I4;
Authentication-Results: ams-dkim-1; header.From=flefauch@cisco.com; dkim=pass (
	sig from cisco.com/amsdkim1002 verified; ); 
Cc: ancp@ietf.org
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset="us-ascii"; Format="flowed"; DelSp="yes"
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Sven,

On 22 Aug 2008, at 13:33, OOGHE Sven wrote:

> Hello Roberta,
>
> My understanding of the proposal of bandwidth delegation is to have a
> mechanism that allows reducing the signaling load between the Access
> Node and the NAS, and to enable fast acceptance or rejection of a
> channel zap request in the Access Node.

Agreed.

> I'm supportive of that.
>
> What puzzles me a bit is that in the case of grey lists, you seem to
> decouple "conditional access" from "admission control", putting
> "conditional access" in the NAS and "admission control" in the Access
> Node.

This is correct where AN Bandwidth Delegation is used.
Where Bandwidth Delegation is used, both Conditional Access and  
Admission Control are done by NAS.

> If I understand the grey list concept correctly, the Access Node
> will ALWAYS signal to the NAS to know whether or not the request  
> can be
> approved.

Right.

> So whether or not admission control is performed on the Access
> Node will have very limited effect on the overall message load. I  
> would
> argue that for grey lists it would be conceptually much simpler to
> perform both "conditional access" as well as "admission control" in  
> the
> NAS.

I think this is very similar to the comment Tom made in Dublin.
It is correct that for the flow in the Grey list, there will be the  
same amount of signaling whether we do admission control on AN or NAS.
The benefit of doing admission control on the AN (when Bandwidth  
Delegation is activated) is that it is also done on AN for the white  
groups. So we have a single entity performing admission control for  
all multicast flows (whether they are white or grey). This should  
simplify operations of multicast bandwidth management (particularly  
since the expectation is that there would be typically many more  
white groups than Grey groups).

>
> So, in my view, bandwidth delegation makese sense, but I would  
> apply it
> only to white lists.
>
> The general behavior of bandwidth delegation as defined in the draft
> looks fine to me.
>

Great.

Francois

> Regards,
> Sven
>
> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On  
> Behalf Of
> Maglione Roberta
> Sent: vrijdag 22 augustus 2008 12:56
> To: DE CNODDER Stefaan; ancp@ietf.org
> Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
> Delegation
>
> Hi Stefaan,
>     thanks for your comments, please see answers inline [RM].
>
> Regards,
> Roberta
>
> -----Original Message-----
> From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
> Sent: Friday, August 22, 2008 10:58 AM
> To: Maglione Roberta; ancp@ietf.org
> Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
> Delegation
>
>
>  Hi Roberta,
>
> Some comments on the multicast draft:
>
> - Why is there a need to have a flag in the white/grey list to  
> indicate
> whether the AN should do admission control or not?  I would think that
> in case of grey list, when the AN has sufficient delegated bandwidth,
> and the AN has to necessary information to do the admission control,
> then the AN does admission control.  Otherwise, the admission  
> control is
> part of the query to the NAS, which must contain an indication whether
> or not the AN has already done admission control or not.  With the
> current proposal, it might be that the AN does not perform admission
> control, queries the NAS, but then the NAS has not sufficient BW  
> and has
> to go back to the AN to get some delegated BW back.  At the end, it  
> does
> not matter who is reserving the bandwidth, as long it gets reserved.
> Both solutions have pros and cons, and I am not claiming that one is
> better than the other, but I would think it is better to generalize  
> the
> proposal and remove the "Bandwidth Delegation" flag.
>
> [RM]  My vision is slightly different; first of all I'd like to  
> clarify
> that the purpose of this draft is to enhance one CAC approach  
> currently
> defined in the framework introducing Multicast Admission with AN
> Bandwidth Delegation mechanism but still keep the Multicast Admission
> without AN Bandwidth Delegation approach currently defined in the
> framework.
> Now in grey list I can have three types of flows:
>         1. Multicast flows that require some Conditional Access checks
> to be    performed by the NAS and CAC can be performed on AN with AN
> Bandwidth   Delegation;
>         2. Multicast flows that require both Conditional Access checks
> and CAC to be    performed by the NAS.
>         3. Multicast flows that do not require Conditional Access but
> that require CAC          to be performed by the NAS.
>
> Do you agree that we need a way to differentiate case 1 from 2/3? The
> way we thought is to use the "Bandwidth Delegation" flag but if you
> think there is a better way to achieve to same goal I am interested in
> discussing it. In addition when CAC is done by NAS (cases 2 and 3) NAS
> does not delegate any portion of bandwidth to the AN, thus the  
> scenario
> that you described (NAS does not have enough BW and has to query  
> the AN)
> does not apply to these cases.
>
>  For white list,
> there is also no need to have the flag: if the AN is provisioned with
> the necessary information to do admission control, then the AN is  
> doing
> admission control, and otherwise there is no admission control.
>
> [RM] I am no not really sure about that, how do you handle multicast
> flows that do not require CAC?
>
> - the text sometimes mentions "multicast bandwidth" and "unicast
> bandwidth".  I would just make it "bandwidth", after all, it does not
> matter whether the bandwidth is used for unicast or multicast on the
> NAS.  At least, I assume that you are not proposing to split the
> bandwidth in 2 parts: unicast and multicast BW?
>
> [RM] In this proposal bandwidth delegated to AN is just to admit
> multicast flow not unicast VOD requests.
>
> - In the meeting in Dublin, it was said that this generalizes the
> flow-group concept, but I do not see anymore where these flow-groups
> are.  Is there still a grouping of flows?  Is there still a difference
> between a user zapping from a channel to a channel is the same group,
> and zapping from a channel to a channel belonging to a different  
> group?
>
> [RM] The word " flow-group" does not appear in the current  
> framework but
> the concept is there as " multicast flows that share the same  
> rules" the
> proposal is to remove this concept for conditional access and to  
> extend
> it with the "AN Bandwidth Delegation" mechanism in case of CAC.
>
>
> - 5th requirement (new text) in section 4.2: If the bandwidth  
> delegation
> flag is set, the AN first has to do admission control before querying
> the NAS.  However, I would not specify the order and make it
> implementation dependent.  The disadvantage of first allocating the BW
> is that the BW is also reserved while the query is ongoing, and the
> query might fail and then the BW has to be released again.  So doing
> first admission control and then query can save messages, but it can
> result in allocating BW while it was not really needed, and sessions
> setup in parallel might fail because of no sufficient BW.  Other
> implementations might be more intelligent: check BW, but does not
> reserve, then query NAS, and after possitive answer from NAS, check BW
> again and reserve it.
>
> [RM] I see pros and cons in both approaches, but if nobody else has
> additional opinions I'm fine with your suggestion to not specify the
> order and make it implementation dependent.
>
> - 6th requirement (new text) in section 4.2: what is the difference
> between a NAS querying the AN, and a NAS instructing to decrease the
> delegated multicast bandwidth on a given port?
>
> [RM] The difference is that with "instruct" the NAS can force the  
> AN to
> release bandwidth, instead of just asking.
>
> Regards,
> Roberta
>
> regards,
> Stefaan
>
> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On  
> Behalf Of
> Maglione Roberta
> Sent: woensdag 6 augustus 2008 10:39
> To: ancp@ietf.org
> Subject: [ANCP] Multicast Admission Control with AN Bandwidth  
> Delegation
>
> Hello,
>     In Dublin we presented some enhancements to multicast use cases  
> and
> requirements (draft-maglione-ancp-mcast-01) in particular we  
> proposed to
> extend the concept of multicast flows that share the same admission
> control rules (aka flow-group) with a mechanism that we called  
> Admission
> Control with AN bandwidth delegation.
> Proposal would be to incorporate in next revision of the ANCP  
> framework
> I-D the outcomes of the list discussion on this draft, so I'd like to
> invite those of you that have questions or comments to send us  
> feedbacks
> on the list.
>
> Thanks.
> Best Regards,
> Roberta
>
>
> --------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE
>
> This message and its attachments are addressed solely to the persons
> above and may contain confidential information. If you have  
> received the
> message in error, be informed that any use of the content hereof is
> prohibited. Please return it immediately to the sender and delete the
> message. Should you have any questions, please contact us by  
> replying to
> webmaster@telecomitalia.it.
>
>         Thank you
>
>                                         www.telecomitalia.it
>
> --------------------------------------------------------------------
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
> --------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE
>
> This message and its attachments are addressed solely to the persons
> above and may contain confidential information. If you have  
> received the
> message in error, be informed that any use of the content hereof is
> prohibited. Please return it immediately to the sender and delete the
> message. Should you have any questions, please contact us by  
> replying to
> webmaster@telecomitalia.it.
>
>         Thank you
>
>                                         www.telecomitalia.it
>
> --------------------------------------------------------------------
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 06:07:40 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 4BA5C3A6B44;
	Fri, 22 Aug 2008 06:07:36 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id AF20A28C3B0
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 05:47:08 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.249
X-Spam-Level: 
X-Spam-Status: No, score=-2.249 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id FF6lMAokQIdo for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 05:47:07 -0700 (PDT)
Received: from smail5.alcatel.fr (smail5.alcatel.fr [64.208.49.27])
	by core3.amsl.com (Postfix) with ESMTP id 9907828C3A4
	for <ancp@ietf.org>; Fri, 22 Aug 2008 05:47:06 -0700 (PDT)
Received: from FRVELSBHS07.ad2.ad.alcatel.com (frvelsbhs07.ad2.ad.alcatel.com
	[155.132.6.79])
	by smail5.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m7MCidD1002571; 
	Fri, 22 Aug 2008 14:44:39 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.54]) by
	FRVELSBHS07.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Fri, 22 Aug 2008 14:44:40 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Fri, 22 Aug 2008 14:44:38 +0200
Message-ID: <7A6D4815C5E8F74988784FEBD50F2F1E022364C8@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A20129491F23@GRFMBX702RM001.griffon.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
Thread-Index: AQHI96AkriBQdwMat0KR7FcVajDUPo52eu7AgAAEAVCAACNrkIAACQ0wgAANOhA=
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
	<7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F23@GRFMBX702RM001.griffon.local>
From: "DE CNODDER Stefaan" <stefaan.de_cnodder@alcatel-lucent.be>
To: "Maglione Roberta" <roberta.maglione@telecomitalia.it>,
	"OOGHE Sven" <Sven.Ooghe@alcatel-lucent.be>, <ancp@ietf.org>
X-OriginalArrivalTime: 22 Aug 2008 12:44:40.0003 (UTC)
	FILETIME=[D7BD2130:01C90454]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.13
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Hi Roberta,

Maybe I misunderstood something.  The "bandwidth delegation flag", is
this a flag per multicast stream or is it a general flag that applies to
all multicast streams?  I mean: is it possible to do CAC on NAS for some
flows and CAC on AN for some other flows, or is it always on one place?

regards,
Stefaan

-----Original Message-----
From: Maglione Roberta [mailto:roberta.maglione@telecomitalia.it] 
Sent: vrijdag 22 augustus 2008 14:00
To: OOGHE Sven; DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hello Sven,
    I see your point, but if for one user, most of the channels are in
white list  and few of them are in grey then it seems reasonable for
bandwidth accounting reasons to have CAC handled on the AN, with AN
bandwidth delegation mechanism, for all the multicast flows (both in
white and grey list.)

Thanks,
Regards
Roberta


-----Original Message-----
From: OOGHE Sven [mailto:Sven.Ooghe@alcatel-lucent.be]
Sent: Friday, August 22, 2008 1:33 PM
To: Maglione Roberta; DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hello Roberta,

My understanding of the proposal of bandwidth delegation is to have a
mechanism that allows reducing the signaling load between the Access
Node and the NAS, and to enable fast acceptance or rejection of a
channel zap request in the Access Node. I'm supportive of that.

What puzzles me a bit is that in the case of grey lists, you seem to
decouple "conditional access" from "admission control", putting
"conditional access" in the NAS and "admission control" in the Access
Node. If I understand the grey list concept correctly, the Access Node
will ALWAYS signal to the NAS to know whether or not the request can be
approved. So whether or not admission control is performed on the Access
Node will have very limited effect on the overall message load. I would
argue that for grey lists it would be conceptually much simpler to
perform both "conditional access" as well as "admission control" in the
NAS.

So, in my view, bandwidth delegation makese sense, but I would apply it
only to white lists.

The general behavior of bandwidth delegation as defined in the draft
looks fine to me.

Regards,
Sven

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: vrijdag 22 augustus 2008 12:56
To: DE CNODDER Stefaan; ancp@ietf.org
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hi Stefaan,
    thanks for your comments, please see answers inline [RM].

