From nsis-bounces@ietf.org Wed Nov 01 07:38:30 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfFLr-0005ku-78; Wed, 01 Nov 2006 07:37:51 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfFLq-0005kj-A1
	for nsis@ietf.org; Wed, 01 Nov 2006 07:37:50 -0500
Received: from mgw-ext13.nokia.com ([131.228.20.172])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfFLl-000310-OW
	for nsis@ietf.org; Wed, 01 Nov 2006 07:37:50 -0500
Received: from esebh105.NOE.Nokia.com (esebh105.ntc.nokia.com [172.21.138.211])
	by mgw-ext13.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA1CaLIQ019426; Wed, 1 Nov 2006 14:37:20 +0200
Received: from esebh101.NOE.Nokia.com ([172.21.138.177]) by
	esebh105.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 1 Nov 2006 14:37:00 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh101.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 1 Nov 2006 14:37:00 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] The version 3 of the InterDomain-QOSM Draft: the NSIS QOS
	Model to Fulfill the E2E QoS Control in the ITU-T RACF
	Functional Architecture
Date: Wed, 1 Nov 2006 14:36:59 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED0466@esebe199.NOE.Nokia.com>
In-Reply-To: <45476930.3020304@dcs.qmul.ac.uk>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] The version 3 of the InterDomain-QOSM Draft: the NSIS QOS
	Model to Fulfill the E2E QoS Control in the ITU-T RACF
	Functional Architecture
Thread-Index: Acb9AEdUxP3jpPnVTTq3dSvmzACrfAAshgRw
From: <john.loughney@nokia.com>
To: <jian.zhang@dcs.qmul.ac.uk>, <nsis@ietf.org>
X-OriginalArrivalTime: 01 Nov 2006 12:37:00.0330 (UTC)
	FILETIME=[6D1D98A0:01C6FDB2]
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 25620135586de10c627e3628c432b04a
Cc: edmundo@dei.uc.pt, karagian@cs.utwente.nl, mendes@docomolab-euro.com
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Jian,

Are any of the authors participating in the ITU-T RACF discussions?

thanks,
John=20

>-----Original Message-----
>From: ext Jian Zhang [mailto:jian.zhang@dcs.qmul.ac.uk]=20
>Sent: 31 October, 2006 17:18
>To: nsis@ietf.org
>Cc: edmundo@dei.uc.pt; karagian@cs.utwente.nl;=20
>mendes@docomolab-euro.com
>Subject: [NSIS] The version 3 of the InterDomain-QOSM Draft:=20
>the NSIS QOS Model to Fulfill the E2E QoS Control in the ITU-T=20
>RACF Functional Architecture
>
>Dear all,
>
>Please kindly read the version 3 of the InterDomain-QOSM=20
>Draft: our effort to define a NSIS QOS Model to fulfill the=20
>E2E QoS control in the ITU-T RACF functional architecture,=20
>which has been submitted to the NSIS WG for discussion at the=20
>coming IETF meeting.
>
>And if you have any questions or suggestions about this draft,=20
>we would like to answer or discuss with you in this mailing=20
>list. The abstract of this draft is attached below and the=20
>link which holds the whole contents of the draft is also=20
>provided below.
>
>Cheers,
>
>Jian Zhang
>
>Dept. of Computer Science
>Queen Mary, Univ. of London
>
>---------------------------------------
>
>Abstract
>
>   This document has three goals. First of all, it presents=20
>our analysis
>   of how to use the NSIS signaling (inter-domain QOSM and intra-domain
>   QOSM) to fulfill the QoS control in accord with the ITU-T RACF
>   functional architecture. For this goal, we discuss how the=20
>ITU-T RACF
>   entities in the ITU-T RACF functional architecture can be mapped to
>   the NSIS entities and how the RACF reference points can be
>   implemented by using the NSIS protocol suites and QOSMs.=20
>Secondly, we
>   aim at specifying an NSIS Inter-domain QOSM for E2E QoS control
>   across heterogeneous IP networks and applying this Inter-domain QOSM
>   to the e2e QoS control in the ITU-T RACF functional architecture
>   based on the above ITU-T RACF analysis. The detailed description of
>   the NSIS Inter-domain QOSM are given and the e2e QoS control
>   scenarios in the ITU-T RACF architecture (including RACF Push and
>   Pull resource control modes), which will be covered by the NSIS
>   Inter-domain QOSM are described and implemented. Thirdly, we specify
>   and implement those QSPECs that will be used by the=20
>Inter-domain QOSM
>   for the e2e QoS control in the ITU-T RACF architecture.
>
>The link is
>http://ietfreport.isoc.org/ids/draft-zhang-nsis-interdomain-qos
>m-03.txt .
>
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 01 10:27:38 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfHz4-00042x-B5; Wed, 01 Nov 2006 10:26:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfHxw-0003dW-CJ
	for nsis@ietf.org; Wed, 01 Nov 2006 10:25:20 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfHq9-0003xp-Ck
	for nsis@ietf.org; Wed, 01 Nov 2006 10:17:21 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Wed, 1 Nov 2006 10:16:48 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006110110164631372 ; Wed, 01 Nov 2006 10:16:48 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <VX80LPJA>; Wed, 1 Nov 2006 10:16:44 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EE2@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: john.loughney@nokia.com,
    jian.zhang@dcs.qmul.ac.uk,
    nsis@ietf.org
Subject: RE: [NSIS] The version 3 of the InterDomain-QOSM Draft: the NSIS 
	QOSModel to Fulfill the E2E QoS Control in the ITU-T RACFFunctional
	Archi tecture
Date: Wed, 1 Nov 2006 10:16:36 -0500 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 01 Nov 2006 15:16:37.0260 (UTC)
	FILETIME=[B96924C0:01C6FDC8]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cf3becbbd6d1a45acbe2ffd4ab88bdc2
Cc: edmundo@dei.uc.pt, karagian@cs.utwente.nl, mendes@docomolab-euro.com
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Jian and John:

It is nice that a meeting will take place like this. Hope that the
following questions will be discussed in the Architecture:

1. Each QOSM model will work individually with NSIS QOS Signaling:
DiffServ QOSM <-> NSIS, RSVP QOSM <-> NSIS, IntServ <-> NSIS, DSTE <->
NSIS, 3GPP/Y.1541 <-> NSIS, or Y.1541 <-> NSIS. 

However, the implementation-specific value-based DSTE, 3GPP/Y.1541, and
Y.1541 do not have any signaling messages. So, the only way they use
NSIS QOS signaling messages to have some specific values for some set of
QOS parameters that become some specific value-based
implementation-specific instances of NSIS QOS signaling messages. Why do
these three QOSM models want to be treated in the same level like those
of DiffServ, RSVP, and IntServ QOSM?

Even DSTE and Y.1541 will NEVER interwork because their "values" of all
QOS parameters are different. If so, why is NSIS WG addressing all of
those (DiffServ, RSVP, IntServ, Y.1541, DSTE, and 3GPP-Y.1541) in the
same QSPEC draft knowing the interworking problem? Will RACFF
Architecture be able to solve this Interworking problem? If so, how?

2. Interworking between DiffServ, RSVP, and IntServ QOSM using NSIS QOS
signaling protocols are there and there are no issues for mapping as
each one of them has its own non-implementation-specific non-value-based
signaling messages. So, the Interworking problems among these QOS
signaling protocols are NOT an issue.

If so, will the RACFF make a separate arch only for these QOS protocols?

3. Will the RACFF Arch qualify the implementation-specific value-based
Y.1541 QOS scheme that does not any signaling messages defining the
"generalized" properties of QOS characteristics of each performance
parameter in the same level of DIffServ, RSVP, and IntServ? If so, will
be able to explain how?

4. Will the RACFF arch be able to create the implementation-specific
value-based QOS Interworking architecture along the line of OSP protocol
where all implementers agree on some SLA agreements including QOS (e.g.,
specific values of delay, jitter, and packet loss for DiffServ code
points, already specified values for delay, jitter, and packet loss in
Y.1541)?

We would be interested to know the outcome of the meeting in getting the
answers of the above questions.

Best regards,
Radhika

-----Original Message-----
From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On Behalf Of
john.loughney@nokia.com
Sent: Wednesday, November 01, 2006 7:37 AM
To: jian.zhang@dcs.qmul.ac.uk; nsis@ietf.org
Cc: edmundo@dei.uc.pt; karagian@cs.utwente.nl; mendes@docomolab-euro.com
Subject: RE: [NSIS] The version 3 of the InterDomain-QOSM Draft: the
NSIS QOSModel to Fulfill the E2E QoS Control in the ITU-T RACFFunctional
Architecture

Jian,

Are any of the authors participating in the ITU-T RACF discussions?

thanks,
John 

>-----Original Message-----
>From: ext Jian Zhang [mailto:jian.zhang@dcs.qmul.ac.uk] 
>Sent: 31 October, 2006 17:18
>To: nsis@ietf.org
>Cc: edmundo@dei.uc.pt; karagian@cs.utwente.nl; 
>mendes@docomolab-euro.com
>Subject: [NSIS] The version 3 of the InterDomain-QOSM Draft: 
>the NSIS QOS Model to Fulfill the E2E QoS Control in the ITU-T 
>RACF Functional Architecture
>
>Dear all,
>
>Please kindly read the version 3 of the InterDomain-QOSM 
>Draft: our effort to define a NSIS QOS Model to fulfill the 
>E2E QoS control in the ITU-T RACF functional architecture, 
>which has been submitted to the NSIS WG for discussion at the 
>coming IETF meeting.
>
>And if you have any questions or suggestions about this draft, 
>we would like to answer or discuss with you in this mailing 
>list. The abstract of this draft is attached below and the 
>link which holds the whole contents of the draft is also 
>provided below.
>
>Cheers,
>
>Jian Zhang
>
>Dept. of Computer Science
>Queen Mary, Univ. of London
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 01 17:13:29 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfOKM-0000Kc-41; Wed, 01 Nov 2006 17:12:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfOKL-0000KP-33
	for nsis@ietf.org; Wed, 01 Nov 2006 17:12:53 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfOKJ-0006rJ-L3
	for nsis@ietf.org; Wed, 01 Nov 2006 17:12:53 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-12.tower-146.messagelabs.com!1162419170!180845!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 10442 invoked from network); 1 Nov 2006 22:12:50 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-12.tower-146.messagelabs.com with SMTP;
	1 Nov 2006 22:12:50 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA1MCoJU014329; 
	Wed, 1 Nov 2006 17:12:50 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA1MCjMD014306; Wed, 1 Nov 2006 17:12:46 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Wed, 1 Nov 2006 16:12:45 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <BAA65A575825454CBB0103267553FCCCED02E0@esebe199.NOE.Nokia.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Proposed Updates to the QSPEC Draft
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQ
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ded6070f7eed56e10c4f4d0d5043d9c7
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>,
	Cornelia Kappler <cornelia.kappler@gmx.net>,
	David R Oran <oran@cisco.com>, "Kappler,
	Cornelia" <cornelia.kappler@siemens.com>
Subject: [NSIS] Proposed Updates to the QSPEC Draft
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi All,

Per John's suggestion below we've discussed Dave Oran's suggestions at
length and have the following proposal. =20

We think the proposed changes address many concerns raised on the list
over the last few months and will require some rather basic changes in
the QSPEC document
http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.

We appreciate Dave's suggestions and careful review of the document.

Here is a summary of the proposed changes:

1. The notion of QOSMs will be played down in the updated draft.
Language to the effect that 'QNSLP/QSPEC can signal for different QOSMs
across multiple domains' will be replaced by the notion that QNSLP/QSPEC
allows different elements on the path to implement different data plane
QoS mechanisms that meets the constraints included in the QSPEC.  A QOSM
describes common capabilities among QNEs to act consistently when
requested to admit traffic and in treating admitted traffic.  A QOSM ID
need not be defined or signaled and a QNE need not support any
particular QOSM.  A 'QOSM specification' would become a 'QNSLP/QSPEC
applicability statement', which documents how a QNE/RMF interprets and
treats the various elements in a QSPEC. =20

An updated QOSM definition is as follows:

QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
constraints and traffic handling directives a QNE implementing that QOSM
is capable of supporting and how resources will be managed by the RMF in
such a QNE.

2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
semantics:
A) The source traffic description is mandatory to include by the QNI and
mandatory to interpret by downstream QNEs.  The traffic description is
specified by a traffic model (TMOD) parameter consisting of 4
sub-parameters:
- rate (r)
- bucket size (b)
- peak rate (p)
- minimum policed unit (m)
The TMOD parameter is a mathematically complete way to describe a
traffic source.  Examples will be given for setting TMOD to correspond
to specific cases, e.g. where only bandwidth is available (r=3Dp; b and =
m
set to default values) and other cases (e.g., TCP).  As such, the
bandwidth parameter is eliminated.
B) All other QSPEC parameters are optional to include by the QNI, and
may be either mandatory or optional to interpret by downstream QNEs.
That is, the QNI can explicitly specify for each constraint parameter
whether it wishes admission control to succeed or fail if that
constraint cannot be met.  A downstream QNE distinguishes whether it is
failure to understand the parameter or failure to meet the constraint
that causes the admission control failure (more detail on this below).
C) Additional QSPEC parameters can be defined in updates to the QSPEC
document (or alternatively specified in QOSM applicability statement
documents).  There is already an IANA registry established for
additional QSPEC parameters.