Regards,
Roberta

-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Friday, August 22, 2008 10:58 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation


 Hi Roberta,

Some comments on the multicast draft:

- Why is there a need to have a flag in the white/grey list to indicate
whether the AN should do admission control or not?  I would think that
in case of grey list, when the AN has sufficient delegated bandwidth,
and the AN has to necessary information to do the admission control,
then the AN does admission control.  Otherwise, the admission control is
part of the query to the NAS, which must contain an indication whether
or not the AN has already done admission control or not.  With the
current proposal, it might be that the AN does not perform admission
control, queries the NAS, but then the NAS has not sufficient BW and has
to go back to the AN to get some delegated BW back.  At the end, it does
not matter who is reserving the bandwidth, as long it gets reserved.
Both solutions have pros and cons, and I am not claiming that one is
better than the other, but I would think it is better to generalize the
proposal and remove the "Bandwidth Delegation" flag.

[RM]  My vision is slightly different; first of all I'd like to clarify
that the purpose of this draft is to enhance one CAC approach currently
defined in the framework introducing Multicast Admission with AN
Bandwidth Delegation mechanism but still keep the Multicast Admission
without AN Bandwidth Delegation approach currently defined in the
framework.
Now in grey list I can have three types of flows:
        1. Multicast flows that require some Conditional Access checks
to be    performed by the NAS and CAC can be performed on AN with AN
Bandwidth   Delegation;
        2. Multicast flows that require both Conditional Access checks
and CAC to be    performed by the NAS.
        3. Multicast flows that do not require Conditional Access but
that require CAC          to be performed by the NAS.

Do you agree that we need a way to differentiate case 1 from 2/3? The
way we thought is to use the "Bandwidth Delegation" flag but if you
think there is a better way to achieve to same goal I am interested in
discussing it. In addition when CAC is done by NAS (cases 2 and 3) NAS
does not delegate any portion of bandwidth to the AN, thus the scenario
that you described (NAS does not have enough BW and has to query the AN)
does not apply to these cases.

 For white list,
there is also no need to have the flag: if the AN is provisioned with
the necessary information to do admission control, then the AN is doing
admission control, and otherwise there is no admission control.

[RM] I am no not really sure about that, how do you handle multicast
flows that do not require CAC?

- the text sometimes mentions "multicast bandwidth" and "unicast
bandwidth".  I would just make it "bandwidth", after all, it does not
matter whether the bandwidth is used for unicast or multicast on the
NAS.  At least, I assume that you are not proposing to split the
bandwidth in 2 parts: unicast and multicast BW?

[RM] In this proposal bandwidth delegated to AN is just to admit
multicast flow not unicast VOD requests.

- In the meeting in Dublin, it was said that this generalizes the
flow-group concept, but I do not see anymore where these flow-groups
are.  Is there still a grouping of flows?  Is there still a difference
between a user zapping from a channel to a channel is the same group,
and zapping from a channel to a channel belonging to a different group?

[RM] The word " flow-group" does not appear in the current framework but
the concept is there as " multicast flows that share the same rules" the
proposal is to remove this concept for conditional access and to extend
it with the "AN Bandwidth Delegation" mechanism in case of CAC.


- 5th requirement (new text) in section 4.2: If the bandwidth delegation
flag is set, the AN first has to do admission control before querying
the NAS.  However, I would not specify the order and make it
implementation dependent.  The disadvantage of first allocating the BW
is that the BW is also reserved while the query is ongoing, and the
query might fail and then the BW has to be released again.  So doing
first admission control and then query can save messages, but it can
result in allocating BW while it was not really needed, and sessions
setup in parallel might fail because of no sufficient BW.  Other
implementations might be more intelligent: check BW, but does not
reserve, then query NAS, and after possitive answer from NAS, check BW
again and reserve it.

[RM] I see pros and cons in both approaches, but if nobody else has
additional opinions I'm fine with your suggestion to not specify the
order and make it implementation dependent.

- 6th requirement (new text) in section 4.2: what is the difference
between a NAS querying the AN, and a NAS instructing to decrease the
delegated multicast bandwidth on a given port?

[RM] The difference is that with "instruct" the NAS can force the AN to
release bandwidth, instead of just asking.

Regards,
Roberta

regards,
Stefaan

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: woensdag 6 augustus 2008 10:39
To: ancp@ietf.org
Subject: [ANCP] Multicast Admission Control with AN Bandwidth Delegation

Hello,
    In Dublin we presented some enhancements to multicast use cases and
requirements (draft-maglione-ancp-mcast-01) in particular we proposed to
extend the concept of multicast flows that share the same admission
control rules (aka flow-group) with a mechanism that we called Admission
Control with AN bandwidth delegation.
Proposal would be to incorporate in next revision of the ANCP framework
I-D the outcomes of the list discussion on this draft, so I'd like to
invite those of you that have questions or comments to send us feedbacks
on the list.

Thanks.
Best Regards,
Roberta


--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 06:07:43 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 795403A6BF8;
	Fri, 22 Aug 2008 06:07:43 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 034BD28C353
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 05:58:39 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.877
X-Spam-Level: 
X-Spam-Status: No, score=-1.877 tagged_above=-999 required=5 tests=[AWL=0.372, 
	BAYES_00=-2.599, HELO_EQ_FR=0.35]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id 1XKCsc2Xwj3J for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 05:58:37 -0700 (PDT)
Received: from smail6.alcatel.fr (gc-na5.alcatel.fr [64.208.49.5])
	by core3.amsl.com (Postfix) with ESMTP id 36CD928C365
	for <ancp@ietf.org>; Fri, 22 Aug 2008 05:58:36 -0700 (PDT)
Received: from FRVELSBHS03.ad2.ad.alcatel.com (frvelsbhs03.ad2.ad.alcatel.com
	[155.132.6.75])
	by smail6.alcatel.fr (8.13.8/8.13.8/ICT) with ESMTP id m7MCw6Ox022658
	for <ancp@ietf.org>; Fri, 22 Aug 2008 14:58:06 +0200
Received: from FRVELSMBS22.ad2.ad.alcatel.com ([155.132.6.54]) by
	FRVELSBHS03.ad2.ad.alcatel.com with Microsoft
	SMTPSVC(6.0.3790.2499); Fri, 22 Aug 2008 14:58:06 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Fri, 22 Aug 2008 14:58:05 +0200
Message-ID: <7168964CAC5C144685272595141B6DBA0101D4CB@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <7168964CAC5C144685272595141B6DBA0101D49E@FRVELSMBS22.ad2.ad.alcatel.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
Thread-Index: AckEUFMQLQJeXMYTSGuwcPI0jwsanQAAMRzgAAFl01A=
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
	<7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
	<8BD954D6-D65F-4AF8-A4F5-9BA93BA81BBF@cisco.com>
	<7168964CAC5C144685272595141B6DBA0101D49E@FRVELSMBS22.ad2.ad.alcatel.com>
From: "OOGHE Sven" <Sven.Ooghe@alcatel-lucent.be>
To: <ancp@ietf.org>
X-OriginalArrivalTime: 22 Aug 2008 12:58:06.0267 (UTC)
	FILETIME=[B84F40B0:01C90456]
X-Scanned-By: MIMEDefang 2.57 on 155.132.188.84
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Francois and Roberta,

When the Access Node is using bandwidth delegation for many flows in a
white list and there are only a few flows in the grey list, I can see it
makes sense to have bandwidth delegation for all multicast flows. It's
probably useful to clarify that in the framework. At least I didn't
consider that scenario when reading the draft.

Regards,
Sven

-----Original Message-----
From: Francois Le Faucheur IMAP [mailto:flefauch@cisco.com]
Sent: vrijdag 22 augustus 2008 14:12
To: OOGHE Sven
Cc: Francois Le Faucheur IMAP; Maglione Roberta; DE CNODDER Stefaan;
ancp@ietf.org
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hi Sven,

On 22 Aug 2008, at 13:33, OOGHE Sven wrote:

> Hello Roberta,
>
> My understanding of the proposal of bandwidth delegation is to have a 
> mechanism that allows reducing the signaling load between the Access 
> Node and the NAS, and to enable fast acceptance or rejection of a 
> channel zap request in the Access Node.

Agreed.

> I'm supportive of that.
>
> What puzzles me a bit is that in the case of grey lists, you seem to 
> decouple "conditional access" from "admission control", putting 
> "conditional access" in the NAS and "admission control" in the Access 
> Node.

This is correct where AN Bandwidth Delegation is used.
Where Bandwidth Delegation is used, both Conditional Access and
Admission Control are done by NAS.

> If I understand the grey list concept correctly, the Access Node will 
> ALWAYS signal to the NAS to know whether or not the request can be 
> approved.

Right.

> So whether or not admission control is performed on the Access Node 
> will have very limited effect on the overall message load. I would 
> argue that for grey lists it would be conceptually much simpler to 
> perform both "conditional access" as well as "admission control" in 
> the NAS.

I think this is very similar to the comment Tom made in Dublin.
It is correct that for the flow in the Grey list, there will be the same
amount of signaling whether we do admission control on AN or NAS.
The benefit of doing admission control on the AN (when Bandwidth
Delegation is activated) is that it is also done on AN for the white
groups. So we have a single entity performing admission control for all
multicast flows (whether they are white or grey). This should simplify
operations of multicast bandwidth management (particularly since the
expectation is that there would be typically many more white groups than
Grey groups).

>
> So, in my view, bandwidth delegation makese sense, but I would apply 
> it only to white lists.
>
> The general behavior of bandwidth delegation as defined in the draft 
> looks fine to me.
>

Great.

Francois

> Regards,
> Sven
>
> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf 
> Of Maglione Roberta
> Sent: vrijdag 22 augustus 2008 12:56
> To: DE CNODDER Stefaan; ancp@ietf.org
> Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth 
> Delegation
>
> Hi Stefaan,
>     thanks for your comments, please see answers inline [RM].
>
> Regards,
> Roberta
>
> -----Original Message-----
> From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
> Sent: Friday, August 22, 2008 10:58 AM
> To: Maglione Roberta; ancp@ietf.org
> Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth 
> Delegation
>
>
>  Hi Roberta,
>
> Some comments on the multicast draft:
>
> - Why is there a need to have a flag in the white/grey list to 
> indicate whether the AN should do admission control or not?  I would 
> think that in case of grey list, when the AN has sufficient delegated 
> bandwidth, and the AN has to necessary information to do the admission

> control, then the AN does admission control.  Otherwise, the admission

> control is part of the query to the NAS, which must contain an 
> indication whether or not the AN has already done admission control or

> not.  With the current proposal, it might be that the AN does not 
> perform admission control, queries the NAS, but then the NAS has not 
> sufficient BW and has to go back to the AN to get some delegated BW 
> back.  At the end, it does not matter who is reserving the bandwidth, 
> as long it gets reserved.
> Both solutions have pros and cons, and I am not claiming that one is 
> better than the other, but I would think it is better to generalize 
> the proposal and remove the "Bandwidth Delegation" flag.
>
> [RM]  My vision is slightly different; first of all I'd like to 
> clarify that the purpose of this draft is to enhance one CAC approach 
> currently defined in the framework introducing Multicast Admission 
> with AN Bandwidth Delegation mechanism but still keep the Multicast 
> Admission without AN Bandwidth Delegation approach currently defined 
> in the framework.
> Now in grey list I can have three types of flows:
>         1. Multicast flows that require some Conditional Access checks
> to be    performed by the NAS and CAC can be performed on AN with AN
> Bandwidth   Delegation;
>         2. Multicast flows that require both Conditional Access checks
> and CAC to be    performed by the NAS.
>         3. Multicast flows that do not require Conditional Access but
> that require CAC          to be performed by the NAS.
>
> Do you agree that we need a way to differentiate case 1 from 2/3? The 
> way we thought is to use the "Bandwidth Delegation" flag but if you 
> think there is a better way to achieve to same goal I am interested in

> discussing it. In addition when CAC is done by NAS (cases 2 and 3) NAS

> does not delegate any portion of bandwidth to the AN, thus the 
> scenario that you described (NAS does not have enough BW and has to 
> query the AN) does not apply to these cases.
>
>  For white list,
> there is also no need to have the flag: if the AN is provisioned with 
> the necessary information to do admission control, then the AN is 
> doing admission control, and otherwise there is no admission control.
>
> [RM] I am no not really sure about that, how do you handle multicast 
> flows that do not require CAC?
>
> - the text sometimes mentions "multicast bandwidth" and "unicast 
> bandwidth".  I would just make it "bandwidth", after all, it does not 
> matter whether the bandwidth is used for unicast or multicast on the 
> NAS.  At least, I assume that you are not proposing to split the 
> bandwidth in 2 parts: unicast and multicast BW?
>
> [RM] In this proposal bandwidth delegated to AN is just to admit 
> multicast flow not unicast VOD requests.
>
> - In the meeting in Dublin, it was said that this generalizes the 
> flow-group concept, but I do not see anymore where these flow-groups 
> are.  Is there still a grouping of flows?  Is there still a difference

> between a user zapping from a channel to a channel is the same group, 
> and zapping from a channel to a channel belonging to a different 
> group?
>
> [RM] The word " flow-group" does not appear in the current framework 
> but the concept is there as " multicast flows that share the same 
> rules" the proposal is to remove this concept for conditional access 
> and to extend it with the "AN Bandwidth Delegation" mechanism in case 
> of CAC.
>
>
> - 5th requirement (new text) in section 4.2: If the bandwidth 
> delegation flag is set, the AN first has to do admission control 
> before querying the NAS.  However, I would not specify the order and 
> make it implementation dependent.  The disadvantage of first 
> allocating the BW is that the BW is also reserved while the query is 
> ongoing, and the query might fail and then the BW has to be released 
> again.  So doing first admission control and then query can save 
> messages, but it can result in allocating BW while it was not really 
> needed, and sessions setup in parallel might fail because of no 
> sufficient BW.  Other implementations might be more intelligent: check

> BW, but does not reserve, then query NAS, and after possitive answer 
> from NAS, check BW again and reserve it.
>
> [RM] I see pros and cons in both approaches, but if nobody else has 
> additional opinions I'm fine with your suggestion to not specify the 
> order and make it implementation dependent.
>
> - 6th requirement (new text) in section 4.2: what is the difference 
> between a NAS querying the AN, and a NAS instructing to decrease the 
> delegated multicast bandwidth on a given port?
>
> [RM] The difference is that with "instruct" the NAS can force the AN 
> to release bandwidth, instead of just asking.
>
> Regards,
> Roberta
>
> regards,
> Stefaan
>
> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf 
> Of Maglione Roberta
> Sent: woensdag 6 augustus 2008 10:39
> To: ancp@ietf.org
> Subject: [ANCP] Multicast Admission Control with AN Bandwidth 
> Delegation
>
> Hello,
>     In Dublin we presented some enhancements to multicast use cases 
> and requirements (draft-maglione-ancp-mcast-01) in particular we 
> proposed to extend the concept of multicast flows that share the same 
> admission control rules (aka flow-group) with a mechanism that we 
> called Admission Control with AN bandwidth delegation.
> Proposal would be to incorporate in next revision of the ANCP 
> framework I-D the outcomes of the list discussion on this draft, so 
> I'd like to invite those of you that have questions or comments to 
> send us feedbacks on the list.
>
> Thanks.
> Best Regards,
> Roberta
>
>
> --------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE
>
> This message and its attachments are addressed solely to the persons 
> above and may contain confidential information. If you have received 
> the message in error, be informed that any use of the content hereof 
> is prohibited. Please return it immediately to the sender and delete 
> the message. Should you have any questions, please contact us by 
> replying to webmaster@telecomitalia.it.
>
>         Thank you
>
>                                         www.telecomitalia.it
>
> --------------------------------------------------------------------
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
> --------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE
>
> This message and its attachments are addressed solely to the persons 
> above and may contain confidential information. If you have received 
> the message in error, be informed that any use of the content hereof 
> is prohibited. Please return it immediately to the sender and delete 
> the message. Should you have any questions, please contact us by 
> replying to webmaster@telecomitalia.it.
>
>         Thank you
>
>                                         www.telecomitalia.it
>
> --------------------------------------------------------------------
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 06:32:38 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 39C213A6826;
	Fri, 22 Aug 2008 06:32:38 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id DA4F73A680D
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 06:17:16 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -6.599
X-Spam-Level: 
X-Spam-Status: No, score=-6.599 tagged_above=-999 required=5 tests=[AWL=0.000, 
	BAYES_00=-2.599, RCVD_IN_DNSWL_MED=-4]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id CWMFeSyYyZeD for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 06:17:15 -0700 (PDT)
Received: from ams-iport-1.cisco.com (ams-iport-1.cisco.com [144.254.224.140])
	by core3.amsl.com (Postfix) with ESMTP id 52E723A6768
	for <ancp@ietf.org>; Fri, 22 Aug 2008 06:17:14 -0700 (PDT)
X-IronPort-AV: E=Sophos;i="4.32,251,1217808000"; d="scan'208";a="17950764"
Received: from ams-dkim-1.cisco.com ([144.254.224.138])
	by ams-iport-1.cisco.com with ESMTP; 22 Aug 2008 13:16:50 +0000
Received: from ams-core-1.cisco.com (ams-core-1.cisco.com [144.254.224.150])
	by ams-dkim-1.cisco.com (8.12.11/8.12.11) with ESMTP id m7MDGoZ4030413; 
	Fri, 22 Aug 2008 15:16:50 +0200
Received: from xbh-ams-331.emea.cisco.com (xbh-ams-331.cisco.com
	[144.254.231.71])
	by ams-core-1.cisco.com (8.13.8/8.13.8) with ESMTP id m7MDGovO002996;
	Fri, 22 Aug 2008 13:16:50 GMT
Received: from xfe-ams-331.emea.cisco.com ([144.254.231.72]) by
	xbh-ams-331.emea.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 22 Aug 2008 15:16:50 +0200
Received: from [144.254.53.136] ([144.254.53.136]) by
	xfe-ams-331.emea.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 22 Aug 2008 15:16:49 +0200
In-Reply-To: <7168964CAC5C144685272595141B6DBA0101D4CB@FRVELSMBS22.ad2.ad.alcatel.com>
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
	<7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
	<8BD954D6-D65F-4AF8-A4F5-9BA93BA81BBF@cisco.com>
	<7168964CAC5C144685272595141B6DBA0101D49E@FRVELSMBS22.ad2.ad.alcatel.com>
	<7168964CAC5C144685272595141B6DBA0101D4CB@FRVELSMBS22.ad2.ad.alcatel.com>
Mime-Version: 1.0 (Apple Message framework v753.1)
Message-Id: <117DEC22-2FA3-41D6-98C6-8B55C7BFBB77@cisco.com>
From: Francois Le Faucheur IMAP <flefauch@cisco.com>
Date: Fri, 22 Aug 2008 15:16:47 +0200
To: "OOGHE Sven" <Sven.Ooghe@alcatel-lucent.be>
X-Mailer: Apple Mail (2.753.1)
X-OriginalArrivalTime: 22 Aug 2008 13:16:49.0707 (UTC)
	FILETIME=[55EE97B0:01C90459]
DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; l=12029; t=1219411010;
	x=1220275010; c=relaxed/simple; s=amsdkim1002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=flefauch@cisco.com;
	z=From:=20Francois=20Le=20Faucheur=20IMAP=20<flefauch@cisco.
	com>
	|Subject:=20Re=3A=20[ANCP]=20Multicast=20Admission=20Contro
	l=20with=20AN=20Bandwidth=20Delegation |Sender:=20;
	bh=MRYulfi4kg7zfV4vwLnxOxSDWHGI2lutnHmfjQ99oAw=;
	b=rmVKJRXZbjov3iTLozZDPRa9nzyplKlgR5fBNEhRi4r0K82mYBs91Pz2sH
	18wLa1+UOqG2z4tT6VX4gSK6H+tDTtEtMThlxnutQ+po1k6NS8JpqhRxw0SS
	KeardWAH51;
Authentication-Results: ams-dkim-1; header.From=flefauch@cisco.com; dkim=pass (
	sig from cisco.com/amsdkim1002 verified; ); 
Cc: ancp@ietf.org
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset="us-ascii"; Format="flowed"; DelSp="yes"
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


On 22 Aug 2008, at 14:58, OOGHE Sven wrote:

> Hi Francois and Roberta,
>
> When the Access Node is using bandwidth delegation for many flows in a
> white list and there are only a few flows in the grey list, I can  
> see it
> makes sense to have bandwidth delegation for all multicast flows. It's
> probably useful to clarify that in the framework.