3. The concept of remapping QSPEC parameters is eliminated.  This
concept is ill defined and has caused much confusion (there are few if
any sensible remappings).  Rather we define 'interpret' a QSPEC
parameter to only mean what we now call 'strictly interpret' (i.e., must
conform to the RFCs defining the parameter and procedures) and eliminate
the remap option part of 'interpret'.

4. Since remapping is eliminated, the concept of stacking QSPECs is also
eliminated.  Discussion of tunneling of QNSLP messages is eliminated
since there are no impacts on tunneling required by QSPEC.

5. Current QSPEC flags are modified as follows:
A) The QNI sets the M flag for each QSPEC parameter it populates that it
intends must be interpreted or the reservation fails.  Currently the M
flag is statically set for every QSPEC parameter.
B) If the M flag is set, a downstream QNE must interpret the parameter
or the reservation fails.  If a QNE does not support the parameter with
the M flag set, it sets the N flag and rejects the reservation.  If the
QNE supports the parameter but cannot meet the parameter, it sets the E
flag and rejects the reservation.=20
C) If the M flag is not set, a downstream QNE should interpret the
parameter.  If a QNE does not support the parameter, it sets the N flag
and can optionally accept or reject the reservation.  If a QNE supports
the parameter but cannot meet the parameter, it sets the E flag and
optional accepts or rejects the reservation.
D) The R (remapped parameter) flag and Q (non QOSM) flag are eliminated.

6. Summary of proposed QSPEC parameters:

Source traffic Description:
 TMOD-1 (mandatory to include)
 TMOD-2 (optional to include)

Constraints (optional to include):
 Path Latency
 Path Jitter
 Path PLR
 Path PER
 Y.1541 class (shorthand notation for above 4 parameters)
 Slack term
 Priority (Preemption, Defending, Admission, RPH Priority)
 DSTE class

Handling directives (optional to include):
 excess treatment

Traffic Classifiers (optional to include):
 PHB Class (note that PHB class is set by downstream QNE to tell QNI how
to mark traffic to ensure it will be treated as committed by admission
control)

Eliminated:
 bandwidth
 Ctot, Dtot, Csum, Dsum

7. Editorial changes:
A) rewrite description of read-only and read-write parameters to make
the rules more clear:=20
- which parameters can be ro/rw (function of object in which parameter
included)
- can only increase rw values and not delete, etc.
B) eliminate the distinction of 'control' and 'qos description'
parameters and replace with suggested taxonomy, as described above
C) clarify discussion of packet outcomes from Y.1540 and place in
appropriate sections
D) clarify which parameters can be included in <QoS Desired> <QoS
Available>, <QoS Reserved>, or <QoS Minimum>

Please review and comment on the proposed changes.

Thanks,
Jerry, Dave, Cornelia


-----Original Message-----
From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]=20
Sent: Monday, October 23, 2006 2:59 AM
To: nsis@ietf.org
Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft

Hi all,

We have had a very large thread on the current
draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear consensus, so
I'd like to try to summarize the discussions.

Several people feel that the current draft contains too many mandatory
parameters and mapping some QoS Models to others is mixing
apples-to-oranges.  The feeling is that interoperability is hard to
achieve and therefore we should not try to gain interoperability.  Jukka
Manner's message on 17th of October proposed a way forward.
http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html

However, several (other) people feel that interoperability is important
and is needed. Giving up on interoperability would not be a good thing,
and that we are very close to a solution.  Dave Oran proposed a
mechanism to reach a solution on the 19th October:
http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html

I have asked Jerry and Dave to make a proposal along the lines of Dave's
message above, and the working group can review this, to see if we can
gain consensus on this.  I think it would be sufficient to send some
rough text to the mailing list, and if this is OK, then we can revise
the draft.

One open question to Dave/Jerry about the proposal is, in the event that
the source cannot be described a mathematically complete way, is there
any value (or does it make any sense at all) in sending an incomplete
qspec, or potentially "wildcarding" sending some of the mandatory
parameters.  I'm wondering if some sources / deployments may have
certain parameters that are relatively well defined, but not all are so
well defined.

thanks,
John

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 01 17:26:34 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfOWu-00086S-IT; Wed, 01 Nov 2006 17:25:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfOWt-00086N-Jf
	for nsis@ietf.org; Wed, 01 Nov 2006 17:25:51 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfOWp-0000yV-RI
	for nsis@ietf.org; Wed, 01 Nov 2006 17:25:51 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Thu, 02 Nov 2006 00:25:42 +0200
	id 000AFD46.45491EE6.00001F0D
Date: Thu, 2 Nov 2006 00:25:40 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: "Ash, Gerald R \\(Jerry\\), ALABS" <gash@att.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
Message-ID: <Pine.LNX.4.64.0611020022340.13096@sbz-31.cs.Helsinki.FI>
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 88b11fc64c1bfdb4425294ef5374ca07
Cc: "Kappler, Cornelia" <cornelia.kappler@siemens.com>,
	David R Oran <oran@cisco.com>, nsis@ietf.org,
	Cornelia Kappler <cornelia.kappler@gmx.net>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi all,

After a first read through the proposed changes, I have a good fealing 
about this. I still have to re-read and think about the proposal a few 
times more, but my first impression was very positive. :)

As QoS NSLP editor and author, I'll need to remove the notion of QSPEC 
stacking from the NSLP spec, right?

Regards,
Jukka

On Wed, 1 Nov 2006, Ash, Gerald R \(Jerry\), ALABS wrote:

> Hi All,
> 
> Per John's suggestion below we've discussed Dave Oran's suggestions at
> length and have the following proposal.  
> 
> We think the proposed changes address many concerns raised on the list
> over the last few months and will require some rather basic changes in
> the QSPEC document
> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> 
> We appreciate Dave's suggestions and careful review of the document.
> 
> Here is a summary of the proposed changes:
> 
> 1. The notion of QOSMs will be played down in the updated draft.
> Language to the effect that 'QNSLP/QSPEC can signal for different QOSMs
> across multiple domains' will be replaced by the notion that QNSLP/QSPEC
> allows different elements on the path to implement different data plane
> QoS mechanisms that meets the constraints included in the QSPEC.  A QOSM
> describes common capabilities among QNEs to act consistently when
> requested to admit traffic and in treating admitted traffic.  A QOSM ID
> need not be defined or signaled and a QNE need not support any
> particular QOSM.  A 'QOSM specification' would become a 'QNSLP/QSPEC
> applicability statement', which documents how a QNE/RMF interprets and
> treats the various elements in a QSPEC.  
> 
> An updated QOSM definition is as follows:
> 
> QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
> IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
> constraints and traffic handling directives a QNE implementing that QOSM
> is capable of supporting and how resources will be managed by the RMF in
> such a QNE.
> 
> 2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
> semantics:
> A) The source traffic description is mandatory to include by the QNI and
> mandatory to interpret by downstream QNEs.  The traffic description is
> specified by a traffic model (TMOD) parameter consisting of 4
> sub-parameters:
> - rate (r)
> - bucket size (b)
> - peak rate (p)
> - minimum policed unit (m)
> The TMOD parameter is a mathematically complete way to describe a
> traffic source.  Examples will be given for setting TMOD to correspond
> to specific cases, e.g. where only bandwidth is available (r=p; b and m
> set to default values) and other cases (e.g., TCP).  As such, the
> bandwidth parameter is eliminated.
> B) All other QSPEC parameters are optional to include by the QNI, and
> may be either mandatory or optional to interpret by downstream QNEs.
> That is, the QNI can explicitly specify for each constraint parameter
> whether it wishes admission control to succeed or fail if that
> constraint cannot be met.  A downstream QNE distinguishes whether it is
> failure to understand the parameter or failure to meet the constraint
> that causes the admission control failure (more detail on this below).
> C) Additional QSPEC parameters can be defined in updates to the QSPEC
> document (or alternatively specified in QOSM applicability statement
> documents).  There is already an IANA registry established for
> additional QSPEC parameters.
> 
> 3. The concept of remapping QSPEC parameters is eliminated.  This
> concept is ill defined and has caused much confusion (there are few if
> any sensible remappings).  Rather we define 'interpret' a QSPEC
> parameter to only mean what we now call 'strictly interpret' (i.e., must
> conform to the RFCs defining the parameter and procedures) and eliminate
> the remap option part of 'interpret'.
> 
> 4. Since remapping is eliminated, the concept of stacking QSPECs is also
> eliminated.  Discussion of tunneling of QNSLP messages is eliminated
> since there are no impacts on tunneling required by QSPEC.
> 
> 5. Current QSPEC flags are modified as follows:
> A) The QNI sets the M flag for each QSPEC parameter it populates that it
> intends must be interpreted or the reservation fails.  Currently the M
> flag is statically set for every QSPEC parameter.
> B) If the M flag is set, a downstream QNE must interpret the parameter
> or the reservation fails.  If a QNE does not support the parameter with
> the M flag set, it sets the N flag and rejects the reservation.  If the
> QNE supports the parameter but cannot meet the parameter, it sets the E
> flag and rejects the reservation. 
> C) If the M flag is not set, a downstream QNE should interpret the
> parameter.  If a QNE does not support the parameter, it sets the N flag
> and can optionally accept or reject the reservation.  If a QNE supports
> the parameter but cannot meet the parameter, it sets the E flag and
> optional accepts or rejects the reservation.
> D) The R (remapped parameter) flag and Q (non QOSM) flag are eliminated.
> 
> 6. Summary of proposed QSPEC parameters:
> 
> Source traffic Description:
>  TMOD-1 (mandatory to include)
>  TMOD-2 (optional to include)
> 
> Constraints (optional to include):
>  Path Latency
>  Path Jitter
>  Path PLR
>  Path PER
>  Y.1541 class (shorthand notation for above 4 parameters)
>  Slack term
>  Priority (Preemption, Defending, Admission, RPH Priority)
>  DSTE class
> 
> Handling directives (optional to include):
>  excess treatment
> 
> Traffic Classifiers (optional to include):
>  PHB Class (note that PHB class is set by downstream QNE to tell QNI how
> to mark traffic to ensure it will be treated as committed by admission
> control)
> 
> Eliminated:
>  bandwidth
>  Ctot, Dtot, Csum, Dsum
> 
> 7. Editorial changes:
> A) rewrite description of read-only and read-write parameters to make
> the rules more clear: 
> - which parameters can be ro/rw (function of object in which parameter
> included)
> - can only increase rw values and not delete, etc.
> B) eliminate the distinction of 'control' and 'qos description'
> parameters and replace with suggested taxonomy, as described above
> C) clarify discussion of packet outcomes from Y.1540 and place in
> appropriate sections
> D) clarify which parameters can be included in <QoS Desired> <QoS
> Available>, <QoS Reserved>, or <QoS Minimum>
> 
> Please review and comment on the proposed changes.
> 
> Thanks,
> Jerry, Dave, Cornelia
> 
> 
> -----Original Message-----
> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com] 
> Sent: Monday, October 23, 2006 2:59 AM
> To: nsis@ietf.org
> Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
> 
> Hi all,
> 
> We have had a very large thread on the current
> draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear consensus, so
> I'd like to try to summarize the discussions.
> 
> Several people feel that the current draft contains too many mandatory
> parameters and mapping some QoS Models to others is mixing
> apples-to-oranges.  The feeling is that interoperability is hard to
> achieve and therefore we should not try to gain interoperability.  Jukka
> Manner's message on 17th of October proposed a way forward.
> http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> 
> However, several (other) people feel that interoperability is important
> and is needed. Giving up on interoperability would not be a good thing,
> and that we are very close to a solution.  Dave Oran proposed a
> mechanism to reach a solution on the 19th October:
> http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> 
> I have asked Jerry and Dave to make a proposal along the lines of Dave's
> message above, and the working group can review this, to see if we can
> gain consensus on this.  I think it would be sufficient to send some
> rough text to the mailing list, and if this is OK, then we can revise
> the draft.
> 
> One open question to Dave/Jerry about the proposal is, in the event that
> the source cannot be described a mathematically complete way, is there
> any value (or does it make any sense at all) in sending an incomplete
> qspec, or potentially "wildcarding" sending some of the mandatory
> parameters.  I'm wondering if some sources / deployments may have
> certain parameters that are relatively well defined, but not all are so
> well defined.
> 
> thanks,
> John
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 01 20:59:42 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfRrC-0007ME-QR; Wed, 01 Nov 2006 20:59:02 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfRrB-0007GU-7Y
	for nsis@ietf.org; Wed, 01 Nov 2006 20:59:01 -0500
Received: from smtp.mei.co.jp ([133.183.129.25])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfRqx-0002wU-U4
	for nsis@ietf.org; Wed, 01 Nov 2006 20:59:01 -0500
Received: from mail-gw.jp.panasonic.com (dodgers.mei.co.jp [157.8.1.150])
	by smtp.mei.co.jp (8.12.11.20060614/3.7W/kings) with ESMTP id
	kA21whZb019143; Thu, 2 Nov 2006 10:58:43 +0900 (JST)
Received: by mail-gw.jp.panasonic.com (8.11.6p2/3.7W/somlx3) with ESMTP id
	kA21wj027651; Thu, 2 Nov 2006 10:58:45 +0900 (JST)
Received: from localhost (localhost [127.0.0.1])
	by mail.jp.panasonic.com (8.11.6p2/3.7W/astros) with SMTP id
	kA21wjj05812; Thu, 2 Nov 2006 10:58:45 +0900 (JST)
X-MimeOLE: Produced By Microsoft Exchange V6.5.7226.0
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-15"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] [Fwd: I-D ACTION:draft-bless-nsis-resv-aggr-00.txt]
Date: Thu, 2 Nov 2006 09:57:15 +0800
Message-ID: <5F09D220B62F79418461A978CA0921BD0148AB0E@pslexc01.psl.local>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] [Fwd: I-D ACTION:draft-bless-nsis-resv-aggr-00.txt]
Thread-Index: AcbzZLFbclxC9l7EQ8ywgRULW1uXqQKvKuYg
X-Priority: 1
Priority: Urgent
Importance: high
From: "Shivanajay Marwaha" <Shivanajay.Marwaha@sg.panasonic.com>
To: "Roland Bless" <bless@tm.uka.de>, <nsis@ietf.org>
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 10ba05e7e8a9aa6adb025f426bef3a30
Cc: Takako Sanda <sanda.takako@jp.panasonic.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland,
=20
I read your I-D and I have two questions:-
=20
1. Are you also considering route change detection (in section 3.3) =
caused by end-host's mobility.=20
2. Based on the solution in section 4.4, in a multi-homing scenario =
(mentioned in section 3.4), will multiple Reserve messages be sent to =
multiple De-aggregators on individual paths?
=20
Regards,
Shiv=20