I agree.
Francois

> At least I didn't
> consider that scenario when reading the draft.
>
> Regards,
> Sven
>
> -----Original Message-----
> From: Francois Le Faucheur IMAP [mailto:flefauch@cisco.com]
> Sent: vrijdag 22 augustus 2008 14:12
> To: OOGHE Sven
> Cc: Francois Le Faucheur IMAP; Maglione Roberta; DE CNODDER Stefaan;
> ancp@ietf.org
> Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
> Delegation
>
> Hi Sven,
>
> On 22 Aug 2008, at 13:33, OOGHE Sven wrote:
>
>> Hello Roberta,
>>
>> My understanding of the proposal of bandwidth delegation is to have a
>> mechanism that allows reducing the signaling load between the Access
>> Node and the NAS, and to enable fast acceptance or rejection of a
>> channel zap request in the Access Node.
>
> Agreed.
>
>> I'm supportive of that.
>>
>> What puzzles me a bit is that in the case of grey lists, you seem to
>> decouple "conditional access" from "admission control", putting
>> "conditional access" in the NAS and "admission control" in the Access
>> Node.
>
> This is correct where AN Bandwidth Delegation is used.
> Where Bandwidth Delegation is used, both Conditional Access and
> Admission Control are done by NAS.
>
>> If I understand the grey list concept correctly, the Access Node will
>> ALWAYS signal to the NAS to know whether or not the request can be
>> approved.
>
> Right.
>
>> So whether or not admission control is performed on the Access Node
>> will have very limited effect on the overall message load. I would
>> argue that for grey lists it would be conceptually much simpler to
>> perform both "conditional access" as well as "admission control" in
>> the NAS.
>
> I think this is very similar to the comment Tom made in Dublin.
> It is correct that for the flow in the Grey list, there will be the  
> same
> amount of signaling whether we do admission control on AN or NAS.
> The benefit of doing admission control on the AN (when Bandwidth
> Delegation is activated) is that it is also done on AN for the white
> groups. So we have a single entity performing admission control for  
> all
> multicast flows (whether they are white or grey). This should simplify
> operations of multicast bandwidth management (particularly since the
> expectation is that there would be typically many more white groups  
> than
> Grey groups).
>
>>
>> So, in my view, bandwidth delegation makese sense, but I would apply
>> it only to white lists.
>>
>> The general behavior of bandwidth delegation as defined in the draft
>> looks fine to me.
>>
>
> Great.
>
> Francois
>
>> Regards,
>> Sven
>>
>> -----Original Message-----
>> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf
>> Of Maglione Roberta
>> Sent: vrijdag 22 augustus 2008 12:56
>> To: DE CNODDER Stefaan; ancp@ietf.org
>> Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
>> Delegation
>>
>> Hi Stefaan,
>>     thanks for your comments, please see answers inline [RM].
>>
>> Regards,
>> Roberta
>>
>> -----Original Message-----
>> From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel- 
>> lucent.be]
>> Sent: Friday, August 22, 2008 10:58 AM
>> To: Maglione Roberta; ancp@ietf.org
>> Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
>> Delegation
>>
>>
>>  Hi Roberta,
>>
>> Some comments on the multicast draft:
>>
>> - Why is there a need to have a flag in the white/grey list to
>> indicate whether the AN should do admission control or not?  I would
>> think that in case of grey list, when the AN has sufficient delegated
>> bandwidth, and the AN has to necessary information to do the  
>> admission
>
>> control, then the AN does admission control.  Otherwise, the  
>> admission
>
>> control is part of the query to the NAS, which must contain an
>> indication whether or not the AN has already done admission  
>> control or
>
>> not.  With the current proposal, it might be that the AN does not
>> perform admission control, queries the NAS, but then the NAS has not
>> sufficient BW and has to go back to the AN to get some delegated BW
>> back.  At the end, it does not matter who is reserving the bandwidth,
>> as long it gets reserved.
>> Both solutions have pros and cons, and I am not claiming that one is
>> better than the other, but I would think it is better to generalize
>> the proposal and remove the "Bandwidth Delegation" flag.
>>
>> [RM]  My vision is slightly different; first of all I'd like to
>> clarify that the purpose of this draft is to enhance one CAC approach
>> currently defined in the framework introducing Multicast Admission
>> with AN Bandwidth Delegation mechanism but still keep the Multicast
>> Admission without AN Bandwidth Delegation approach currently defined
>> in the framework.
>> Now in grey list I can have three types of flows:
>>         1. Multicast flows that require some Conditional Access  
>> checks
>> to be    performed by the NAS and CAC can be performed on AN with AN
>> Bandwidth   Delegation;
>>         2. Multicast flows that require both Conditional Access  
>> checks
>> and CAC to be    performed by the NAS.
>>         3. Multicast flows that do not require Conditional Access but
>> that require CAC          to be performed by the NAS.
>>
>> Do you agree that we need a way to differentiate case 1 from 2/3? The
>> way we thought is to use the "Bandwidth Delegation" flag but if you
>> think there is a better way to achieve to same goal I am  
>> interested in
>
>> discussing it. In addition when CAC is done by NAS (cases 2 and 3)  
>> NAS
>
>> does not delegate any portion of bandwidth to the AN, thus the
>> scenario that you described (NAS does not have enough BW and has to
>> query the AN) does not apply to these cases.
>>
>>  For white list,
>> there is also no need to have the flag: if the AN is provisioned with
>> the necessary information to do admission control, then the AN is
>> doing admission control, and otherwise there is no admission control.
>>
>> [RM] I am no not really sure about that, how do you handle multicast
>> flows that do not require CAC?
>>
>> - the text sometimes mentions "multicast bandwidth" and "unicast
>> bandwidth".  I would just make it "bandwidth", after all, it does not
>> matter whether the bandwidth is used for unicast or multicast on the
>> NAS.  At least, I assume that you are not proposing to split the
>> bandwidth in 2 parts: unicast and multicast BW?
>>
>> [RM] In this proposal bandwidth delegated to AN is just to admit
>> multicast flow not unicast VOD requests.
>>
>> - In the meeting in Dublin, it was said that this generalizes the
>> flow-group concept, but I do not see anymore where these flow-groups
>> are.  Is there still a grouping of flows?  Is there still a  
>> difference
>
>> between a user zapping from a channel to a channel is the same group,
>> and zapping from a channel to a channel belonging to a different
>> group?
>>
>> [RM] The word " flow-group" does not appear in the current framework
>> but the concept is there as " multicast flows that share the same
>> rules" the proposal is to remove this concept for conditional access
>> and to extend it with the "AN Bandwidth Delegation" mechanism in case
>> of CAC.
>>
>>
>> - 5th requirement (new text) in section 4.2: If the bandwidth
>> delegation flag is set, the AN first has to do admission control
>> before querying the NAS.  However, I would not specify the order and
>> make it implementation dependent.  The disadvantage of first
>> allocating the BW is that the BW is also reserved while the query is
>> ongoing, and the query might fail and then the BW has to be released
>> again.  So doing first admission control and then query can save
>> messages, but it can result in allocating BW while it was not really
>> needed, and sessions setup in parallel might fail because of no
>> sufficient BW.  Other implementations might be more intelligent:  
>> check
>
>> BW, but does not reserve, then query NAS, and after possitive answer
>> from NAS, check BW again and reserve it.
>>
>> [RM] I see pros and cons in both approaches, but if nobody else has
>> additional opinions I'm fine with your suggestion to not specify the
>> order and make it implementation dependent.
>>
>> - 6th requirement (new text) in section 4.2: what is the difference
>> between a NAS querying the AN, and a NAS instructing to decrease the
>> delegated multicast bandwidth on a given port?
>>
>> [RM] The difference is that with "instruct" the NAS can force the AN
>> to release bandwidth, instead of just asking.
>>
>> Regards,
>> Roberta
>>
>> regards,
>> Stefaan
>>
>> -----Original Message-----
>> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf
>> Of Maglione Roberta
>> Sent: woensdag 6 augustus 2008 10:39
>> To: ancp@ietf.org
>> Subject: [ANCP] Multicast Admission Control with AN Bandwidth
>> Delegation
>>
>> Hello,
>>     In Dublin we presented some enhancements to multicast use cases
>> and requirements (draft-maglione-ancp-mcast-01) in particular we
>> proposed to extend the concept of multicast flows that share the same
>> admission control rules (aka flow-group) with a mechanism that we
>> called Admission Control with AN bandwidth delegation.
>> Proposal would be to incorporate in next revision of the ANCP
>> framework I-D the outcomes of the list discussion on this draft, so
>> I'd like to invite those of you that have questions or comments to
>> send us feedbacks on the list.
>>
>> Thanks.
>> Best Regards,
>> Roberta
>>
>>
>> --------------------------------------------------------------------
>>
>> CONFIDENTIALITY NOTICE
>>
>> This message and its attachments are addressed solely to the persons
>> above and may contain confidential information. If you have received
>> the message in error, be informed that any use of the content hereof
>> is prohibited. Please return it immediately to the sender and delete
>> the message. Should you have any questions, please contact us by
>> replying to webmaster@telecomitalia.it.
>>
>>         Thank you
>>
>>                                         www.telecomitalia.it
>>
>> --------------------------------------------------------------------
>>
>> _______________________________________________
>> ANCP mailing list
>> ANCP@ietf.org
>> https://www.ietf.org/mailman/listinfo/ancp
>> --------------------------------------------------------------------
>>
>> CONFIDENTIALITY NOTICE
>>
>> This message and its attachments are addressed solely to the persons
>> above and may contain confidential information. If you have received
>> the message in error, be informed that any use of the content hereof
>> is prohibited. Please return it immediately to the sender and delete
>> the message. Should you have any questions, please contact us by
>> replying to webmaster@telecomitalia.it.
>>
>>         Thank you
>>
>>                                         www.telecomitalia.it
>>
>> --------------------------------------------------------------------
>>
>> _______________________________________________
>> ANCP mailing list
>> ANCP@ietf.org
>> https://www.ietf.org/mailman/listinfo/ancp
>> _______________________________________________
>> ANCP mailing list
>> ANCP@ietf.org
>> https://www.ietf.org/mailman/listinfo/ancp
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 06:35:15 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 5B6773A6964;
	Fri, 22 Aug 2008 06:35:15 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 637B23A6AB5
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 06:29:10 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 0.534
X-Spam-Level: 
X-Spam-Status: No, score=0.534 tagged_above=-999 required=5 tests=[AWL=1.254, 
	BAYES_00=-2.599, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id omH9qZLc3LSJ for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 06:29:09 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id 647383A694D
	for <ancp@ietf.org>; Fri, 22 Aug 2008 06:29:08 -0700 (PDT)
Received: from ptpxch008rm001.idc.cww.telecomitalia.it ([156.54.205.170]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Fri, 22 Aug 2008 15:28:10 +0200
Received: from grfhub701rm001.griffon.local ([10.19.3.8]) by
	ptpxch008rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Fri, 22 Aug 2008 15:28:10 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub701rm001.griffon.local ([10.19.3.8]) with mapi;
	Fri, 22 Aug 2008 15:28:09 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "'DE CNODDER Stefaan'" <stefaan.de_cnodder@alcatel-lucent.be>,
	<ancp@ietf.org>
Date: Fri, 22 Aug 2008 15:28:09 +0200
Thread-Topic: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
thread-index: AQHI96AkriBQdwMat0KR7FcVajDUPo52eu7AgAAEAVCAACNrkIAACQ0wgAANOhCAAA1EYA==
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491F24@GRFMBX702RM001.griffon.local>
References: <A1F769BC58A8B146B2EEA818EAE052A201294A2BEE@GRFMBX702RM001.griffon.local><7A6D4815C5E8F74988784FEBD50F2F1E0223638D@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F22@GRFMBX702RM001.griffon.local>
	<7168964CAC5C144685272595141B6DBA0101D44F@FRVELSMBS22.ad2.ad.alcatel.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F23@GRFMBX702RM001.griffon.local>
	<7A6D4815C5E8F74988784FEBD50F2F1E022364C8@FRVELSMBS22.ad2.ad.alcatel.com>
In-Reply-To: <7A6D4815C5E8F74988784FEBD50F2F1E022364C8@FRVELSMBS22.ad2.ad.alcatel.com>
Accept-Language: en-US, it-IT
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US, it-IT
MIME-Version: 1.0
X-OriginalArrivalTime: 22 Aug 2008 13:28:10.0263 (UTC)
	FILETIME=[EB933270:01C9045A]
Cc: OOGHE Sven <Sven.Ooghe@alcatel-lucent.be>
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth Delegation
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

> The "bandwidth delegation flag", is
> this a flag per multicast stream or is it a general flag that applies to
> all multicast streams?

The flag is not for all multicast streams, it is per group of flows, according to the mask used in the ACL.

Regards,
Roberta

-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Friday, August 22, 2008 2:45 PM
To: Maglione Roberta; OOGHE Sven; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth Delegation


Hi Roberta,

Maybe I misunderstood something.  The "bandwidth delegation flag", is
this a flag per multicast stream or is it a general flag that applies to
all multicast streams?  I mean: is it possible to do CAC on NAS for some
flows and CAC on AN for some other flows, or is it always on one place?

regards,
Stefaan

-----Original Message-----
From: Maglione Roberta [mailto:roberta.maglione@telecomitalia.it]
Sent: vrijdag 22 augustus 2008 14:00
To: OOGHE Sven; DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hello Sven,
    I see your point, but if for one user, most of the channels are in
white list  and few of them are in grey then it seems reasonable for
bandwidth accounting reasons to have CAC handled on the AN, with AN
bandwidth delegation mechanism, for all the multicast flows (both in
white and grey list.)

Thanks,
Regards
Roberta


-----Original Message-----
From: OOGHE Sven [mailto:Sven.Ooghe@alcatel-lucent.be]
Sent: Friday, August 22, 2008 1:33 PM
To: Maglione Roberta; DE CNODDER Stefaan; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hello Roberta,

My understanding of the proposal of bandwidth delegation is to have a
mechanism that allows reducing the signaling load between the Access
Node and the NAS, and to enable fast acceptance or rejection of a
channel zap request in the Access Node. I'm supportive of that.

What puzzles me a bit is that in the case of grey lists, you seem to
decouple "conditional access" from "admission control", putting
"conditional access" in the NAS and "admission control" in the Access
Node. If I understand the grey list concept correctly, the Access Node
will ALWAYS signal to the NAS to know whether or not the request can be
approved. So whether or not admission control is performed on the Access
Node will have very limited effect on the overall message load. I would
argue that for grey lists it would be conceptually much simpler to
perform both "conditional access" as well as "admission control" in the
NAS.

So, in my view, bandwidth delegation makese sense, but I would apply it
only to white lists.

The general behavior of bandwidth delegation as defined in the draft
looks fine to me.

Regards,
Sven

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: vrijdag 22 augustus 2008 12:56
To: DE CNODDER Stefaan; ancp@ietf.org
Subject: Re: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation

Hi Stefaan,
    thanks for your comments, please see answers inline [RM].

Regards,
Roberta

-----Original Message-----
From: DE CNODDER Stefaan [mailto:stefaan.de_cnodder@alcatel-lucent.be]
Sent: Friday, August 22, 2008 10:58 AM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: [ANCP] Multicast Admission Control with AN Bandwidth
Delegation


 Hi Roberta,

Some comments on the multicast draft:

- Why is there a need to have a flag in the white/grey list to indicate
whether the AN should do admission control or not?  I would think that
in case of grey list, when the AN has sufficient delegated bandwidth,
and the AN has to necessary information to do the admission control,
then the AN does admission control.  Otherwise, the admission control is
part of the query to the NAS, which must contain an indication whether
or not the AN has already done admission control or not.  With the
current proposal, it might be that the AN does not perform admission
control, queries the NAS, but then the NAS has not sufficient BW and has
to go back to the AN to get some delegated BW back.  At the end, it does
not matter who is reserving the bandwidth, as long it gets reserved.
Both solutions have pros and cons, and I am not claiming that one is
better than the other, but I would think it is better to generalize the
proposal and remove the "Bandwidth Delegation" flag.

[RM]  My vision is slightly different; first of all I'd like to clarify
that the purpose of this draft is to enhance one CAC approach currently
defined in the framework introducing Multicast Admission with AN
Bandwidth Delegation mechanism but still keep the Multicast Admission
without AN Bandwidth Delegation approach currently defined in the
framework.
Now in grey list I can have three types of flows:
        1. Multicast flows that require some Conditional Access checks
to be    performed by the NAS and CAC can be performed on AN with AN
Bandwidth   Delegation;
        2. Multicast flows that require both Conditional Access checks
and CAC to be    performed by the NAS.
        3. Multicast flows that do not require Conditional Access but
that require CAC          to be performed by the NAS.

Do you agree that we need a way to differentiate case 1 from 2/3? The
way we thought is to use the "Bandwidth Delegation" flag but if you
think there is a better way to achieve to same goal I am interested in
discussing it. In addition when CAC is done by NAS (cases 2 and 3) NAS
does not delegate any portion of bandwidth to the AN, thus the scenario
that you described (NAS does not have enough BW and has to query the AN)
does not apply to these cases.

 For white list,
there is also no need to have the flag: if the AN is provisioned with
the necessary information to do admission control, then the AN is doing
admission control, and otherwise there is no admission control.

[RM] I am no not really sure about that, how do you handle multicast
flows that do not require CAC?

- the text sometimes mentions "multicast bandwidth" and "unicast
bandwidth".  I would just make it "bandwidth", after all, it does not
matter whether the bandwidth is used for unicast or multicast on the
NAS.  At least, I assume that you are not proposing to split the
bandwidth in 2 parts: unicast and multicast BW?

[RM] In this proposal bandwidth delegated to AN is just to admit
multicast flow not unicast VOD requests.

- In the meeting in Dublin, it was said that this generalizes the
flow-group concept, but I do not see anymore where these flow-groups
are.  Is there still a grouping of flows?  Is there still a difference
between a user zapping from a channel to a channel is the same group,
and zapping from a channel to a channel belonging to a different group?

[RM] The word " flow-group" does not appear in the current framework but
the concept is there as " multicast flows that share the same rules" the
proposal is to remove this concept for conditional access and to extend
it with the "AN Bandwidth Delegation" mechanism in case of CAC.


- 5th requirement (new text) in section 4.2: If the bandwidth delegation
flag is set, the AN first has to do admission control before querying
the NAS.  However, I would not specify the order and make it
implementation dependent.  The disadvantage of first allocating the BW
is that the BW is also reserved while the query is ongoing, and the
query might fail and then the BW has to be released again.  So doing
first admission control and then query can save messages, but it can
result in allocating BW while it was not really needed, and sessions
setup in parallel might fail because of no sufficient BW.  Other
implementations might be more intelligent: check BW, but does not
reserve, then query NAS, and after possitive answer from NAS, check BW
again and reserve it.

[RM] I see pros and cons in both approaches, but if nobody else has
additional opinions I'm fine with your suggestion to not specify the
order and make it implementation dependent.

- 6th requirement (new text) in section 4.2: what is the difference
between a NAS querying the AN, and a NAS instructing to decrease the
delegated multicast bandwidth on a given port?

[RM] The difference is that with "instruct" the NAS can force the AN to
release bandwidth, instead of just asking.

Regards,
Roberta

regards,
Stefaan

-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of
Maglione Roberta
Sent: woensdag 6 augustus 2008 10:39
To: ancp@ietf.org
Subject: [ANCP] Multicast Admission Control with AN Bandwidth Delegation

Hello,
    In Dublin we presented some enhancements to multicast use cases and
requirements (draft-maglione-ancp-mcast-01) in particular we proposed to
extend the concept of multicast flows that share the same admission
control rules (aka flow-group) with a mechanism that we called Admission
Control with AN bandwidth delegation.
Proposal would be to incorporate in next revision of the ANCP framework
I-D the outcomes of the list discussion on this draft, so I'd like to
invite those of you that have questions or comments to send us feedbacks
on the list.

Thanks.
Best Regards,
Roberta


--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons
above and may contain confidential information. If you have received the
message in error, be informed that any use of the content hereof is
prohibited. Please return it immediately to the sender and delete the
message. Should you have any questions, please contact us by replying to
webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------

--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 07:55:46 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id C357D3A6803;
	Fri, 22 Aug 2008 07:55:46 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 789073A6803
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 07:52:46 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 0.117
X-Spam-Level: 
X-Spam-Status: No, score=0.117 tagged_above=-999 required=5 tests=[AWL=0.836, 
	BAYES_00=-2.599, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id phNc-448a-Om for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 07:52:45 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id 215503A67A8
	for <ancp@ietf.org>; Fri, 22 Aug 2008 07:52:43 -0700 (PDT)
Received: from ptpxch008rm001.idc.cww.telecomitalia.it ([156.54.205.170]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Fri, 22 Aug 2008 16:52:39 +0200
Received: from grfhub702rm001.griffon.local ([10.19.3.9]) by
	ptpxch008rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Fri, 22 Aug 2008 16:52:39 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub702rm001.griffon.local ([10.19.3.9]) with mapi;
	Fri, 22 Aug 2008 16:52:39 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: <ancp@ietf.org>
Date: Fri, 22 Aug 2008 16:52:38 +0200
Thread-Topic: Re: [ANCP] On TLVs
thread-index: AcjyRs3clz+hA956R722OxvE/NRFgwAH4hjwACkDViAAJ6D2cAQvJmcA
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491F26@GRFMBX702RM001.griffon.local>
References: <48906A22.7070409@cisco.com><D9872168DBD43A41BD71FFC4713274D40579A4D8@xmb-ams-33b.emea.cisco.com><48C276B536524E478C659685995F6AA57C597A4E73@RBSJX.ad.redback.com>
	<D9872168DBD43A41BD71FFC4713274D40579AA4B@xmb-ams-33b.emea.cisco.com>
In-Reply-To: <D9872168DBD43A41BD71FFC4713274D40579AA4B@xmb-ams-33b.emea.cisco.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 22 Aug 2008 14:52:39.0343 (UTC)
	FILETIME=[B8FB6BF0:01C90466]
Subject: Re: [ANCP] On TLVs
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Hello,
     as Design Team is going to start working on other Multicast messages that may re-use Target Type TLV  we'd like to know if anybody has additional opinions/comments or would like have more clarifications on the proposed new TLV.

Thanks and best regards.
Roberta on behalf of the Design Team.



-----Original Message-----
From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On Behalf Of Wojciech Dec (wdec)
Sent: Friday, August 01, 2008 9:42 AM
To: Peter Arberg; Mark Townsley (townsley); ancp@ietf.org
Subject: Re: [ANCP] On TLVs

Hi,

> -----Original Message-----
> From: Peter Arberg [mailto:parberg@redback.com]
> Sent: 31 July 2008 15:03
> To: Wojciech Dec (wdec); Mark Townsley (townsley); ancp@ietf.org
> Subject: RE: Re: [ANCP] On TLVs
>
> Hi,
>
> my comments in the meeting was exactly what you write Mark.
> Woj and others have after the meeting explained the reasoning
> for the TLV, as they have expanded the transactional
> multicast case from being a 1 target (single dsl line) to
> potentially in the future be multiple targets enabled simultaneously.
>
> This can then with the TLV be done in 2 ways, either by
> including sub-tlvs for additional ACI identifiers, or add a
> new identifier which could be learned or configured by some
> today non identified mechanism.
>
> As such the TLV is in my mind okay if people in general buy
> into the usability of having multiple simultaneous targets
> for transactional multicast.

The re-use of such TLV is not just intended for transactional multicast,
but pretty much any (future) message type. Furthermore, it's not just a
case of multiple targets, but totally different target types, eg an ACL
is very much a target, and it's identifier (a string say) is not much
different from that of an ACI. It is very useful to have this traget
type concept in the protocol in order to have it extensible.

>
> In my mind I do not see a need for the capability of a NAS
> telling the AN to update the multicast forwarding state for
> many ports simultaneously, as individual clients in the
> multicast use-case discussed so far have to send individual
> joins, and as such it can not be a goal to be able to add
> many when they have to join individually.

The use-case for addressing multiple ports through a command (report) is
already in the framework/requirements doc.

>
> So as I see it at the moment, there seems to be an argument
> that a generic target TLV can be useful in other/many ANCP
> messages, but it is defined in a transactional multicast
> draft, without any explanation why it is suggested, and what
> it will solve.
>
> Also it is not clear to me what the authors have in mind for
> a use-case of multi targets in transactional multicast draft,
> which are trying to solve the framework requirements of the
> multicast use-case which is mainly focused around IPTV.
>
> So it will be nice to either have a better explanation why
> multiple targets are useful, if this is aimed at having the
> transactional multicast solution handle a future use-case
> where many clients can be added to a multicast stream which
> the NAS learns out-of-band what clients should be added, or
> if the intention is to add a capability for the NAS to delay
> multicast joins so they can be batched in the ancp message.
>
> In my mind the usage of this target TLV is just adding
> complexity to the transactional multicast communication, and
> it is in my mind not adding any significant value.

Granted, a better explanation is due and will be put in place. As
mentioned earlier, the complexity is trivial; it's just a TLV # wrapper,
ie an extra few bytes, nothing more.

Thanks,
Woj.

>
> Peter
>
> > -----Original Message-----
> > From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org]
> On Behalf
> > Of Wojciech Dec (wdec)
> > Sent: 30. juli 2008 18:32
> > To: Mark Townsley (townsley); ancp@ietf.org
> > Subject: Re: [ANCP] On TLVs
> >
> > Hi Mark,
> >
> > Let me go over what the rationale for this TLV was. You and others
> > feel free to comment/re-asses/advise.
> >
> > 1. The target type wrapper TLV
> > There is a very reasonable need to have a protocol wide (ie
> re-usable
> > across messages) container TLV that carries in its payload
> information
> > allowing the recipient to identify the target of a command.
> > There may be
> > different target type identifiers, besides different types
> of target
> > (these would need to be definied in the future).
> > The information that allows an access node to identify the traget
> > (access port) is currently the ACI, but others may get defined,
> > inluding things like group of ports say, or an access-list. If the
> > access port, and/or the ACI were the only type of object and
> > identifier respectively that we envisage commands to act
> upon, then we
> > can do away with the target type wrapper, gaining a few bytes of
> > efficency, at the cost of extensibility. Given though other
> types of
> > targets and identifiers are a definite possility, the wrapper TLV
> > would appear to be a small cost/complication to have.
> >
> > 2. Fixed position of target wrapper TLV It was thought that
> having the
> > target be the 1st information TLV in a multicast request
> message (the
> > same TLV could however be anywhere else in other messages)
> would allow
> > the recipient to be more efficent in determining whether
> the target is
> > valid. Processing any commands before identifying the target makes
> > little sense, and since a single message may contain N commands, it
> > appearead resonable to have the target be before the commands. That
> > said, nothing is set in stone, so if the WG is more
> comfortable with
> > allowing this TLV to be anywhere in the message, that's a
> change that
> > can be easily done.
> >
> > Regards,
> > Woj.
> >
> >
> > > -----Original Message-----
> > > From: ancp-bounces@ietf.org
> [mailto:ancp-bounces@ietf.org] On Behalf
> > > Of Mark Townsley (townsley)
> > > Sent: 30 July 2008 15:18
> > > To: ancp@ietf.org
> > > Subject: [ANCP] On TLVs
> > >
> > >
> > > Taking this to the list...  There was a discussion on
> whether a TLV
> > > was the right construct to use for something.
> > >
> > > First, TLVs are great, they allow extensibility, are
> proven protocol
> > > mechanics, etc. etc.
> > >
> > > But. If you restrict the TLV to be in a fixed location, always be
> > > present, and never be repeated, it effectively neuters all of the
> > > good reasons to have a TLV.
> > >
> > > For example:
> > >
> > > Type = X
> > > Length = Y
> > > Value = Information
> > >
> > > If this is always present, always present at the same
> location, and
> > > never repeated in a list of TLVs, you are effectively
> putting an "X"
> > > and "Y" in your protocol message which will always be present and
> > > adds no value at all. It's always there. If it is *always* there,
> > > and *always* the same, what's the point of it even
> existing? Just to
> > > see if one implementation can put it in the right place and the
> > > other can check to see if it is there?
> > >
> > > In actuality, you could just put "Information" the
> meesage as part
> > > of the fixed header that precedes it.
> > >
> > > It's really not a big deal at the end of the day, one can
> skip over
> > > "X and Y" when processing. But, it's one more thing to
> get wrong in
> > > terms of interoperability. i.e., make something *look*
> like a TLV,
> > > and some coder might expect it to *be* a TLV when it is not by
> > > mandate from the RFC.
> > >
> > > I speak from experience here. L2TP has its Message Type
> encoded as a
> > > TLV, with the same kinds of rules I heard proposed in the meeting
> > > today.
> > > It was a mistake, and we witnessed actual interop
> problems from it
> > > in the beginning. It either should have been a proper TLV placed
> > > anywhere, or a part of the fixed header at the start of
> the message.
> > >
> > > - Mark
> > > _______________________________________________
> > > ANCP mailing list
> > > ANCP@ietf.org
> > > https://www.ietf.org/mailman/listinfo/ancp
> > >
> > _______________________________________________
> > ANCP mailing list
> > ANCP@ietf.org
> > https://www.ietf.org/mailman/listinfo/ancp
> >
>
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Fri Aug 22 09:45:30 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 4C9F53A69E6;
	Fri, 22 Aug 2008 09:45:30 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id EB4273A69E6
	for <ancp@core3.amsl.com>; Fri, 22 Aug 2008 09:45:28 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.599
X-Spam-Level: 
X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5
	tests=[BAYES_00=-2.599]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id q+zmH2U7Kq76 for <ancp@core3.amsl.com>;
	Fri, 22 Aug 2008 09:45:27 -0700 (PDT)
Received: from prattle.redback.com (prattle.redback.com [155.53.12.9])
	by core3.amsl.com (Postfix) with ESMTP id E1AD83A68B5
	for <ancp@ietf.org>; Fri, 22 Aug 2008 09:45:26 -0700 (PDT)
Received: from localhost (localhost [127.0.0.1])
	by prattle.redback.com (Postfix) with ESMTP id A3B23D20470;
	Fri, 22 Aug 2008 09:25:19 -0700 (PDT)
Received: from prattle.redback.com ([127.0.0.1])
	by localhost (prattle [127.0.0.1]) (amavisd-new, port 10024) with ESMTP
	id 23925-01; Fri, 22 Aug 2008 09:25:19 -0700 (PDT)
Received: from rbsjx.ad.redback.com (rbsjxh1.redback.com [155.53.14.105])
	by prattle.redback.com (Postfix) with ESMTP id 4ED21D2046F;
	Fri, 22 Aug 2008 09:25:19 -0700 (PDT)
Received: from RBSJX.ad.redback.com ([155.53.14.103]) by
	rbsjxh1.ad.redback.com ([155.53.14.105]) with mapi;
	Fri, 22 Aug 2008 09:25:19 -0700
From: Peter Arberg <parberg@redback.com>
To: Maglione Roberta <roberta.maglione@telecomitalia.it>,
	"ancp@ietf.org" <ancp@ietf.org>
Date: Fri, 22 Aug 2008 09:25:17 -0700
Thread-Topic: Re: [ANCP] On TLVs
Thread-Index: AcjyRs3clz+hA956R722OxvE/NRFgwAH4hjwACkDViAAJ6D2cAQvJmcAAANUniA=
Message-ID: <48C276B536524E478C659685995F6AA57C5E9272B6@RBSJX.ad.redback.com>
References: <48906A22.7070409@cisco.com><D9872168DBD43A41BD71FFC4713274D40579A4D8@xmb-ams-33b.emea.cisco.com><48C276B536524E478C659685995F6AA57C597A4E73@RBSJX.ad.redback.com>
	<D9872168DBD43A41BD71FFC4713274D40579AA4B@xmb-ams-33b.emea.cisco.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F26@GRFMBX702RM001.griffon.local>
In-Reply-To: <A1F769BC58A8B146B2EEA818EAE052A20129491F26@GRFMBX702RM001.griffon.local>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-Virus-Scanned: by amavisd-new at redback.com
Subject: Re: [ANCP] On TLVs
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Maglione,

I have nothing against a Target Type TLV, my comments is much more
towards where and when you use it.