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de]=20
> Sent: Thursday, October 19, 2006 5:55 PM
> To: nsis@ietf.org
> Subject: [NSIS] [Fwd: I-D ACTION:draft-bless-nsis-resv-aggr-00.txt]
>=20
> Hi all,
>=20
> the earlier mentioned draft was now published in the I-D repository:
>=20
> -------- Original-Message --------
> A New Internet-Draft is available from the on-line=20
> Internet-Drafts directories.
>=20
>=20
> 	Title		: Inter-Domain Reservation Aggregation=20
> for QoS NSLP
> 	Author(s)	: R. Bless, M. Doll
> 	Filename	: draft-bless-nsis-resv-aggr-00.txt
> 	Pages		: 19
> 	Date		: 2006-10-18
> =09
>    QoS NSLP is a recently proposed signaling protocol that allows to
>    establish QoS reservations in the Internet.  In order to=20
> enable large
>    scale deployment, inter-domain aggregation should be considered as
>    mechanism to allow for the necessary scalability.  This draft
>    describes the major problems that must be solved and proposes also
>    solutions to these problems, requiring only modest=20
> modifications and
>    extensions to the currently defined GIST and QoS NSLP=20
> specifications.
>=20
>=20
> A URL for this Internet-Draft is:
> http://www.ietf.org/internet-drafts/draft-bless-nsis-resv-aggr-00.txt
>=20
> To remove yourself from the I-D Announcement list, send a=20
> message to i-d-announce-request@ietf.org with the word=20
> unsubscribe in the body of the message.
> You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce
> to change your subscription settings.
>=20
> Internet-Drafts are also available by anonymous FTP. Login=20
> with the username "anonymous" and a password of your e-mail=20
> address. After logging in, type "cd internet-drafts" and then=20
> "get draft-bless-nsis-resv-aggr-00.txt".
>=20
> A list of Internet-Drafts directories can be found in=20
> http://www.ietf.org/shadow.html or=20
> ftp://ftp.ietf.org/ietf/1shadow-sites.txt
>=20
> Internet-Drafts can also be obtained by e-mail.
>=20
> Send a message to:
> 	mailserv@ietf.org.
> In the body type:
> 	"FILE /internet-drafts/draft-bless-nsis-resv-aggr-00.txt".
> =09
> NOTE:	The mail server at ietf.org can return the document in
> 	MIME-encoded form by using the "mpack" utility.  To use this
> 	feature, insert the command "ENCODING mime" before the "FILE"
> 	command.  To decode the response(s), you will need "munpack" or
> 	a MIME-compliant mail reader.  Different MIME-compliant=20
> mail readers
> 	exhibit different behavior, especially when dealing with
> 	"multipart" MIME messages (i.e. documents which have been split
> 	up into multiple messages), so check your local documentation on
> 	how to manipulate these messages.
>=20
> Below is the data which will enable a MIME compliant mail=20
> reader implementation to automatically retrieve the ASCII=20
> version of the Internet-Draft.
>=20
>=20


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 02:24:47 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfWvD-0008QW-J4; Thu, 02 Nov 2006 02:23:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfWvC-0008QR-Vx
	for nsis@ietf.org; Thu, 02 Nov 2006 02:23:30 -0500
Received: from tcmail31.telekom.de ([217.6.95.238])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfWvB-0003Mb-98
	for nsis@ietf.org; Thu, 02 Nov 2006 02:23:30 -0500
Received: from S4DE8PSAAGT.mitte.t-com.de by tcmail31.telekom.de with ESMTP;
	Thu, 2 Nov 2006 08:23:27 +0100
Received: from S4DE8PSAAFQ.mitte.t-com.de ([10.151.180.5]) by
	S4DE8PSAAGT.mitte.t-com.de with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 2 Nov 2006 08:23:27 +0100
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
X-MimeOLE: Produced By Microsoft Exchange V6.5
Date: Thu, 2 Nov 2006 08:23:26 +0100
Message-Id: <6439282641581441A36F7F6F83ED2ED2CBF91E@S4DE8PSAAFQ.mitte.t-com.de>
in-reply-to: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Proposed Updates to the QSPEC Draft
Thread-Index: Acb+AwzAtZ9rHtIFRwGQZQCF7FeA7QATBZog
From: "Geib, Ruediger" <Ruediger.Geib@t-systems.com>
To: <gash@att.com>
X-OriginalArrivalTime: 02 Nov 2006 07:23:27.0236 (UTC)
	FILETIME=[CA0D0440:01C6FE4F]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bc6181926481d86059e678c9f7cb8b34
Cc: cornelia.kappler@siemens.com, oran@cisco.com, nsis@ietf.org,
	cornelia.kappler@gmx.net
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Thanks for agreeing on such a pragmatic approach. As Jukka,=20
my impression by reading your message was a pretty good=20
one too.

Regards,

Rudiger

|-----Original Message-----
|From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
|Sent: Wednesday, November 01, 2006 11:13 PM
|To: nsis@ietf.org
|Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran;
|Kappler, Cornelia
|Subject: [NSIS] Proposed Updates to the QSPEC Draft
|
|
|Hi All,
|
|Per John's suggestion below we've discussed Dave Oran's suggestions at
|length and have the following proposal. =20
|
|We think the proposed changes address many concerns raised on the list
|over the last few months and will require some rather basic changes in
|the QSPEC document
|http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
|
|We appreciate Dave's suggestions and careful review of the document.
|
|Here is a summary of the proposed changes:
|
|1. The notion of QOSMs will be played down in the updated draft.
|Language to the effect that 'QNSLP/QSPEC can signal for different QOSMs
|across multiple domains' will be replaced by the notion that=20
|QNSLP/QSPEC
|allows different elements on the path to implement different data plane
|QoS mechanisms that meets the constraints included in the=20
|QSPEC.  A QOSM
|describes common capabilities among QNEs to act consistently when
|requested to admit traffic and in treating admitted traffic.  A QOSM ID
|need not be defined or signaled and a QNE need not support any
|particular QOSM.  A 'QOSM specification' would become a 'QNSLP/QSPEC
|applicability statement', which documents how a QNE/RMF interprets and
|treats the various elements in a QSPEC. =20
|
|An updated QOSM definition is as follows:
|
|QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
|IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
|constraints and traffic handling directives a QNE implementing=20
|that QOSM
|is capable of supporting and how resources will be managed by=20
|the RMF in
|such a QNE.
|
|2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
|semantics:
|A) The source traffic description is mandatory to include by=20
|the QNI and
|mandatory to interpret by downstream QNEs.  The traffic description is
|specified by a traffic model (TMOD) parameter consisting of 4
|sub-parameters:
|- rate (r)
|- bucket size (b)
|- peak rate (p)
|- minimum policed unit (m)
|The TMOD parameter is a mathematically complete way to describe a
|traffic source.  Examples will be given for setting TMOD to correspond
|to specific cases, e.g. where only bandwidth is available (r=3Dp; b and =
m
|set to default values) and other cases (e.g., TCP).  As such, the
|bandwidth parameter is eliminated.
|B) All other QSPEC parameters are optional to include by the QNI, and
|may be either mandatory or optional to interpret by downstream QNEs.
|That is, the QNI can explicitly specify for each constraint parameter
|whether it wishes admission control to succeed or fail if that
|constraint cannot be met.  A downstream QNE distinguishes whether it is
|failure to understand the parameter or failure to meet the constraint
|that causes the admission control failure (more detail on this below).
|C) Additional QSPEC parameters can be defined in updates to the QSPEC
|document (or alternatively specified in QOSM applicability statement
|documents).  There is already an IANA registry established for
|additional QSPEC parameters.
|
|3. The concept of remapping QSPEC parameters is eliminated.  This
|concept is ill defined and has caused much confusion (there are few if
|any sensible remappings).  Rather we define 'interpret' a QSPEC
|parameter to only mean what we now call 'strictly interpret'=20
|(i.e., must
|conform to the RFCs defining the parameter and procedures) and=20
|eliminate
|the remap option part of 'interpret'.
|
|4. Since remapping is eliminated, the concept of stacking=20
|QSPECs is also
|eliminated.  Discussion of tunneling of QNSLP messages is eliminated
|since there are no impacts on tunneling required by QSPEC.
|
|5. Current QSPEC flags are modified as follows:
|A) The QNI sets the M flag for each QSPEC parameter it=20
|populates that it
|intends must be interpreted or the reservation fails.  Currently the M
|flag is statically set for every QSPEC parameter.
|B) If the M flag is set, a downstream QNE must interpret the parameter
|or the reservation fails.  If a QNE does not support the parameter with
|the M flag set, it sets the N flag and rejects the reservation.  If the
|QNE supports the parameter but cannot meet the parameter, it sets the E
|flag and rejects the reservation.=20
|C) If the M flag is not set, a downstream QNE should interpret the
|parameter.  If a QNE does not support the parameter, it sets the N flag
|and can optionally accept or reject the reservation.  If a QNE supports
|the parameter but cannot meet the parameter, it sets the E flag and
|optional accepts or rejects the reservation.
|D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
|eliminated.
|
|6. Summary of proposed QSPEC parameters:
|
|Source traffic Description:
| TMOD-1 (mandatory to include)
| TMOD-2 (optional to include)
|
|Constraints (optional to include):
| Path Latency
| Path Jitter
| Path PLR
| Path PER
| Y.1541 class (shorthand notation for above 4 parameters)
| Slack term
| Priority (Preemption, Defending, Admission, RPH Priority)
| DSTE class
|
|Handling directives (optional to include):
| excess treatment
|
|Traffic Classifiers (optional to include):
| PHB Class (note that PHB class is set by downstream QNE to=20
|tell QNI how
|to mark traffic to ensure it will be treated as committed by admission
|control)
|
|Eliminated:
| bandwidth
| Ctot, Dtot, Csum, Dsum
|
|7. Editorial changes:
|A) rewrite description of read-only and read-write parameters to make
|the rules more clear:=20
|- which parameters can be ro/rw (function of object in which parameter
|included)
|- can only increase rw values and not delete, etc.
|B) eliminate the distinction of 'control' and 'qos description'
|parameters and replace with suggested taxonomy, as described above
|C) clarify discussion of packet outcomes from Y.1540 and place in
|appropriate sections
|D) clarify which parameters can be included in <QoS Desired> <QoS
|Available>, <QoS Reserved>, or <QoS Minimum>
|
|Please review and comment on the proposed changes.
|
|Thanks,
|Jerry, Dave, Cornelia
|
|
|-----Original Message-----
|From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]=20
|Sent: Monday, October 23, 2006 2:59 AM
|To: nsis@ietf.org
|Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
|
|Hi all,
|
|We have had a very large thread on the current
|draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear=20
|consensus, so
|I'd like to try to summarize the discussions.
|
|Several people feel that the current draft contains too many mandatory
|parameters and mapping some QoS Models to others is mixing
|apples-to-oranges.  The feeling is that interoperability is hard to
|achieve and therefore we should not try to gain=20
|interoperability.  Jukka
|Manner's message on 17th of October proposed a way forward.
|http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
|
|However, several (other) people feel that interoperability is important
|and is needed. Giving up on interoperability would not be a good thing,
|and that we are very close to a solution.  Dave Oran proposed a
|mechanism to reach a solution on the 19th October:
|http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
|
|I have asked Jerry and Dave to make a proposal along the lines=20
|of Dave's
|message above, and the working group can review this, to see if we can
|gain consensus on this.  I think it would be sufficient to send some
|rough text to the mailing list, and if this is OK, then we can revise
|the draft.
|
|One open question to Dave/Jerry about the proposal is, in the=20
|event that
|the source cannot be described a mathematically complete way, is there
|any value (or does it make any sense at all) in sending an incomplete
|qspec, or potentially "wildcarding" sending some of the mandatory
|parameters.  I'm wondering if some sources / deployments may have
|certain parameters that are relatively well defined, but not all are so
|well defined.
|
|thanks,
|John
|
|_______________________________________________
|nsis mailing list
|nsis@ietf.org
|https://www1.ietf.org/mailman/listinfo/nsis
|
|_______________________________________________
|nsis mailing list
|nsis@ietf.org
|https://www1.ietf.org/mailman/listinfo/nsis
|