I will still like to see a description on the TLV, what the idea is
by introducing it, generic usage of it and the rules you envision
it to play by.

One of my comments on the TLV was that if you really think this is a
generic TLV which can be used in many places, then maybe it is not
the multicast draft which is the right place to define it, maybe it
needs to be defined in the protocol draft, and then simply used in the
new messages where it makes sense.

So I will say, please update the existing drafts with the Target Type
TLV explanation, then it is much easier to make comments if people
are satisfied with the TLV and its usage or not.

cheers,
Peter


> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On
> Behalf Of Maglione Roberta
> Sent: 22. august 2008 07:53
> To: ancp@ietf.org
> Subject: Re: [ANCP] On TLVs
>
>
> Hello,
>      as Design Team is going to start working on other
> Multicast messages that may re-use Target Type TLV  we'd like
> to know if anybody has additional opinions/comments or would
> like have more clarifications on the proposed new TLV.
>
> Thanks and best regards.
> Roberta on behalf of the Design Team.
>
>
>
> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On
> Behalf Of Wojciech Dec (wdec)
> Sent: Friday, August 01, 2008 9:42 AM
> To: Peter Arberg; Mark Townsley (townsley); ancp@ietf.org
> Subject: Re: [ANCP] On TLVs
>
> Hi,
>
> > -----Original Message-----
> > From: Peter Arberg [mailto:parberg@redback.com]
> > Sent: 31 July 2008 15:03
> > To: Wojciech Dec (wdec); Mark Townsley (townsley); ancp@ietf.org
> > Subject: RE: Re: [ANCP] On TLVs
> >
> > Hi,
> >
> > my comments in the meeting was exactly what you write Mark.
> > Woj and others have after the meeting explained the reasoning
> > for the TLV, as they have expanded the transactional
> > multicast case from being a 1 target (single dsl line) to
> > potentially in the future be multiple targets enabled
> simultaneously.
> >
> > This can then with the TLV be done in 2 ways, either by
> > including sub-tlvs for additional ACI identifiers, or add a
> > new identifier which could be learned or configured by some
> > today non identified mechanism.
> >
> > As such the TLV is in my mind okay if people in general buy
> > into the usability of having multiple simultaneous targets
> > for transactional multicast.
>
> The re-use of such TLV is not just intended for transactional
> multicast,
> but pretty much any (future) message type. Furthermore, it's
> not just a
> case of multiple targets, but totally different target types,
> eg an ACL
> is very much a target, and it's identifier (a string say) is not much
> different from that of an ACI. It is very useful to have this traget
> type concept in the protocol in order to have it extensible.
>
> >
> > In my mind I do not see a need for the capability of a NAS
> > telling the AN to update the multicast forwarding state for
> > many ports simultaneously, as individual clients in the
> > multicast use-case discussed so far have to send individual
> > joins, and as such it can not be a goal to be able to add
> > many when they have to join individually.
>
> The use-case for addressing multiple ports through a command
> (report) is
> already in the framework/requirements doc.
>
> >
> > So as I see it at the moment, there seems to be an argument
> > that a generic target TLV can be useful in other/many ANCP
> > messages, but it is defined in a transactional multicast
> > draft, without any explanation why it is suggested, and what
> > it will solve.
> >
> > Also it is not clear to me what the authors have in mind for
> > a use-case of multi targets in transactional multicast draft,
> > which are trying to solve the framework requirements of the
> > multicast use-case which is mainly focused around IPTV.
> >
> > So it will be nice to either have a better explanation why
> > multiple targets are useful, if this is aimed at having the
> > transactional multicast solution handle a future use-case
> > where many clients can be added to a multicast stream which
> > the NAS learns out-of-band what clients should be added, or
> > if the intention is to add a capability for the NAS to delay
> > multicast joins so they can be batched in the ancp message.
> >
> > In my mind the usage of this target TLV is just adding
> > complexity to the transactional multicast communication, and
> > it is in my mind not adding any significant value.
>
> Granted, a better explanation is due and will be put in place. As
> mentioned earlier, the complexity is trivial; it's just a TLV
> # wrapper,
> ie an extra few bytes, nothing more.
>
> Thanks,
> Woj.
>
> >
> > Peter
> >
> > > -----Original Message-----
> > > From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org]
> > On Behalf
> > > Of Wojciech Dec (wdec)
> > > Sent: 30. juli 2008 18:32
> > > To: Mark Townsley (townsley); ancp@ietf.org
> > > Subject: Re: [ANCP] On TLVs
> > >
> > > Hi Mark,
> > >
> > > Let me go over what the rationale for this TLV was. You and others
> > > feel free to comment/re-asses/advise.
> > >
> > > 1. The target type wrapper TLV
> > > There is a very reasonable need to have a protocol wide (ie
> > re-usable
> > > across messages) container TLV that carries in its payload
> > information
> > > allowing the recipient to identify the target of a command.
> > > There may be
> > > different target type identifiers, besides different types
> > of target
> > > (these would need to be definied in the future).
> > > The information that allows an access node to identify the traget
> > > (access port) is currently the ACI, but others may get defined,
> > > inluding things like group of ports say, or an access-list. If the
> > > access port, and/or the ACI were the only type of object and
> > > identifier respectively that we envisage commands to act
> > upon, then we
> > > can do away with the target type wrapper, gaining a few bytes of
> > > efficency, at the cost of extensibility. Given though other
> > types of
> > > targets and identifiers are a definite possility, the wrapper TLV
> > > would appear to be a small cost/complication to have.
> > >
> > > 2. Fixed position of target wrapper TLV It was thought that
> > having the
> > > target be the 1st information TLV in a multicast request
> > message (the
> > > same TLV could however be anywhere else in other messages)
> > would allow
> > > the recipient to be more efficent in determining whether
> > the target is
> > > valid. Processing any commands before identifying the target makes
> > > little sense, and since a single message may contain N
> commands, it
> > > appearead resonable to have the target be before the
> commands. That
> > > said, nothing is set in stone, so if the WG is more
> > comfortable with
> > > allowing this TLV to be anywhere in the message, that's a
> > change that
> > > can be easily done.
> > >
> > > Regards,
> > > Woj.
> > >
> > >
> > > > -----Original Message-----
> > > > From: ancp-bounces@ietf.org
> > [mailto:ancp-bounces@ietf.org] On Behalf
> > > > Of Mark Townsley (townsley)
> > > > Sent: 30 July 2008 15:18
> > > > To: ancp@ietf.org
> > > > Subject: [ANCP] On TLVs
> > > >
> > > >
> > > > Taking this to the list...  There was a discussion on
> > whether a TLV
> > > > was the right construct to use for something.
> > > >
> > > > First, TLVs are great, they allow extensibility, are
> > proven protocol
> > > > mechanics, etc. etc.
> > > >
> > > > But. If you restrict the TLV to be in a fixed location,
> always be
> > > > present, and never be repeated, it effectively neuters
> all of the
> > > > good reasons to have a TLV.
> > > >
> > > > For example:
> > > >
> > > > Type = X
> > > > Length = Y
> > > > Value = Information
> > > >
> > > > If this is always present, always present at the same
> > location, and
> > > > never repeated in a list of TLVs, you are effectively
> > putting an "X"
> > > > and "Y" in your protocol message which will always be
> present and
> > > > adds no value at all. It's always there. If it is
> *always* there,
> > > > and *always* the same, what's the point of it even
> > existing? Just to
> > > > see if one implementation can put it in the right place and the
> > > > other can check to see if it is there?
> > > >
> > > > In actuality, you could just put "Information" the
> > meesage as part
> > > > of the fixed header that precedes it.
> > > >
> > > > It's really not a big deal at the end of the day, one can
> > skip over
> > > > "X and Y" when processing. But, it's one more thing to
> > get wrong in
> > > > terms of interoperability. i.e., make something *look*
> > like a TLV,
> > > > and some coder might expect it to *be* a TLV when it is not by
> > > > mandate from the RFC.
> > > >
> > > > I speak from experience here. L2TP has its Message Type
> > encoded as a
> > > > TLV, with the same kinds of rules I heard proposed in
> the meeting
> > > > today.
> > > > It was a mistake, and we witnessed actual interop
> > problems from it
> > > > in the beginning. It either should have been a proper TLV placed
> > > > anywhere, or a part of the fixed header at the start of
> > the message.
> > > >
> > > > - Mark
> > > > _______________________________________________
> > > > ANCP mailing list
> > > > ANCP@ietf.org
> > > > https://www.ietf.org/mailman/listinfo/ancp
> > > >
> > > _______________________________________________
> > > ANCP mailing list
> > > ANCP@ietf.org
> > > https://www.ietf.org/mailman/listinfo/ancp
> > >
> >
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
> --------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE
>
> This message and its attachments are addressed solely to the
> persons above and may contain confidential information. If
> you have received the message in error, be informed that any
> use of the content hereof is prohibited. Please return it
> immediately to the sender and delete the message. Should you
> have any questions, please contact us by replying to
> webmaster@telecomitalia.it.
>
>         Thank you
>
>                                         www.telecomitalia.it
>
> --------------------------------------------------------------------
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
>
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Mon Aug 25 02:26:55 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id A6F8B3A69A0;
	Mon, 25 Aug 2008 02:26:55 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id EFC473A69C1
	for <ancp@core3.amsl.com>; Mon, 25 Aug 2008 02:26:54 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 1.881
X-Spam-Level: *
X-Spam-Status: No, score=1.881 tagged_above=-999 required=5
	tests=[BAYES_50=0.001, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id g3dmjoYYD-Yi for <ancp@core3.amsl.com>;
	Mon, 25 Aug 2008 02:26:53 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id 8C9BF3A6969
	for <ancp@ietf.org>; Mon, 25 Aug 2008 02:26:52 -0700 (PDT)
Received: from ptpxch007rm001.idc.cww.telecomitalia.it ([156.54.205.169]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Mon, 25 Aug 2008 11:14:26 +0200
Received: from grfhub701rm001.griffon.local ([10.19.3.8]) by
	ptpxch007rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Mon, 25 Aug 2008 11:14:25 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub701rm001.griffon.local ([10.19.3.8]) with mapi;
	Mon, 25 Aug 2008 11:14:25 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: "'Peter Arberg'" <parberg@redback.com>,
	<ancp@ietf.org>
Date: Mon, 25 Aug 2008 11:14:24 +0200
Thread-Topic: Re: [ANCP] On TLVs
thread-index: AcjyRs3clz+hA956R722OxvE/NRFgwAH4hjwACkDViAAJ6D2cAQvJmcAAANUniAAhoY3kA==
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491F29@GRFMBX702RM001.griffon.local>
References: <48906A22.7070409@cisco.com><D9872168DBD43A41BD71FFC4713274D40579A4D8@xmb-ams-33b.emea.cisco.com><48C276B536524E478C659685995F6AA57C597A4E73@RBSJX.ad.redback.com><D9872168DBD43A41BD71FFC4713274D40579AA4B@xmb-ams-33b.emea.cisco.com>
	<A1F769BC58A8B146B2EEA818EAE052A20129491F26@GRFMBX702RM001.griffon.local>
	<48C276B536524E478C659685995F6AA57C5E9272B6@RBSJX.ad.redback.com>
In-Reply-To: <48C276B536524E478C659685995F6AA57C5E9272B6@RBSJX.ad.redback.com>
Accept-Language: en-US, it-IT
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US, it-IT
MIME-Version: 1.0
X-OriginalArrivalTime: 25 Aug 2008 09:14:25.0759 (UTC)
	FILETIME=[F84D7EF0:01C90692]
Subject: Re: [ANCP] On TLVs
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org


Hi Peter,
    thanks for your feedback.

> I have nothing against a Target Type TLV, my comments is much more
> towards where and when you use it.

Great.

> I will still like to see a description on the TLV, what the idea is
> by introducing it, generic usage of it and the rules you envision
> it to play by.

Indeed, I agree with you we need to update the draft with a better description for this TLV and clearer explanation of its uses.

> One of my comments on the TLV was that if you really think this is a
> generic TLV which can be used in many places, then maybe it is not
> the multicast draft which is the right place to define

I understand your point, our initial intention was to incorporate the whole Multicast extensions in the ANCP Protocol I-D, thus we didn't clearly differentiated between Multicast specific items and more generic elements, like for example Target Type TLV.
Now, in case this draft may live as standalone, we may consider following your suggestion and moving some stuff like, Target Type TLV, in ANCP Protocol I-D.
We will discuss this point in the design team meeting and we will update the draft(s) accordingly.

In the meanwhile if nobody is against Target Type TLV, Design Team will consider re-using it across other messages.

Thanks and best regards,
Roberta


-----Original Message-----
From: Peter Arberg [mailto:parberg@redback.com]
Sent: Friday, August 22, 2008 6:25 PM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: Re: [ANCP] On TLVs

Hi Maglione,

I have nothing against a Target Type TLV, my comments is much more
towards where and when you use it.

I will still like to see a description on the TLV, what the idea is
by introducing it, generic usage of it and the rules you envision
it to play by.

One of my comments on the TLV was that if you really think this is a
generic TLV which can be used in many places, then maybe it is not
the multicast draft which is the right place to define it, maybe it
needs to be defined in the protocol draft, and then simply used in the
new messages where it makes sense.

So I will say, please update the existing drafts with the Target Type
TLV explanation, then it is much easier to make comments if people
are satisfied with the TLV and its usage or not.

cheers,
Peter


> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On
> Behalf Of Maglione Roberta
> Sent: 22. august 2008 07:53
> To: ancp@ietf.org
> Subject: Re: [ANCP] On TLVs
>
>
> Hello,
>      as Design Team is going to start working on other
> Multicast messages that may re-use Target Type TLV  we'd like
> to know if anybody has additional opinions/comments or would
> like have more clarifications on the proposed new TLV.
>
> Thanks and best regards.
> Roberta on behalf of the Design Team.
>
>
>
> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On
> Behalf Of Wojciech Dec (wdec)
> Sent: Friday, August 01, 2008 9:42 AM
> To: Peter Arberg; Mark Townsley (townsley); ancp@ietf.org
> Subject: Re: [ANCP] On TLVs
>
> Hi,
>
> > -----Original Message-----
> > From: Peter Arberg [mailto:parberg@redback.com]
> > Sent: 31 July 2008 15:03
> > To: Wojciech Dec (wdec); Mark Townsley (townsley); ancp@ietf.org
> > Subject: RE: Re: [ANCP] On TLVs
> >
> > Hi,
> >
> > my comments in the meeting was exactly what you write Mark.
> > Woj and others have after the meeting explained the reasoning
> > for the TLV, as they have expanded the transactional
> > multicast case from being a 1 target (single dsl line) to
> > potentially in the future be multiple targets enabled
> simultaneously.
> >
> > This can then with the TLV be done in 2 ways, either by
> > including sub-tlvs for additional ACI identifiers, or add a
> > new identifier which could be learned or configured by some
> > today non identified mechanism.
> >
> > As such the TLV is in my mind okay if people in general buy
> > into the usability of having multiple simultaneous targets
> > for transactional multicast.
>
> The re-use of such TLV is not just intended for transactional
> multicast,
> but pretty much any (future) message type. Furthermore, it's
> not just a
> case of multiple targets, but totally different target types,
> eg an ACL
> is very much a target, and it's identifier (a string say) is not much
> different from that of an ACI. It is very useful to have this traget
> type concept in the protocol in order to have it extensible.
>
> >
> > In my mind I do not see a need for the capability of a NAS
> > telling the AN to update the multicast forwarding state for
> > many ports simultaneously, as individual clients in the
> > multicast use-case discussed so far have to send individual
> > joins, and as such it can not be a goal to be able to add
> > many when they have to join individually.
>
> The use-case for addressing multiple ports through a command
> (report) is
> already in the framework/requirements doc.
>
> >
> > So as I see it at the moment, there seems to be an argument
> > that a generic target TLV can be useful in other/many ANCP
> > messages, but it is defined in a transactional multicast
> > draft, without any explanation why it is suggested, and what
> > it will solve.
> >
> > Also it is not clear to me what the authors have in mind for
> > a use-case of multi targets in transactional multicast draft,
> > which are trying to solve the framework requirements of the
> > multicast use-case which is mainly focused around IPTV.
> >
> > So it will be nice to either have a better explanation why
> > multiple targets are useful, if this is aimed at having the
> > transactional multicast solution handle a future use-case
> > where many clients can be added to a multicast stream which
> > the NAS learns out-of-band what clients should be added, or
> > if the intention is to add a capability for the NAS to delay
> > multicast joins so they can be batched in the ancp message.
> >
> > In my mind the usage of this target TLV is just adding
> > complexity to the transactional multicast communication, and
> > it is in my mind not adding any significant value.
>
> Granted, a better explanation is due and will be put in place. As
> mentioned earlier, the complexity is trivial; it's just a TLV
> # wrapper,
> ie an extra few bytes, nothing more.
>
> Thanks,
> Woj.
>
> >
> > Peter
> >
> > > -----Original Message-----
> > > From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org]
> > On Behalf
> > > Of Wojciech Dec (wdec)
> > > Sent: 30. juli 2008 18:32
> > > To: Mark Townsley (townsley); ancp@ietf.org
> > > Subject: Re: [ANCP] On TLVs
> > >
> > > Hi Mark,
> > >
> > > Let me go over what the rationale for this TLV was. You and others
> > > feel free to comment/re-asses/advise.
> > >
> > > 1. The target type wrapper TLV
> > > There is a very reasonable need to have a protocol wide (ie
> > re-usable
> > > across messages) container TLV that carries in its payload
> > information
> > > allowing the recipient to identify the target of a command.
> > > There may be
> > > different target type identifiers, besides different types
> > of target
> > > (these would need to be definied in the future).
> > > The information that allows an access node to identify the traget
> > > (access port) is currently the ACI, but others may get defined,
> > > inluding things like group of ports say, or an access-list. If the
> > > access port, and/or the ACI were the only type of object and
> > > identifier respectively that we envisage commands to act
> > upon, then we
> > > can do away with the target type wrapper, gaining a few bytes of
> > > efficency, at the cost of extensibility. Given though other
> > types of
> > > targets and identifiers are a definite possility, the wrapper TLV
> > > would appear to be a small cost/complication to have.
> > >
> > > 2. Fixed position of target wrapper TLV It was thought that
> > having the
> > > target be the 1st information TLV in a multicast request
> > message (the
> > > same TLV could however be anywhere else in other messages)
> > would allow
> > > the recipient to be more efficent in determining whether
> > the target is
> > > valid. Processing any commands before identifying the target makes
> > > little sense, and since a single message may contain N
> commands, it
> > > appearead resonable to have the target be before the
> commands. That
> > > said, nothing is set in stone, so if the WG is more
> > comfortable with
> > > allowing this TLV to be anywhere in the message, that's a
> > change that
> > > can be easily done.
> > >
> > > Regards,
> > > Woj.
> > >
> > >
> > > > -----Original Message-----
> > > > From: ancp-bounces@ietf.org
> > [mailto:ancp-bounces@ietf.org] On Behalf
> > > > Of Mark Townsley (townsley)
> > > > Sent: 30 July 2008 15:18
> > > > To: ancp@ietf.org
> > > > Subject: [ANCP] On TLVs
> > > >
> > > >
> > > > Taking this to the list...  There was a discussion on
> > whether a TLV
> > > > was the right construct to use for something.
> > > >
> > > > First, TLVs are great, they allow extensibility, are
> > proven protocol
> > > > mechanics, etc. etc.
> > > >
> > > > But. If you restrict the TLV to be in a fixed location,
> always be
> > > > present, and never be repeated, it effectively neuters
> all of the
> > > > good reasons to have a TLV.
> > > >
> > > > For example:
> > > >
> > > > Type = X
> > > > Length = Y
> > > > Value = Information
> > > >
> > > > If this is always present, always present at the same
> > location, and
> > > > never repeated in a list of TLVs, you are effectively
> > putting an "X"
> > > > and "Y" in your protocol message which will always be
> present and
> > > > adds no value at all. It's always there. If it is
> *always* there,
> > > > and *always* the same, what's the point of it even
> > existing? Just to
> > > > see if one implementation can put it in the right place and the
> > > > other can check to see if it is there?
> > > >
> > > > In actuality, you could just put "Information" the
> > meesage as part
> > > > of the fixed header that precedes it.
> > > >
> > > > It's really not a big deal at the end of the day, one can
> > skip over
> > > > "X and Y" when processing. But, it's one more thing to
> > get wrong in
> > > > terms of interoperability. i.e., make something *look*
> > like a TLV,
> > > > and some coder might expect it to *be* a TLV when it is not by
> > > > mandate from the RFC.
> > > >
> > > > I speak from experience here. L2TP has its Message Type
> > encoded as a
> > > > TLV, with the same kinds of rules I heard proposed in
> the meeting
> > > > today.
> > > > It was a mistake, and we witnessed actual interop
> > problems from it
> > > > in the beginning. It either should have been a proper TLV placed
> > > > anywhere, or a part of the fixed header at the start of
> > the message.
> > > >
> > > > - Mark
> > > > _______________________________________________
> > > > ANCP mailing list
> > > > ANCP@ietf.org
> > > > https://www.ietf.org/mailman/listinfo/ancp
> > > >
> > > _______________________________________________
> > > ANCP mailing list
> > > ANCP@ietf.org
> > > https://www.ietf.org/mailman/listinfo/ancp
> > >
> >
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
> --------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE
>
> This message and its attachments are addressed solely to the
> persons above and may contain confidential information. If
> you have received the message in error, be informed that any
> use of the content hereof is prohibited. Please return it
> immediately to the sender and delete the message. Should you
> have any questions, please contact us by replying to
> webmaster@telecomitalia.it.
>
>         Thank you
>
>                                         www.telecomitalia.it
>
> --------------------------------------------------------------------
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
>
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


From ancp-bounces@ietf.org  Mon Aug 25 02:26:56 2008
Return-Path: <ancp-bounces@ietf.org>
X-Original-To: ancp-archive@optimus.ietf.org
Delivered-To: ietfarch-ancp-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id C44FF3A69D8;
	Mon, 25 Aug 2008 02:26:56 -0700 (PDT)
X-Original-To: ancp@core3.amsl.com
Delivered-To: ancp@core3.amsl.com
Received: from localhost (localhost [127.0.0.1])
	by core3.amsl.com (Postfix) with ESMTP id 8B9AC3A6969
	for <ancp@core3.amsl.com>; Mon, 25 Aug 2008 02:26:55 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 1.881
X-Spam-Level: *
X-Spam-Status: No, score=1.881 tagged_above=-999 required=5
	tests=[BAYES_50=0.001, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245]
Received: from mail.ietf.org ([64.170.98.32])
	by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024)
	with ESMTP id PrLF6ex6c73k for <ancp@core3.amsl.com>;
	Mon, 25 Aug 2008 02:26:54 -0700 (PDT)
Received: from maill.telecomitalia.it (maill.telecomitalia.it [217.169.121.52])
	by core3.amsl.com (Postfix) with ESMTP id D33CF3A69A0
	for <ancp@ietf.org>; Mon, 25 Aug 2008 02:26:53 -0700 (PDT)
Received: from ptpxch008rm001.idc.cww.telecomitalia.it ([156.54.205.170]) by
	maill.telecomitalia.it with Microsoft SMTPSVC(6.0.3790.1830);
	Mon, 25 Aug 2008 11:25:12 +0200
Received: from grfhub702rm001.griffon.local ([10.19.3.9]) by
	ptpxch008rm001.idc.cww.telecomitalia.it with Microsoft
	SMTPSVC(6.0.3790.1830); Mon, 25 Aug 2008 11:25:12 +0200
Content-Class: urn:content-classes:message
Importance: normal
Priority: normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.3790.2992
Received: from GRFMBX702RM001.griffon.local ([10.19.3.19]) by
	grfhub702rm001.griffon.local ([10.19.3.9]) with mapi;
	Mon, 25 Aug 2008 11:25:11 +0200
From: "Maglione Roberta" <roberta.maglione@telecomitalia.it>
To: <ancp@ietf.org>,
	"'Peter Arberg'" <parberg@redback.com>
Date: Mon, 25 Aug 2008 11:25:10 +0200
Thread-Topic: Re: [ANCP] On TLVs
thread-index: AcjyRs3clz+hA956R722OxvE/NRFgwAH4hjwACkDViAAJ6D2cAQvJmcAAANUniAAhoY3kAAB2mFQ
Message-ID: <A1F769BC58A8B146B2EEA818EAE052A20129491F2A@GRFMBX702RM001.griffon.local>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
acceptlanguage: en-US
MIME-Version: 1.0
X-OriginalArrivalTime: 25 Aug 2008 09:25:12.0095 (UTC)
	FILETIME=[798C8AF0:01C90694]
Subject: [ANCP] FW: Re:  On TLVs
X-BeenThere: ancp@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Access Node Control Protocol working group mailing list
	<ancp.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/pipermail/ancp>
List-Post: <mailto:ancp@ietf.org>
List-Help: <mailto:ancp-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ancp>,
	<mailto:ancp-request@ietf.org?subject=subscribe>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: ancp-bounces@ietf.org
Errors-To: ancp-bounces@ietf.org

Hi Peter,
    thanks for your feedback.