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 02:40:04 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfXAh-0000cR-GU; Thu, 02 Nov 2006 02:39:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfXAe-0000ZM-8k
	for nsis@ietf.org; Thu, 02 Nov 2006 02:39:28 -0500
Received: from mgw-ext12.nokia.com ([131.228.20.171])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfX6b-0004sk-19
	for nsis@ietf.org; Thu, 02 Nov 2006 02:35:17 -0500
Received: from esebh108.NOE.Nokia.com (esebh108.ntc.nokia.com [172.21.143.145])
	by mgw-ext12.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA27YxWf013750; Thu, 2 Nov 2006 09:35:09 +0200
Received: from esebh104.NOE.Nokia.com ([172.21.143.34]) by
	esebh108.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 2 Nov 2006 09:35:04 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh104.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 2 Nov 2006 09:35:03 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Thu, 2 Nov 2006 09:35:03 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED049A@esebe199.NOE.Nokia.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQAEfCIzA=
From: <john.loughney@nokia.com>
To: <gash@att.com>, <nsis@ietf.org>
X-OriginalArrivalTime: 02 Nov 2006 07:35:03.0973 (UTC)
	FILETIME=[6956A550:01C6FE51]
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 325b777e1a3a618c889460b612a65510
Cc: cornelia.kappler@siemens.com, oran@cisco.com, cornelia.kappler@gmx.net
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Jerry, Dave and Cornelia;

This looks like a very good simplification of the QSPEC, and I think
that is a good idea.  I think that there will be some implications to
the QoSM drafts - I think QoS Model drafts will need to contain some
more=20
applicability statement type of text, along with deployment
considerations.

A plus side of this is that I think that the QoS template will become
more broadly applicable to other uses than just NSIS QoS Signaling,
so I think that is a gain as well.

I'd really like people to review this and send comments on it so that
we could really focus the discussion next week and get this work
finally resolved.

Thanks a lot for working through this, I think this is a good basis
for a proposal.

thanks,
John

>-----Original Message-----
>From: ext Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]=20
>Sent: 02 November, 2006 00:13
>To: nsis@ietf.org
>Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R=20
>Oran; Kappler,Cornelia
>Subject: [NSIS] Proposed Updates to the QSPEC Draft
>
>Hi All,
>
>Per John's suggestion below we've discussed Dave Oran's=20
>suggestions at length and have the following proposal. =20
>
>We think the proposed changes address many concerns raised on=20
>the list over the last few months and will require some rather=20
>basic changes in the QSPEC document=20
>http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
>
>We appreciate Dave's suggestions and careful review of the document.
>
>Here is a summary of the proposed changes:
>
>1. The notion of QOSMs will be played down in the updated draft.
>Language to the effect that 'QNSLP/QSPEC can signal for=20
>different QOSMs across multiple domains' will be replaced by=20
>the notion that QNSLP/QSPEC allows different elements on the=20
>path to implement different data plane QoS mechanisms that=20
>meets the constraints included in the QSPEC.  A QOSM describes=20
>common capabilities among QNEs to act consistently when=20
>requested to admit traffic and in treating admitted traffic. =20
>A QOSM ID need not be defined or signaled and a QNE need not=20
>support any particular QOSM.  A 'QOSM specification' would=20
>become a 'QNSLP/QSPEC applicability statement', which=20
>documents how a QNE/RMF interprets and treats the various=20
>elements in a QSPEC. =20
>
>An updated QOSM definition is as follows:
>
>QoS Model (QOSM): A method to achieve QoS for a traffic flow,=20
>e.g., IntServ Controlled Load.  A QOSM specifies what sub-set=20
>of QSPEC QoS constraints and traffic handling directives a QNE=20
>implementing that QOSM is capable of supporting and how=20
>resources will be managed by the RMF in such a QNE.
>
>2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
>semantics:
>A) The source traffic description is mandatory to include by=20
>the QNI and mandatory to interpret by downstream QNEs.  The=20
>traffic description is specified by a traffic model (TMOD)=20
>parameter consisting of 4
>sub-parameters:
>- rate (r)
>- bucket size (b)
>- peak rate (p)
>- minimum policed unit (m)
>The TMOD parameter is a mathematically complete way to=20
>describe a traffic source.  Examples will be given for setting=20
>TMOD to correspond to specific cases, e.g. where only=20
>bandwidth is available (r=3Dp; b and m set to default values)=20
>and other cases (e.g., TCP).  As such, the bandwidth parameter=20
>is eliminated.
>B) All other QSPEC parameters are optional to include by the=20
>QNI, and may be either mandatory or optional to interpret by=20
>downstream QNEs.
>That is, the QNI can explicitly specify for each constraint=20
>parameter whether it wishes admission control to succeed or=20
>fail if that constraint cannot be met.  A downstream QNE=20
>distinguishes whether it is failure to understand the=20
>parameter or failure to meet the constraint that causes the=20
>admission control failure (more detail on this below).
>C) Additional QSPEC parameters can be defined in updates to=20
>the QSPEC document (or alternatively specified in QOSM=20
>applicability statement documents).  There is already an IANA=20
>registry established for additional QSPEC parameters.
>
>3. The concept of remapping QSPEC parameters is eliminated. =20
>This concept is ill defined and has caused much confusion=20
>(there are few if any sensible remappings).  Rather we define=20
>'interpret' a QSPEC parameter to only mean what we now call=20
>'strictly interpret' (i.e., must conform to the RFCs defining=20
>the parameter and procedures) and eliminate the remap option=20
>part of 'interpret'.
>
>4. Since remapping is eliminated, the concept of stacking=20
>QSPECs is also eliminated.  Discussion of tunneling of QNSLP=20
>messages is eliminated since there are no impacts on tunneling=20
>required by QSPEC.
>
>5. Current QSPEC flags are modified as follows:
>A) The QNI sets the M flag for each QSPEC parameter it=20
>populates that it intends must be interpreted or the=20
>reservation fails.  Currently the M flag is statically set for=20
>every QSPEC parameter.
>B) If the M flag is set, a downstream QNE must interpret the=20
>parameter or the reservation fails.  If a QNE does not support=20
>the parameter with the M flag set, it sets the N flag and=20
>rejects the reservation.  If the QNE supports the parameter=20
>but cannot meet the parameter, it sets the E flag and rejects=20
>the reservation.=20
>C) If the M flag is not set, a downstream QNE should interpret=20
>the parameter.  If a QNE does not support the parameter, it=20
>sets the N flag and can optionally accept or reject the=20
>reservation.  If a QNE supports the parameter but cannot meet=20
>the parameter, it sets the E flag and optional accepts or=20
>rejects the reservation.
>D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
>eliminated.
>
>6. Summary of proposed QSPEC parameters:
>
>Source traffic Description:
> TMOD-1 (mandatory to include)
> TMOD-2 (optional to include)
>
>Constraints (optional to include):
> Path Latency
> Path Jitter
> Path PLR
> Path PER
> Y.1541 class (shorthand notation for above 4 parameters) =20
>Slack term  Priority (Preemption, Defending, Admission, RPH=20
>Priority)  DSTE class
>
>Handling directives (optional to include):
> excess treatment
>
>Traffic Classifiers (optional to include):
> PHB Class (note that PHB class is set by downstream QNE to=20
>tell QNI how to mark traffic to ensure it will be treated as=20
>committed by admission
>control)
>
>Eliminated:
> bandwidth
> Ctot, Dtot, Csum, Dsum
>
>7. Editorial changes:
>A) rewrite description of read-only and read-write parameters=20
>to make the rules more clear:=20
>- which parameters can be ro/rw (function of object in which parameter
>included)
>- can only increase rw values and not delete, etc.
>B) eliminate the distinction of 'control' and 'qos description'
>parameters and replace with suggested taxonomy, as described above
>C) clarify discussion of packet outcomes from Y.1540 and place=20
>in appropriate sections
>D) clarify which parameters can be included in <QoS Desired> <QoS
>Available>, <QoS Reserved>, or <QoS Minimum>
>
>Please review and comment on the proposed changes.
>
>Thanks,
>Jerry, Dave, Cornelia
>
>
>-----Original Message-----
>From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
>Sent: Monday, October 23, 2006 2:59 AM
>To: nsis@ietf.org
>Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
>
>Hi all,
>
>We have had a very large thread on the current=20
>draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear=20
>consensus, so I'd like to try to summarize the discussions.
>
>Several people feel that the current draft contains too many=20
>mandatory parameters and mapping some QoS Models to others is=20
>mixing apples-to-oranges.  The feeling is that=20
>interoperability is hard to achieve and therefore we should=20
>not try to gain interoperability.  Jukka Manner's message on=20
>17th of October proposed a way forward.
>http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
>
>However, several (other) people feel that interoperability is=20
>important and is needed. Giving up on interoperability would=20
>not be a good thing, and that we are very close to a solution.=20
> Dave Oran proposed a mechanism to reach a solution on the=20
>19th October:
>http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
>
>I have asked Jerry and Dave to make a proposal along the lines=20
>of Dave's message above, and the working group can review=20
>this, to see if we can gain consensus on this.  I think it=20
>would be sufficient to send some rough text to the mailing=20
>list, and if this is OK, then we can revise the draft.
>
>One open question to Dave/Jerry about the proposal is, in the=20
>event that the source cannot be described a mathematically=20
>complete way, is there any value (or does it make any sense at=20
>all) in sending an incomplete qspec, or potentially=20
>"wildcarding" sending some of the mandatory parameters.  I'm=20
>wondering if some sources / deployments may have certain=20
>parameters that are relatively well defined, but not all are=20
>so well defined.
>
>thanks,
>John
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 02:40:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfXAp-0000l9-PR; Thu, 02 Nov 2006 02:39:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfXAp-0000iZ-0o
	for nsis@ietf.org; Thu, 02 Nov 2006 02:39:39 -0500
Received: from nz-out-0102.google.com ([64.233.162.198])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfX0w-00043K-Pq
	for nsis@ietf.org; Thu, 02 Nov 2006 02:29:29 -0500
Received: by nz-out-0102.google.com with SMTP id 13so71809nzn
	for <nsis@ietf.org>; Wed, 01 Nov 2006 23:29:26 -0800 (PST)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com;
	h=received:message-id:date:from:to:subject:mime-version:content-type;
	b=H02XLGS1TQCH2FzKPgF/81fGrIp0Q9HbxM5WvPauCIZHAZZ3180DJdLj/CY/wIgRYjsoEQKvI/STHOz7oFOZrL0iR6OxiCKVSG9y93kEgJf2q9NUt0pNW74/HtAqcq0RpUtkz9mHqzvIiGvPALzG1ITYmziu1Yssi0zNYclEmHw=
Received: by 10.64.253.12 with SMTP id a12mr183601qbi.1162452565650;
	Wed, 01 Nov 2006 23:29:25 -0800 (PST)
Received: by 10.64.27.8 with HTTP; Wed, 1 Nov 2006 23:29:25 -0800 (PST)
Message-ID: <5a26be460611012329p547880b9k800be90516dace8d@mail.gmail.com>
Date: Thu, 2 Nov 2006 15:29:25 +0800
From: "=?GB2312?B?sfOx8831?=" <binbinwang118@gmail.com>
To: nsis@ietf.org
MIME-Version: 1.0
X-Spam-Score: 0.5 (/)
X-Scan-Signature: 32b73d73e8047ed17386f9799119ce43
Subject: [NSIS] question about GIST-Query message
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0987469230=="
Errors-To: nsis-bounces@ietf.org

--===============0987469230==
Content-Type: multipart/alternative; 
	boundary="----=_Part_925_1952694.1162452565619"

------=_Part_925_1952694.1162452565619
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hi all,

I have a question about GIST-Query message. Whether the GIST-Query message
is group sending in networking?
                         +--+
                          |B |
                        .+--+.
                       .        .
                  +-+          . +--+
                  |A |.            | D|
                  +-+.          .+--+
                        .        .
                         .+--+.
                          |C  |
                          +--+
Assume node A, node B, node C, and node D, they all support the same NSLP.
Node A sends a GIST-Query message, namely a UDP datagram addressed to the
node D and include an IP RAO. whether both node B and node C will receive
the Query message? And which node that node A will establishes GIST state
with it? Finally the signaling path is A >> B >> D, or A >> C >> D ?

Bestwishes!

Wang
2006-11-2



-- 
skyWang.bupt

------=_Part_925_1952694.1162452565619
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

<div>Hi all,</div>
<div>&nbsp;</div>
<div>I have a question about GIST-Query message. Whether the&nbsp;GIST-Quer=
y message is group sending in networking?</div>
<div>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; =
+--+&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
; <br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp=
; |B |<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; .=
+--+.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p; <br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; .&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp; &nbsp; .&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <br>&nbs=
p;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp; +-+&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=
&nbsp; . +--+<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |A |.&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; | D|
<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; +-+.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp; .+--+<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; .&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp; <br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=
&nbsp;&nbsp;&nbsp; .+--+.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; <br>&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; |C=
 &nbsp;|<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=
&nbsp;&nbsp; +--+&nbsp;</div>
<div>Assume node A, node B, node C, and node D, they all support the same N=
SLP. Node A sends a GIST-Query message, namely a UDP datagram addressed to =
the node D and include an IP RAO. whether both node B and node C will recei=
ve the Query message? And which node that node A will establishes GIST stat=
e with it? Finally the signaling path is A &gt;&gt; B &gt;&gt; D, or A &gt;=
&gt; C &gt;&gt; D ?
</div>
<div>&nbsp;</div>
<div>Bestwishes!</div>
<div>&nbsp;</div>
<div>Wang</div>
<div>2006-11-2</div>
<div>&nbsp;</div>
<div>&nbsp; </div>
<div><br>-- <br>skyWang.bupt </div>

------=_Part_925_1952694.1162452565619--


--===============0987469230==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============0987469230==--




From nsis-bounces@ietf.org Thu Nov 02 05:40:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfZyy-0006fM-DQ; Thu, 02 Nov 2006 05:39:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfZyx-0006dT-4V
	for nsis@ietf.org; Thu, 02 Nov 2006 05:39:35 -0500
Received: from tart.dcs.qmul.ac.uk ([138.37.95.139] helo=mail.dcs.qmul.ac.uk)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfZyv-0004Da-MB
	for nsis@ietf.org; Thu, 02 Nov 2006 05:39:35 -0500
Received: from hst138-37-90-32.dcs.qmul.ac.uk ([138.37.90.32] helo=[127.0.0.1])
	by mail.dcs.qmul.ac.uk with esmtpsa (TLSv1:DHE-RSA-AES256-SHA:256)
	(Exim 4.62) (envelope-from <jian.zhang@dcs.qmul.ac.uk>)
	id 1GfZyW-0003os-Tv; Thu, 02 Nov 2006 10:39:09 +0000
Message-ID: <4549CACD.2000209@dcs.qmul.ac.uk>
Date: Thu, 02 Nov 2006 10:39:09 +0000
From: Jian Zhang <jian.zhang@dcs.qmul.ac.uk>
Organization: Queen Mary, Univ. of London
User-Agent: Thunderbird 1.5.0.7 (Windows/20060909)
MIME-Version: 1.0
To: john.loughney@nokia.com
Subject: Re: [NSIS] The version 3 of the InterDomain-QOSM Draft: the NSIS
	QOS Model to Fulfill the E2E QoS Control in the ITU-T RACF Functional
	Architecture
References: <BAA65A575825454CBB0103267553FCCCED0466@esebe199.NOE.Nokia.com>
In-Reply-To: <BAA65A575825454CBB0103267553FCCCED0466@esebe199.NOE.Nokia.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-DCS-Auth-User: jianzhang
X-DCS-Whoson-User: jianzhang
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 386e0819b1192672467565a524848168
Cc: edmundo@dei.uc.pt, karagian@cs.utwente.nl, mendes@docomolab-euro.com,
	nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
Reply-To: jian.zhang@dcs.qmul.ac.uk
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hello John,

Currently I have not involved in the ITU-T RACF activities and I am not 
aware if Paulo and Jorge are involving with these activities
or not, but I believe the EU IST FP6 EuQoS project, as a whole, is 
involved in the discussions (Edmundo knows better about that).

Meanwhile, Georgios will discuss with the ITU-T people about these 
issues during the coming IETF meeting.

Cheers,

Jian

john.loughney@nokia.com wrote:
> Jian,
>
> Are any of the authors participating in the ITU-T RACF discussions?
>
> thanks,
> John 
>
>   
>> -----Original Message-----
>> From: ext Jian Zhang [mailto:jian.zhang@dcs.qmul.ac.uk] 
>> Sent: 31 October, 2006 17:18
>> To: nsis@ietf.org
>> Cc: edmundo@dei.uc.pt; karagian@cs.utwente.nl; 
>> mendes@docomolab-euro.com
>> Subject: [NSIS] The version 3 of the InterDomain-QOSM Draft: 
>> the NSIS QOS Model to Fulfill the E2E QoS Control in the ITU-T 
>> RACF Functional Architecture
>>
>> Dear all,
>>
>> Please kindly read the version 3 of the InterDomain-QOSM 
>> Draft: our effort to define a NSIS QOS Model to fulfill the 
>> E2E QoS control in the ITU-T RACF functional architecture, 
>> which has been submitted to the NSIS WG for discussion at the 
>> coming IETF meeting.
>>
>> And if you have any questions or suggestions about this draft, 
>> we would like to answer or discuss with you in this mailing 
>> list. The abstract of this draft is attached below and the 
>> link which holds the whole contents of the draft is also 
>> provided below.
>>
>> Cheers,
>>
>> Jian Zhang
>>
>> Dept. of Computer Science
>> Queen Mary, Univ. of London
>>
>> ---------------------------------------
>>
>> Abstract
>>
>>   This document has three goals. First of all, it presents 
>> our analysis
>>   of how to use the NSIS signaling (inter-domain QOSM and intra-domain
>>   QOSM) to fulfill the QoS control in accord with the ITU-T RACF
>>   functional architecture. For this goal, we discuss how the 
>> ITU-T RACF
>>   entities in the ITU-T RACF functional architecture can be mapped to
>>   the NSIS entities and how the RACF reference points can be
>>   implemented by using the NSIS protocol suites and QOSMs. 
>> Secondly, we
>>   aim at specifying an NSIS Inter-domain QOSM for E2E QoS control
>>   across heterogeneous IP networks and applying this Inter-domain QOSM
>>   to the e2e QoS control in the ITU-T RACF functional architecture
>>   based on the above ITU-T RACF analysis. The detailed description of
>>   the NSIS Inter-domain QOSM are given and the e2e QoS control
>>   scenarios in the ITU-T RACF architecture (including RACF Push and
>>   Pull resource control modes), which will be covered by the NSIS
>>   Inter-domain QOSM are described and implemented. Thirdly, we specify
>>   and implement those QSPECs that will be used by the 
>> Inter-domain QOSM
>>   for the e2e QoS control in the ITU-T RACF architecture.
>>
>> The link is
>> http://ietfreport.isoc.org/ids/draft-zhang-nsis-interdomain-qos
>> m-03.txt .
>>
>>
>> _______________________________________________
>> nsis mailing list
>> nsis@ietf.org
>> https://www1.ietf.org/mailman/listinfo/nsis
>>
>>     


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 06:01:13 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfaJK-0001wb-02; Thu, 02 Nov 2006 06:00:38 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfaJI-0001nn-2m
	for nsis@ietf.org; Thu, 02 Nov 2006 06:00:36 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfaJH-0000jY-EZ
	for nsis@ietf.org; Thu, 02 Nov 2006 06:00:36 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA2B0QIw014081; Thu, 2 Nov 2006 12:00:31 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <john.loughney@nokia.com>, <nsis@ietf.org>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Thu, 2 Nov 2006 12:00:23 +0100
Message-ID: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQAEfCIzAABuAfQA==
In-Reply-To: <BAA65A575825454CBB0103267553FCCCED049A@esebe199.NOE.Nokia.com>
X-Spam-Score: 0.252 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Thu, 02 Nov 2006 12:00:32 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 918f4bd8440e8de4700bcf6d658bc801
Cc: "'Lars Westberg \(KI/EAB\)'" <Lars.westberg@ericsson.com>,
	=?iso-8859-1?Q?'Attila_B=E1der_=28IJ/ETH=29'?= <attila.bader@ericsson.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi John

I have just come back from a trip and I have just look into point 4):
> >4. Since remapping is eliminated, the concept of stacking QSPECs is 
> >also eliminated.  Discussion of tunneling of QNSLP messages is 
> >eliminated since there are no impacts on tunneling required by QSPEC.

QSPEC stacking is one of the significant concepts in the QoS based NSIS
solution.
This concept provides the additional flexibility that is missing from RSVP.
Furthermore, all the work that we have done until now is based on 
the concept of QSPEC stacking. Thus what you actually say is to just 
forget the work that we have done
until now, in e.g.the RMD-QOSM draft, and start from the beginning.


The main problem that we have had until now was not the QSPEC stacking
issue,
but it was the huge number of mandatory paremeters. Actualy I consider the
issue of removing the 
QSPEC stacking concept a much more severe problem on achieving the
flexibility 
that we wanted to achieve with NSIS than the huge number of mandatory
parameters.

Therefore, I am severely against removing the QSPEC stacking concept!

Best Regards,
Georgios


 
 