> I have nothing against a Target Type TLV, my comments is much more
> towards where and when you use it.

Great.

> I will still like to see a description on the TLV, what the idea is
> by introducing it, generic usage of it and the rules you envision
> it to play by.

Indeed, I agree with you we need to update the draft with a better description for this TLV and clearer explanation of its uses.

> One of my comments on the TLV was that if you really think this is a
> generic TLV which can be used in many places, then maybe it is not
> the multicast draft which is the right place to define

I understand your point, our initial intention was to incorporate the whole Multicast extensions in the ANCP Protocol I-D, thus we didn't clearly differentiate between Multicast specific items and more generic elements, like for example Target Type TLV.
Now, in case this draft may live as standalone, we may consider following your suggestion and moving some stuff like, Target Type TLV, in ANCP Protocol I-D.
We will discuss this point in the design team meeting and we will update the draft(s) accordingly.

In the meanwhile if nobody is against Target Type TLV, Design Team will consider re-using it across other messages.

Thanks and best regards,
Roberta


-----Original Message-----
From: Peter Arberg [mailto:parberg@redback.com]
Sent: Friday, August 22, 2008 6:25 PM
To: Maglione Roberta; ancp@ietf.org
Subject: RE: Re: [ANCP] On TLVs

Hi Maglione,

I have nothing against a Target Type TLV, my comments is much more
towards where and when you use it.

I will still like to see a description on the TLV, what the idea is
by introducing it, generic usage of it and the rules you envision
it to play by.

One of my comments on the TLV was that if you really think this is a
generic TLV which can be used in many places, then maybe it is not
the multicast draft which is the right place to define it, maybe it
needs to be defined in the protocol draft, and then simply used in the
new messages where it makes sense.

So I will say, please update the existing drafts with the Target Type
TLV explanation, then it is much easier to make comments if people
are satisfied with the TLV and its usage or not.

cheers,
Peter


> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On
> Behalf Of Maglione Roberta
> Sent: 22. august 2008 07:53
> To: ancp@ietf.org
> Subject: Re: [ANCP] On TLVs
>
>
> Hello,
>      as Design Team is going to start working on other
> Multicast messages that may re-use Target Type TLV  we'd like
> to know if anybody has additional opinions/comments or would
> like have more clarifications on the proposed new TLV.
>
> Thanks and best regards.
> Roberta on behalf of the Design Team.
>
>
>
> -----Original Message-----
> From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org] On
> Behalf Of Wojciech Dec (wdec)
> Sent: Friday, August 01, 2008 9:42 AM
> To: Peter Arberg; Mark Townsley (townsley); ancp@ietf.org
> Subject: Re: [ANCP] On TLVs
>
> Hi,
>
> > -----Original Message-----
> > From: Peter Arberg [mailto:parberg@redback.com]
> > Sent: 31 July 2008 15:03
> > To: Wojciech Dec (wdec); Mark Townsley (townsley); ancp@ietf.org
> > Subject: RE: Re: [ANCP] On TLVs
> >
> > Hi,
> >
> > my comments in the meeting was exactly what you write Mark.
> > Woj and others have after the meeting explained the reasoning
> > for the TLV, as they have expanded the transactional
> > multicast case from being a 1 target (single dsl line) to
> > potentially in the future be multiple targets enabled
> simultaneously.
> >
> > This can then with the TLV be done in 2 ways, either by
> > including sub-tlvs for additional ACI identifiers, or add a
> > new identifier which could be learned or configured by some
> > today non identified mechanism.
> >
> > As such the TLV is in my mind okay if people in general buy
> > into the usability of having multiple simultaneous targets
> > for transactional multicast.
>
> The re-use of such TLV is not just intended for transactional
> multicast,
> but pretty much any (future) message type. Furthermore, it's
> not just a
> case of multiple targets, but totally different target types,
> eg an ACL
> is very much a target, and it's identifier (a string say) is not much
> different from that of an ACI. It is very useful to have this traget
> type concept in the protocol in order to have it extensible.
>
> >
> > In my mind I do not see a need for the capability of a NAS
> > telling the AN to update the multicast forwarding state for
> > many ports simultaneously, as individual clients in the
> > multicast use-case discussed so far have to send individual
> > joins, and as such it can not be a goal to be able to add
> > many when they have to join individually.
>
> The use-case for addressing multiple ports through a command
> (report) is
> already in the framework/requirements doc.
>
> >
> > So as I see it at the moment, there seems to be an argument
> > that a generic target TLV can be useful in other/many ANCP
> > messages, but it is defined in a transactional multicast
> > draft, without any explanation why it is suggested, and what
> > it will solve.
> >
> > Also it is not clear to me what the authors have in mind for
> > a use-case of multi targets in transactional multicast draft,
> > which are trying to solve the framework requirements of the
> > multicast use-case which is mainly focused around IPTV.
> >
> > So it will be nice to either have a better explanation why
> > multiple targets are useful, if this is aimed at having the
> > transactional multicast solution handle a future use-case
> > where many clients can be added to a multicast stream which
> > the NAS learns out-of-band what clients should be added, or
> > if the intention is to add a capability for the NAS to delay
> > multicast joins so they can be batched in the ancp message.
> >
> > In my mind the usage of this target TLV is just adding
> > complexity to the transactional multicast communication, and
> > it is in my mind not adding any significant value.
>
> Granted, a better explanation is due and will be put in place. As
> mentioned earlier, the complexity is trivial; it's just a TLV
> # wrapper,
> ie an extra few bytes, nothing more.
>
> Thanks,
> Woj.
>
> >
> > Peter
> >
> > > -----Original Message-----
> > > From: ancp-bounces@ietf.org [mailto:ancp-bounces@ietf.org]
> > On Behalf
> > > Of Wojciech Dec (wdec)
> > > Sent: 30. juli 2008 18:32
> > > To: Mark Townsley (townsley); ancp@ietf.org
> > > Subject: Re: [ANCP] On TLVs
> > >
> > > Hi Mark,
> > >
> > > Let me go over what the rationale for this TLV was. You and others
> > > feel free to comment/re-asses/advise.
> > >
> > > 1. The target type wrapper TLV
> > > There is a very reasonable need to have a protocol wide (ie
> > re-usable
> > > across messages) container TLV that carries in its payload
> > information
> > > allowing the recipient to identify the target of a command.
> > > There may be
> > > different target type identifiers, besides different types
> > of target
> > > (these would need to be definied in the future).
> > > The information that allows an access node to identify the traget
> > > (access port) is currently the ACI, but others may get defined,
> > > inluding things like group of ports say, or an access-list. If the
> > > access port, and/or the ACI were the only type of object and
> > > identifier respectively that we envisage commands to act
> > upon, then we
> > > can do away with the target type wrapper, gaining a few bytes of
> > > efficency, at the cost of extensibility. Given though other
> > types of
> > > targets and identifiers are a definite possility, the wrapper TLV
> > > would appear to be a small cost/complication to have.
> > >
> > > 2. Fixed position of target wrapper TLV It was thought that
> > having the
> > > target be the 1st information TLV in a multicast request
> > message (the
> > > same TLV could however be anywhere else in other messages)
> > would allow
> > > the recipient to be more efficent in determining whether
> > the target is
> > > valid. Processing any commands before identifying the target makes
> > > little sense, and since a single message may contain N
> commands, it
> > > appearead resonable to have the target be before the
> commands. That
> > > said, nothing is set in stone, so if the WG is more
> > comfortable with
> > > allowing this TLV to be anywhere in the message, that's a
> > change that
> > > can be easily done.
> > >
> > > Regards,
> > > Woj.
> > >
> > >
> > > > -----Original Message-----
> > > > From: ancp-bounces@ietf.org
> > [mailto:ancp-bounces@ietf.org] On Behalf
> > > > Of Mark Townsley (townsley)
> > > > Sent: 30 July 2008 15:18
> > > > To: ancp@ietf.org
> > > > Subject: [ANCP] On TLVs
> > > >
> > > >
> > > > Taking this to the list...  There was a discussion on
> > whether a TLV
> > > > was the right construct to use for something.
> > > >
> > > > First, TLVs are great, they allow extensibility, are
> > proven protocol
> > > > mechanics, etc. etc.
> > > >
> > > > But. If you restrict the TLV to be in a fixed location,
> always be
> > > > present, and never be repeated, it effectively neuters
> all of the
> > > > good reasons to have a TLV.
> > > >
> > > > For example:
> > > >
> > > > Type = X
> > > > Length = Y
> > > > Value = Information
> > > >
> > > > If this is always present, always present at the same
> > location, and
> > > > never repeated in a list of TLVs, you are effectively
> > putting an "X"
> > > > and "Y" in your protocol message which will always be
> present and
> > > > adds no value at all. It's always there. If it is
> *always* there,
> > > > and *always* the same, what's the point of it even
> > existing? Just to
> > > > see if one implementation can put it in the right place and the
> > > > other can check to see if it is there?
> > > >
> > > > In actuality, you could just put "Information" the
> > meesage as part
> > > > of the fixed header that precedes it.
> > > >
> > > > It's really not a big deal at the end of the day, one can
> > skip over
> > > > "X and Y" when processing. But, it's one more thing to
> > get wrong in
> > > > terms of interoperability. i.e., make something *look*
> > like a TLV,
> > > > and some coder might expect it to *be* a TLV when it is not by
> > > > mandate from the RFC.
> > > >
> > > > I speak from experience here. L2TP has its Message Type
> > encoded as a
> > > > TLV, with the same kinds of rules I heard proposed in
> the meeting
> > > > today.
> > > > It was a mistake, and we witnessed actual interop
> > problems from it
> > > > in the beginning. It either should have been a proper TLV placed
> > > > anywhere, or a part of the fixed header at the start of
> > the message.
> > > >
> > > > - Mark
> > > > _______________________________________________
> > > > ANCP mailing list
> > > > ANCP@ietf.org
> > > > https://www.ietf.org/mailman/listinfo/ancp
> > > >
> > > _______________________________________________
> > > ANCP mailing list
> > > ANCP@ietf.org
> > > https://www.ietf.org/mailman/listinfo/ancp
> > >
> >
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
> --------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE
>
> This message and its attachments are addressed solely to the
> persons above and may contain confidential information. If
> you have received the message in error, be informed that any
> use of the content hereof is prohibited. Please return it
> immediately to the sender and delete the message. Should you
> have any questions, please contact us by replying to
> webmaster@telecomitalia.it.
>
>         Thank you
>
>                                         www.telecomitalia.it
>
> --------------------------------------------------------------------
>
> _______________________________________________
> ANCP mailing list
> ANCP@ietf.org
> https://www.ietf.org/mailman/listinfo/ancp
>
--------------------------------------------------------------------

CONFIDENTIALITY NOTICE

This message and its attachments are addressed solely to the persons above and may contain confidential information. If you have received the message in error, be informed that any use of the content hereof is prohibited. Please return it immediately to the sender and delete the message. Should you have any questions, please contact us by replying to webmaster@telecomitalia.it.

        Thank you

                                        www.telecomitalia.it

--------------------------------------------------------------------
                        
_______________________________________________
ANCP mailing list
ANCP@ietf.org
https://www.ietf.org/mailman/listinfo/ancp