> -----Original Message-----
> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com] 
> Sent: donderdag 2 november 2006 8:35
> To: gash@att.com; nsis@ietf.org
> Cc: cornelia.kappler@siemens.com; oran@cisco.com; 
> cornelia.kappler@gmx.net
> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> 
> Jerry, Dave and Cornelia;
> 
> This looks like a very good simplification of the QSPEC, and 
> I think that is a good idea.  I think that there will be some 
> implications to the QoSM drafts - I think QoS Model drafts 
> will need to contain some more applicability statement type 
> of text, along with deployment considerations.
> 
> A plus side of this is that I think that the QoS template 
> will become more broadly applicable to other uses than just 
> NSIS QoS Signaling, so I think that is a gain as well.
> 
> I'd really like people to review this and send comments on it 
> so that we could really focus the discussion next week and 
> get this work finally resolved.
> 
> Thanks a lot for working through this, I think this is a good 
> basis for a proposal.
> 
> thanks,
> John
> 
> >-----Original Message-----
> >From: ext Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
> >Sent: 02 November, 2006 00:13
> >To: nsis@ietf.org
> >Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran; 
> >Kappler,Cornelia
> >Subject: [NSIS] Proposed Updates to the QSPEC Draft
> >
> >Hi All,
> >
> >Per John's suggestion below we've discussed Dave Oran's 
> suggestions at 
> >length and have the following proposal.
> >
> >We think the proposed changes address many concerns raised 
> on the list 
> >over the last few months and will require some rather basic 
> changes in 
> >the QSPEC document 
> >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> >
> >We appreciate Dave's suggestions and careful review of the document.
> >
> >Here is a summary of the proposed changes:
> >
> >1. The notion of QOSMs will be played down in the updated draft.
> >Language to the effect that 'QNSLP/QSPEC can signal for 
> different QOSMs 
> >across multiple domains' will be replaced by the notion that 
> >QNSLP/QSPEC allows different elements on the path to implement 
> >different data plane QoS mechanisms that meets the 
> constraints included 
> >in the QSPEC.  A QOSM describes common capabilities among 
> QNEs to act 
> >consistently when requested to admit traffic and in treating 
> admitted 
> >traffic.
> >A QOSM ID need not be defined or signaled and a QNE need not support 
> >any particular QOSM.  A 'QOSM specification' would become a 
> >'QNSLP/QSPEC applicability statement', which documents how a QNE/RMF 
> >interprets and treats the various elements in a QSPEC.
> >
> >An updated QOSM definition is as follows:
> >
> >QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g., 
> >IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS 
> >constraints and traffic handling directives a QNE implementing that 
> >QOSM is capable of supporting and how resources will be 
> managed by the 
> >RMF in such a QNE.
> >
> >2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
> >semantics:
> >A) The source traffic description is mandatory to include by the QNI 
> >and mandatory to interpret by downstream QNEs.  The traffic 
> description 
> >is specified by a traffic model (TMOD) parameter consisting of 4
> >sub-parameters:
> >- rate (r)
> >- bucket size (b)
> >- peak rate (p)
> >- minimum policed unit (m)
> >The TMOD parameter is a mathematically complete way to describe a 
> >traffic source.  Examples will be given for setting TMOD to 
> correspond 
> >to specific cases, e.g. where only bandwidth is available 
> (r=p; b and m 
> >set to default values) and other cases (e.g., TCP).  As such, the 
> >bandwidth parameter is eliminated.
> >B) All other QSPEC parameters are optional to include by the 
> QNI, and 
> >may be either mandatory or optional to interpret by downstream QNEs.
> >That is, the QNI can explicitly specify for each constraint 
> parameter 
> >whether it wishes admission control to succeed or fail if that 
> >constraint cannot be met.  A downstream QNE distinguishes 
> whether it is 
> >failure to understand the parameter or failure to meet the 
> constraint 
> >that causes the admission control failure (more detail on 
> this below).
> >C) Additional QSPEC parameters can be defined in updates to 
> the QSPEC 
> >document (or alternatively specified in QOSM applicability statement 
> >documents).  There is already an IANA registry established for 
> >additional QSPEC parameters.
> >
> >3. The concept of remapping QSPEC parameters is eliminated.  
> >This concept is ill defined and has caused much confusion (there are 
> >few if any sensible remappings).  Rather we define 
> 'interpret' a QSPEC 
> >parameter to only mean what we now call 'strictly interpret' (i.e., 
> >must conform to the RFCs defining the parameter and procedures) and 
> >eliminate the remap option part of 'interpret'.
> >
> >4. Since remapping is eliminated, the concept of stacking QSPECs is 
> >also eliminated.  Discussion of tunneling of QNSLP messages is 
> >eliminated since there are no impacts on tunneling required by QSPEC.
> >
> >5. Current QSPEC flags are modified as follows:
> >A) The QNI sets the M flag for each QSPEC parameter it 
> populates that 
> >it intends must be interpreted or the reservation fails.  
> Currently the 
> >M flag is statically set for every QSPEC parameter.
> >B) If the M flag is set, a downstream QNE must interpret the 
> parameter 
> >or the reservation fails.  If a QNE does not support the 
> parameter with 
> >the M flag set, it sets the N flag and rejects the 
> reservation.  If the 
> >QNE supports the parameter but cannot meet the parameter, it 
> sets the E 
> >flag and rejects the reservation.
> >C) If the M flag is not set, a downstream QNE should interpret the 
> >parameter.  If a QNE does not support the parameter, it sets 
> the N flag 
> >and can optionally accept or reject the reservation.  If a 
> QNE supports 
> >the parameter but cannot meet the parameter, it sets the E flag and 
> >optional accepts or rejects the reservation.
> >D) The R (remapped parameter) flag and Q (non QOSM) flag are 
> >eliminated.
> >
> >6. Summary of proposed QSPEC parameters:
> >
> >Source traffic Description:
> > TMOD-1 (mandatory to include)
> > TMOD-2 (optional to include)
> >
> >Constraints (optional to include):
> > Path Latency
> > Path Jitter
> > Path PLR
> > Path PER
> > Y.1541 class (shorthand notation for above 4 parameters) 
> Slack term  
> >Priority (Preemption, Defending, Admission, RPH
> >Priority)  DSTE class
> >
> >Handling directives (optional to include):
> > excess treatment
> >
> >Traffic Classifiers (optional to include):
> > PHB Class (note that PHB class is set by downstream QNE to tell QNI 
> >how to mark traffic to ensure it will be treated as committed by 
> >admission
> >control)
> >
> >Eliminated:
> > bandwidth
> > Ctot, Dtot, Csum, Dsum
> >
> >7. Editorial changes:
> >A) rewrite description of read-only and read-write 
> parameters to make 
> >the rules more clear:
> >- which parameters can be ro/rw (function of object in which 
> parameter
> >included)
> >- can only increase rw values and not delete, etc.
> >B) eliminate the distinction of 'control' and 'qos description'
> >parameters and replace with suggested taxonomy, as described above
> >C) clarify discussion of packet outcomes from Y.1540 and place in 
> >appropriate sections
> >D) clarify which parameters can be included in <QoS Desired> <QoS
> >Available>, <QoS Reserved>, or <QoS Minimum>
> >
> >Please review and comment on the proposed changes.
> >
> >Thanks,
> >Jerry, Dave, Cornelia
> >
> >
> >-----Original Message-----
> >From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> >Sent: Monday, October 23, 2006 2:59 AM
> >To: nsis@ietf.org
> >Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
> >
> >Hi all,
> >
> >We have had a very large thread on the current 
> >draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear 
> consensus, 
> >so I'd like to try to summarize the discussions.
> >
> >Several people feel that the current draft contains too many 
> mandatory 
> >parameters and mapping some QoS Models to others is mixing 
> >apples-to-oranges.  The feeling is that interoperability is hard to 
> >achieve and therefore we should not try to gain interoperability.  
> >Jukka Manner's message on 17th of October proposed a way forward.
> >http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> >
> >However, several (other) people feel that interoperability 
> is important 
> >and is needed. Giving up on interoperability would not be a 
> good thing, 
> >and that we are very close to a solution.
> > Dave Oran proposed a mechanism to reach a solution on the 19th 
> >October:
> >http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> >
> >I have asked Jerry and Dave to make a proposal along the lines of 
> >Dave's message above, and the working group can review this, 
> to see if 
> >we can gain consensus on this.  I think it would be 
> sufficient to send 
> >some rough text to the mailing list, and if this is OK, then we can 
> >revise the draft.
> >
> >One open question to Dave/Jerry about the proposal is, in the event 
> >that the source cannot be described a mathematically 
> complete way, is 
> >there any value (or does it make any sense at
> >all) in sending an incomplete qspec, or potentially "wildcarding" 
> >sending some of the mandatory parameters.  I'm wondering if some 
> >sources / deployments may have certain parameters that are 
> relatively 
> >well defined, but not all are so well defined.
> >
> >thanks,
> >John
> >
> >_______________________________________________
> >nsis mailing list
> >nsis@ietf.org
> >https://www1.ietf.org/mailman/listinfo/nsis
> >
> >_______________________________________________
> >nsis mailing list
> >nsis@ietf.org
> >https://www1.ietf.org/mailman/listinfo/nsis
> >
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 08:58:03 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfd4O-00085M-6E; Thu, 02 Nov 2006 08:57:24 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfd4M-00085H-Nn
	for nsis@ietf.org; Thu, 02 Nov 2006 08:57:22 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1Gfd4H-0004vD-TK
	for nsis@ietf.org; Thu, 02 Nov 2006 08:57:22 -0500
Received: from sj-dkim-4.cisco.com ([171.71.179.196])
	by sj-iport-3.cisco.com with ESMTP; 02 Nov 2006 05:57:16 -0800
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AgAAANhNSUWrR7PEh2dsb2JhbACMSgEBCQ4q
X-IronPort-AV: i="4.09,380,1157353200"; 
	d="scan'208"; a="447698922:sNHT38149508"
Received: from sj-core-2.cisco.com (sj-core-2.cisco.com [171.71.177.254])
	by sj-dkim-4.cisco.com (8.12.11.20060308/8.12.11) with ESMTP id
	kA2DvGvA020833; Thu, 2 Nov 2006 05:57:16 -0800
Received: from imail.cisco.com (sjc12-sbr-sw3-3f5.cisco.com [172.19.96.182])
	by sj-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id kA2DvGin011036;
	Thu, 2 Nov 2006 05:57:16 -0800 (PST)
Received: from [10.32.245.152] (stealth-10-32-245-152.cisco.com
	[10.32.245.152])
	by imail.cisco.com (8.12.11/8.12.10) with ESMTP id kA2DiJi3011018;
	Thu, 2 Nov 2006 05:44:19 -0800
In-Reply-To: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>
Content-Transfer-Encoding: 7bit
From: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
Date: Thu, 2 Nov 2006 08:57:10 -0500
To: Georgios Karagiannis <karagian@cs.utwente.nl>
X-Mailer: Apple Mail (2.752.2)
Authentication-Results: sj-dkim-4.cisco.com; header.From=oran@cisco.com;
	dkim=pass ( sig from cisco.com verified; ); 
DKIM-Signature: a=rsa-sha1; q=dns; l=13498; t=1162475836; x=1163339836;
	c=relaxed/simple; s=sjdkim4002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=oran@cisco.com;
	z=From:David=20R=20Oran=20<oran@cisco.com>
	|Subject:Re=3A=20[NSIS]=20Proposed=20Updates=20to=20the=20QSPEC=20Draft;
	X=v=3Dcisco.com=3B=20h=3Da2zo2nRtey2Ue6oN5FICmok8I60=3D;
	b=Cwxi/Jp9bQR3gxwnGIjD2SBm7xTcocamQnVmaXilfWdPHchveDO48XpdWNtCKNy5g7Mcklc+
	QeZtKxYo7TZroi27UhOgz/Vt+y1XF2WobjyzUGNdUH/e2GMD0RfOkkBC;
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 73f7847c44628482de9d5f1018acf469
Cc: john.loughney@nokia.com,
	"'Lars Westberg \(KI/EAB\)'" <Lars.westberg@ericsson.com>, nsis@ietf.org,
	=?ISO-8859-1?Q? 'Attila_B=E1der_=28IJ/ETH=29' ?=
	<attila.bader@ericsson.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

I think some detailed discussion of this is going to be needed,  
because after considerable pondering the three of us could not at all  
figure out what either remapping or QSPEC stacking was good for. It  
certainly introduces a vast amount of complexity and confusion as  
currently formulated.

It appears at least to me that one of three situation obtains:
(a) we (well, at least I) don't understand RMD
(b) there is some general QoS concept that we are missing - we went  
through some complex scenarios like reservation aggregation any  
multiple QOSMs chained together and still could not figure out how  
one needed either stacking or remapping of QSPEC parameters
(c) RMD is really a completely different QoS signaling architecture  
from the rest of NSIS and is being shoehorned in by complicating  
QNSLP and QSPEC to allow multiple QoS signaling schemes to masquerade  
as a single architecture.

I really hope it's (a) or (b) since then we can probably resolve this  
by either putting back the missing functionality or figuring out  
another simple way to accommodate RMD. What I'm most afraid of is  
that this is really (c).

 From the little I understand RMD, it's really like mapping onto a  
custom L2 QoS scheme like ATM or DOCSIS, and that rather than doing  
the mapping the way any conventional L3 RMF would do it, you co- 
mingle RMD's model with NSIS's model.

But, I could of course be wildly wrong here.

Dave.


On Nov 2, 2006, at 6:00 AM, Georgios Karagiannis wrote:

> Hi John
>
> I have just come back from a trip and I have just look into point 4):
>>> 4. Since remapping is eliminated, the concept of stacking QSPECs is
>>> also eliminated.  Discussion of tunneling of QNSLP messages is
>>> eliminated since there are no impacts on tunneling required by  
>>> QSPEC.
>
> QSPEC stacking is one of the significant concepts in the QoS based  
> NSIS
> solution.
> This concept provides the additional flexibility that is missing  
> from RSVP.
> Furthermore, all the work that we have done until now is based on
> the concept of QSPEC stacking. Thus what you actually say is to just
> forget the work that we have done
> until now, in e.g.the RMD-QOSM draft, and start from the beginning.
>
>
> The main problem that we have had until now was not the QSPEC stacking
> issue,
> but it was the huge number of mandatory paremeters. Actualy I  
> consider the
> issue of removing the
> QSPEC stacking concept a much more severe problem on achieving the
> flexibility
> that we wanted to achieve with NSIS than the huge number of mandatory
> parameters.
>
> Therefore, I am severely against removing the QSPEC stacking concept!
>
> Best Regards,
> Georgios
>
>
>
>
>
>> -----Original Message-----
>> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
>> Sent: donderdag 2 november 2006 8:35
>> To: gash@att.com; nsis@ietf.org
>> Cc: cornelia.kappler@siemens.com; oran@cisco.com;
>> cornelia.kappler@gmx.net
>> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
>>
>> Jerry, Dave and Cornelia;
>>
>> This looks like a very good simplification of the QSPEC, and
>> I think that is a good idea.  I think that there will be some
>> implications to the QoSM drafts - I think QoS Model drafts
>> will need to contain some more applicability statement type
>> of text, along with deployment considerations.
>>
>> A plus side of this is that I think that the QoS template
>> will become more broadly applicable to other uses than just
>> NSIS QoS Signaling, so I think that is a gain as well.
>>
>> I'd really like people to review this and send comments on it
>> so that we could really focus the discussion next week and
>> get this work finally resolved.
>>
>> Thanks a lot for working through this, I think this is a good
>> basis for a proposal.
>>
>> thanks,
>> John
>>
>>> -----Original Message-----
>>> From: ext Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
>>> Sent: 02 November, 2006 00:13
>>> To: nsis@ietf.org
>>> Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran;
>>> Kappler,Cornelia
>>> Subject: [NSIS] Proposed Updates to the QSPEC Draft
>>>
>>> Hi All,
>>>
>>> Per John's suggestion below we've discussed Dave Oran's
>> suggestions at
>>> length and have the following proposal.
>>>
>>> We think the proposed changes address many concerns raised
>> on the list
>>> over the last few months and will require some rather basic
>> changes in
>>> the QSPEC document
>>> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
>>>
>>> We appreciate Dave's suggestions and careful review of the document.
>>>
>>> Here is a summary of the proposed changes:
>>>
>>> 1. The notion of QOSMs will be played down in the updated draft.
>>> Language to the effect that 'QNSLP/QSPEC can signal for
>> different QOSMs
>>> across multiple domains' will be replaced by the notion that
>>> QNSLP/QSPEC allows different elements on the path to implement
>>> different data plane QoS mechanisms that meets the
>> constraints included
>>> in the QSPEC.  A QOSM describes common capabilities among
>> QNEs to act
>>> consistently when requested to admit traffic and in treating
>> admitted
>>> traffic.
>>> A QOSM ID need not be defined or signaled and a QNE need not support
>>> any particular QOSM.  A 'QOSM specification' would become a
>>> 'QNSLP/QSPEC applicability statement', which documents how a QNE/RMF
>>> interprets and treats the various elements in a QSPEC.
>>>
>>> An updated QOSM definition is as follows:
>>>
>>> QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
>>> IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
>>> constraints and traffic handling directives a QNE implementing that
>>> QOSM is capable of supporting and how resources will be
>> managed by the
>>> RMF in such a QNE.
>>>
>>> 2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) -- 
>>> > d)
>>> semantics:
>>> A) The source traffic description is mandatory to include by the QNI
>>> and mandatory to interpret by downstream QNEs.  The traffic
>> description
>>> is specified by a traffic model (TMOD) parameter consisting of 4
>>> sub-parameters:
>>> - rate (r)
>>> - bucket size (b)
>>> - peak rate (p)
>>> - minimum policed unit (m)
>>> The TMOD parameter is a mathematically complete way to describe a
>>> traffic source.  Examples will be given for setting TMOD to
>> correspond
>>> to specific cases, e.g. where only bandwidth is available
>> (r=p; b and m
>>> set to default values) and other cases (e.g., TCP).  As such, the
>>> bandwidth parameter is eliminated.
>>> B) All other QSPEC parameters are optional to include by the
>> QNI, and
>>> may be either mandatory or optional to interpret by downstream QNEs.
>>> That is, the QNI can explicitly specify for each constraint
>> parameter
>>> whether it wishes admission control to succeed or fail if that
>>> constraint cannot be met.  A downstream QNE distinguishes
>> whether it is
>>> failure to understand the parameter or failure to meet the
>> constraint
>>> that causes the admission control failure (more detail on
>> this below).
>>> C) Additional QSPEC parameters can be defined in updates to
>> the QSPEC
>>> document (or alternatively specified in QOSM applicability statement
>>> documents).  There is already an IANA registry established for
>>> additional QSPEC parameters.
>>>
>>> 3. The concept of remapping QSPEC parameters is eliminated.
>>> This concept is ill defined and has caused much confusion (there are
>>> few if any sensible remappings).  Rather we define
>> 'interpret' a QSPEC
>>> parameter to only mean what we now call 'strictly interpret' (i.e.,
>>> must conform to the RFCs defining the parameter and procedures) and
>>> eliminate the remap option part of 'interpret'.
>>>
>>> 4. Since remapping is eliminated, the concept of stacking QSPECs is
>>> also eliminated.  Discussion of tunneling of QNSLP messages is
>>> eliminated since there are no impacts on tunneling required by  
>>> QSPEC.
>>>
>>> 5. Current QSPEC flags are modified as follows:
>>> A) The QNI sets the M flag for each QSPEC parameter it
>> populates that
>>> it intends must be interpreted or the reservation fails.
>> Currently the
>>> M flag is statically set for every QSPEC parameter.
>>> B) If the M flag is set, a downstream QNE must interpret the
>> parameter
>>> or the reservation fails.  If a QNE does not support the
>> parameter with
>>> the M flag set, it sets the N flag and rejects the
>> reservation.  If the
>>> QNE supports the parameter but cannot meet the parameter, it
>> sets the E
>>> flag and rejects the reservation.
>>> C) If the M flag is not set, a downstream QNE should interpret the
>>> parameter.  If a QNE does not support the parameter, it sets
>> the N flag
>>> and can optionally accept or reject the reservation.  If a
>> QNE supports
>>> the parameter but cannot meet the parameter, it sets the E flag and
>>> optional accepts or rejects the reservation.
>>> D) The R (remapped parameter) flag and Q (non QOSM) flag are
>>> eliminated.
>>>
>>> 6. Summary of proposed QSPEC parameters:
>>>
>>> Source traffic Description:
>>> TMOD-1 (mandatory to include)
>>> TMOD-2 (optional to include)
>>>
>>> Constraints (optional to include):
>>> Path Latency
>>> Path Jitter
>>> Path PLR
>>> Path PER
>>> Y.1541 class (shorthand notation for above 4 parameters)
>> Slack term
>>> Priority (Preemption, Defending, Admission, RPH
>>> Priority)  DSTE class
>>>
>>> Handling directives (optional to include):
>>> excess treatment
>>>
>>> Traffic Classifiers (optional to include):
>>> PHB Class (note that PHB class is set by downstream QNE to tell QNI
>>> how to mark traffic to ensure it will be treated as committed by
>>> admission
>>> control)
>>>
>>> Eliminated:
>>> bandwidth
>>> Ctot, Dtot, Csum, Dsum
>>>
>>> 7. Editorial changes:
>>> A) rewrite description of read-only and read-write
>> parameters to make
>>> the rules more clear:
>>> - which parameters can be ro/rw (function of object in which
>> parameter
>>> included)
>>> - can only increase rw values and not delete, etc.
>>> B) eliminate the distinction of 'control' and 'qos description'
>>> parameters and replace with suggested taxonomy, as described above
>>> C) clarify discussion of packet outcomes from Y.1540 and place in
>>> appropriate sections
>>> D) clarify which parameters can be included in <QoS Desired> <QoS
>>> Available>, <QoS Reserved>, or <QoS Minimum>
>>>
>>> Please review and comment on the proposed changes.
>>>
>>> Thanks,
>>> Jerry, Dave, Cornelia
>>>
>>>
>>> -----Original Message-----
>>> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
>>> Sent: Monday, October 23, 2006 2:59 AM
>>> To: nsis@ietf.org
>>> Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
>>>
>>> Hi all,
>>>
>>> We have had a very large thread on the current
>>> draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear
>> consensus,
>>> so I'd like to try to summarize the discussions.
>>>
>>> Several people feel that the current draft contains too many
>> mandatory
>>> parameters and mapping some QoS Models to others is mixing
>>> apples-to-oranges.  The feeling is that interoperability is hard to
>>> achieve and therefore we should not try to gain interoperability.
>>> Jukka Manner's message on 17th of October proposed a way forward.
>>> http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
>>>
>>> However, several (other) people feel that interoperability
>> is important
>>> and is needed. Giving up on interoperability would not be a
>> good thing,
>>> and that we are very close to a solution.
>>> Dave Oran proposed a mechanism to reach a solution on the 19th
>>> October:
>>> http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
>>>
>>> I have asked Jerry and Dave to make a proposal along the lines of
>>> Dave's message above, and the working group can review this,
>> to see if
>>> we can gain consensus on this.  I think it would be
>> sufficient to send
>>> some rough text to the mailing list, and if this is OK, then we can
>>> revise the draft.
>>>
>>> One open question to Dave/Jerry about the proposal is, in the event
>>> that the source cannot be described a mathematically
>> complete way, is
>>> there any value (or does it make any sense at
>>> all) in sending an incomplete qspec, or potentially "wildcarding"
>>> sending some of the mandatory parameters.  I'm wondering if some
>>> sources / deployments may have certain parameters that are
>> relatively
>>> well defined, but not all are so well defined.
>>>
>>> thanks,
>>> John
>>>
>>> _______________________________________________
>>> nsis mailing list
>>> nsis@ietf.org
>>> https://www1.ietf.org/mailman/listinfo/nsis
>>>
>>> _______________________________________________
>>> nsis mailing list
>>> nsis@ietf.org
>>> https://www1.ietf.org/mailman/listinfo/nsis
>>>
>>
>> _______________________________________________
>> nsis mailing list
>> nsis@ietf.org
>> https://www1.ietf.org/mailman/listinfo/nsis
>>
>
>
>
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 09:01:00 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfd7H-0003pj-Jf; Thu, 02 Nov 2006 09:00:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfd7G-0003oT-1i
	for nsis@ietf.org; Thu, 02 Nov 2006 09:00:22 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfd78-0005Nn-OC
	for nsis@ietf.org; Thu, 02 Nov 2006 09:00:22 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA2E07RF023992
	for <nsis@ietf.org>; Thu, 2 Nov 2006 15:00:11 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <nsis@ietf.org>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Thu, 2 Nov 2006 15:00:04 +0100
Message-ID: <004201c6fe87$358f1f50$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQAFNCDpAAAftaUA==
In-Reply-To: 
X-Spam-Score: 0.247 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Thu, 02 Nov 2006 15:00:12 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a1f9797ba297220533cb8c3f4bc709a8
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry, Dave and Cornelia

Here are some more detail comments on your proposal!

This proposal actually changes completely the QOS based concept of NSIS and
not only the QSPEC draft. Therefore, I do not agree with the proposal.

The problems that we have had until now, in the nsis mailing discussions,
were related to the huge number of mandatory parameters that all nodes in
the universe had to support. 
This proposal proposes only one parameter Token Bucket, which is good, but
it removes the possibility from NSIS to use QoS signaling for Diffserv
domains, which is against the NSIS charter specification.

Moreover, the remapping possibility from the peak bandwidth (Token Bucket)
value to the Bandwidth value that can be used in RMD-QOSM aware like nodes
is eliminated.
I do not understand why, since we have had a consensus on this point in the
nsis mailing list.
This proposal also removes the QSPEC stacking concept, which is one of the
most significant concepts in the QoS based NSIS solution.
This concept provides the additional flexibility that is missing from RSVP.
Furthermore, all the work that we have done until now is based on the
concept of QSPEC stacking. Thus what you actually say is to just forget the
work that we have done until now, in e.g.the RMD-QOSM draft, and start from
the beginning.

Furthermore, I wonder why Attila Bader has been left out of the discussions
that you have had to come up to this proposal.
>From what I know Jerry, Cornelia and Attila were the editors of the QSPEC
template draft.

More comments in line!


> -----Original Message-----
> From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
> Sent: woensdag 1 november 2006 23:13
> To: nsis@ietf.org
> Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran; 
> Kappler,Cornelia
> Subject: [NSIS] Proposed Updates to the QSPEC Draft
> 
> Hi All,
> 
> Per John's suggestion below we've discussed Dave Oran's suggestions at 
> length and have the following proposal.
> 
> We think the proposed changes address many concerns raised on the list 
> over the last few months and will require some rather basic changes in 
> the QSPEC document 
> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> 
> We appreciate Dave's suggestions and careful review of the document.
> 
> Here is a summary of the proposed changes:
> 
> 1. The notion of QOSMs will be played down in the updated draft.
> Language to the effect that 'QNSLP/QSPEC can signal for different 
> QOSMs across multiple domains' will be replaced by the notion that 
> QNSLP/QSPEC allows different elements on the path to implement 
> different data plane QoS mechanisms that meets the constraints 
> included in the QSPEC.  A QOSM describes common capabilities among 
> QNEs to act consistently when requested to admit traffic and in 
> treating admitted traffic.  A QOSM ID need not be defined or signaled 
> and a QNE need not support any particular QOSM.  A 'QOSM 
> specification'
> would become a 'QNSLP/QSPEC applicability statement', which documents 
> how a QNE/RMF interprets and treats the various elements in a QSPEC.


Georgios: This is a severe change in the NSIS QoS based concept, which we
have agreed
4/5 years ago to use in our work.
The current agreed concept compared to RSVP provides a flexibility advantage
that is missing from RSVP.
 


> An updated QOSM definition is as follows:
> 
> QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g., 
> IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS 
> constraints and traffic handling directives a QNE implementing that 
> QOSM is capable of supporting and how resources will be managed by the 
> RMF in such a QNE.
> 
> 2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
> semantics:
> A) The source traffic description is mandatory to include by the QNI 
> and mandatory to interpret by downstream QNEs.  The traffic 
> description is specified by a traffic model (TMOD) parameter 
> consisting of 4
> sub-parameters:
> - rate (r)
> - bucket size (b)
> - peak rate (p)
> - minimum policed unit (m)
> The TMOD parameter is a mathematically complete way to describe a 
> traffic source.  Examples will be given for setting TMOD to correspond 
> to specific cases, e.g. where only bandwidth is available (r=p; b and 
> m set to default values) and other cases (e.g., TCP).  As such, the 
> bandwidth parameter is eliminated.

Georgios: What you actually now say is that a Diffserv domain cannot use
NSIS for QoS signaling. As you probably know the NSIS charter specifies that
NSIS should provide QoS signaling for Diffserv domains.
This is actually against the NSIS charter specification.


> B) All other QSPEC parameters are optional to include by the QNI, and 
> may be either mandatory or optional to interpret by downstream QNEs.
> That is, the QNI can explicitly specify for each constraint parameter 
> whether it wishes admission control to succeed or fail if that 
> constraint cannot be met.  A downstream QNE distinguishes whether it 
> is failure to understand the parameter or failure to meet the 
> constraint that causes the admission control failure (more detail on 
> this below).

Georgios: In other words, RMD-QOSM cannot any more use NSIS for QoS
signaling ???



> C) Additional QSPEC parameters can be defined in updates to the QSPEC 
> document (or alternatively specified in QOSM applicability statement 
> documents).  There is already an IANA registry established for 
> additional QSPEC parameters.
> 
> 3. The concept of remapping QSPEC parameters is eliminated.  
> This concept is ill defined and has caused much confusion (there are 
> few if any sensible remappings).  Rather we define 'interpret' a QSPEC 
> parameter to only mean what we now call 'strictly interpret' (i.e., 
> must conform to the RFCs defining the parameter and procedures) and 
> eliminate the remap option part of 'interpret'.


Georgios: Again the peak bandwidth in the Token Bucket parameter can be
remapped to Bandwidth. Many people already agreed with this.
 Thus I do not see why you do not want to support this feature.
Probably, because in this way you will actually exclude the use and the
application of the RMD-QOSM work ???



> 4. Since remapping is eliminated, the concept of stacking QSPECs is 
> also eliminated.  Discussion of tunneling of QNSLP messages is 
> eliminated since there are no impacts on tunneling required by QSPEC.

Georgios: 

QSPEC stacking is one of the significant concepts in the QoS based NSIS
solution.
This concept provides the additional flexibility that is missing from RSVP.
Furthermore, all the work that we have done until now is based on the
concept of QSPEC stacking. Thus what you actually say is to just forget the
work that we have done until now, in e.g.the RMD-QOSM draft, and start from
the beginning.
The main problem that we have had until now was not the QSPEC stacking
issue, but it was the huge number of mandatory paremeters. Actualy I
consider the issue of removing the QSPEC stacking concept a much more severe
problem on achieving the flexibility that we wanted to achieve with NSIS
than the huge number of mandatory parameters.
> 
> 5. Current QSPEC flags are modified as follows:
> A) The QNI sets the M flag for each QSPEC parameter it populates that 
> it intends must be interpreted or the reservation fails.  Currently 
> the M flag is statically set for every QSPEC parameter.
> B) If the M flag is set, a downstream QNE must interpret the parameter 
> or the reservation fails.  If a QNE does not support the parameter 
> with the M flag set, it sets the N flag and rejects the reservation.  
> If the QNE supports the parameter but cannot meet the parameter, it 
> sets the E flag and rejects the reservation.
> C) If the M flag is not set, a downstream QNE should interpret the 
> parameter.  If a QNE does not support the parameter, it sets the N 
> flag and can optionally accept or reject the reservation.  If a QNE 
> supports the parameter but cannot meet the parameter, it sets the E 
> flag and optional accepts or rejects the reservation.
> D) The R (remapped parameter) flag and Q (non QOSM) flag are 
> eliminated.

Georgios: I do not agree to completely remove the remapping, see the reason
given on the previous points  and I also disagree to remove the R and Q
flags listed above!


 
> 6. Summary of proposed QSPEC parameters:
> 
> Source traffic Description:
>  TMOD-1 (mandatory to include)
>  TMOD-2 (optional to include)
> 
> Constraints (optional to include):
>  Path Latency
>  Path Jitter
>  Path PLR
>  Path PER
>  Y.1541 class (shorthand notation for above 4 parameters) Slack term  
> Priority (Preemption, Defending, Admission, RPH
> Priority)  DSTE class
> 
> Handling directives (optional to include):
>  excess treatment
> 
> Traffic Classifiers (optional to include):
>  PHB Class (note that PHB class is set by downstream QNE to tell QNI 
> how to mark traffic to ensure it will be treated as committed by 
> admission
> control)
> 
> Eliminated:
>  bandwidth
>  Ctot, Dtot, Csum, Dsum

Georgios: I do not agree, See above!

> 7. Editorial changes:
> A) rewrite description of read-only and read-write parameters to make 
> the rules more clear:
> - which parameters can be ro/rw (function of object in which parameter
> included)
> - can only increase rw values and not delete, etc.
> B) eliminate the distinction of 'control' and 'qos description'
> parameters and replace with suggested taxonomy, as described above
> C) clarify discussion of packet outcomes from Y.1540 and place in 
> appropriate sections
> D) clarify which parameters can be included in <QoS Desired> <QoS
> Available>, <QoS Reserved>, or <QoS Minimum>

Georgios: I do not agree, See above!


> Please review and comment on the proposed changes.
> Thanks,
> Jerry, Dave, Cornelia


Best Regards,
Georgios






_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 09:05:11 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfdBK-0000gs-MO; Thu, 02 Nov 2006 09:04:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfctE-0004mp-DG
	for nsis@ietf.org; Thu, 02 Nov 2006 08:45:52 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfcmc-0000WS-Go
	for nsis@ietf.org; Thu, 02 Nov 2006 08:39:06 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA2Dcelu023135; Thu, 2 Nov 2006 14:38:50 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>, <nsis@ietf.org>,
	"'Cornelia Kappler'" <cornelia.kappler@gmx.net>,
	"'David R Oran'" <oran@cisco.com>,
	"'Kappler,Cornelia'" <cornelia.kappler@siemens.com>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Thu, 2 Nov 2006 14:38:37 +0100
Message-ID: <002e01c6fe84$3919d410$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQAFNCDpA=
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
X-Spam-Score: 0.253 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Thu, 02 Nov 2006 14:38:51 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 162d87dc0b780d17da9b1934777fd451
X-Mailman-Approved-At: Thu, 02 Nov 2006 09:04:33 -0500
Cc: john.loughney@nokia.com, stiemerling@netlab.nec.de,
	"'Lars Westberg \(KI/EAB\)'" <Lars.westberg@ericsson.com>,
	=?iso-8859-1?Q?'Attila_B=E1der_=28IJ/ETH=29'?= <attila.bader@ericsson.com>,
	'Lars Eggert' <lars.eggert@netlab.nec.de>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry, Dave and Cornelia

Here are some more detail comments on your proposal!

This proposal actually changes completely the QOS based concept 
of NSIS and not only the QSPEC draft. Therefore, I do not agree 
with the proposal.

The problems that we have had until now, in the nsis mailing discussions,
were
related to the huge number of mandatory parameters that all nodes in the
universe had to support. 
This proposal proposes only one parameter Token Bucket, which is good, but 
it removes the possibility from NSIS to use QoS signaling for Diffserv
domains, 
which is against the NSIS charter specification.

Moreover, the remapping possibility from the peak bandwidth (Token Bucket)
value to 
the Bandwidth value that can be used in RMD-QOSM aware like nodes is
eliminated.
I do not understand why, since we have had a consensus on this point in the
nsis
mailing list.
This proposal also removes the QSPEC stacking concept, which 
is one of the most significant concepts in the QoS based NSIS solution.
This concept provides the additional flexibility that is missing from RSVP.
Furthermore, all the work that we have done until now is based on the 
concept of QSPEC stacking. Thus what you actually say is to just forget 
the work that we have done until now, in e.g.the RMD-QOSM draft, 
and start from the beginning.

Furthermore, I wonder why Attila Bader has been left out of the 
discussions that you have had to come up to this proposal.
>From what I know Jerry, Cornelia and Attila were the editors of the 
QSPEC template draft.

More comments in line!


> -----Original Message-----
> From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com] 
> Sent: woensdag 1 november 2006 23:13
> To: nsis@ietf.org
> Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R 
> Oran; Kappler,Cornelia
> Subject: [NSIS] Proposed Updates to the QSPEC Draft
> 
> Hi All,
> 
> Per John's suggestion below we've discussed Dave Oran's 
> suggestions at length and have the following proposal.  
> 
> We think the proposed changes address many concerns raised on 
> the list over the last few months and will require some 
> rather basic changes in the QSPEC document 
> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> 
> We appreciate Dave's suggestions and careful review of the document.
> 
> Here is a summary of the proposed changes:
> 
> 1. The notion of QOSMs will be played down in the updated draft.
> Language to the effect that 'QNSLP/QSPEC can signal for 
> different QOSMs across multiple domains' will be replaced by 
> the notion that QNSLP/QSPEC allows different elements on the 
> path to implement different data plane QoS mechanisms that 
> meets the constraints included in the QSPEC.  A QOSM 
> describes common capabilities among QNEs to act consistently 
> when requested to admit traffic and in treating admitted 
> traffic.  A QOSM ID need not be defined or signaled and a QNE 
> need not support any particular QOSM.  A 'QOSM specification' 
> would become a 'QNSLP/QSPEC applicability statement', which 
> documents how a QNE/RMF interprets and treats the various 
> elements in a QSPEC.  


Georgios: This is a severe change in the NSIS QoS based concept, 
which we have agreed 
4/5 years ago to use in our work.
The current agreed concept compared to RSVP provides a 
flexibility advantage that is missing from RSVP.
 


> An updated QOSM definition is as follows:
> 
> QoS Model (QOSM): A method to achieve QoS for a traffic flow, 
> e.g., IntServ Controlled Load.  A QOSM specifies what sub-set 
> of QSPEC QoS constraints and traffic handling directives a 
> QNE implementing that QOSM is capable of supporting and how 
> resources will be managed by the RMF in such a QNE.
> 
> 2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
> semantics:
> A) The source traffic description is mandatory to include by 
> the QNI and mandatory to interpret by downstream QNEs.  The 
> traffic description is specified by a traffic model (TMOD) 
> parameter consisting of 4
> sub-parameters:
> - rate (r)
> - bucket size (b)
> - peak rate (p)
> - minimum policed unit (m)
> The TMOD parameter is a mathematically complete way to 
> describe a traffic source.  Examples will be given for 
> setting TMOD to correspond to specific cases, e.g. where only 
> bandwidth is available (r=p; b and m set to default values) 
> and other cases (e.g., TCP).  As such, the bandwidth 
> parameter is eliminated.

Georgios: What you actually now say is that a Diffserv domain cannot 
use NSIS for QoS signaling. As you probably know the NSIS charter 
specifies that NSIS should provide QoS signaling for Diffserv domains.
This is actually against the NSIS charter specification.


> B) All other QSPEC parameters are optional to include by the 
> QNI, and may be either mandatory or optional to interpret by 
> downstream QNEs.
> That is, the QNI can explicitly specify for each constraint 
> parameter whether it wishes admission control to succeed or 
> fail if that constraint cannot be met.  A downstream QNE 
> distinguishes whether it is failure to understand the 
> parameter or failure to meet the constraint that causes the 
> admission control failure (more detail on this below).

Georgios: In other words, RMD-QOSM cannot any more use NSIS for QoS
signaling ???



> C) Additional QSPEC parameters can be defined in updates to 
> the QSPEC document (or alternatively specified in QOSM 
> applicability statement documents).  There is already an IANA 
> registry established for additional QSPEC parameters.
> 
> 3. The concept of remapping QSPEC parameters is eliminated.  
> This concept is ill defined and has caused much confusion 
> (there are few if any sensible remappings).  Rather we define 
> 'interpret' a QSPEC parameter to only mean what we now call 
> 'strictly interpret' (i.e., must conform to the RFCs defining 
> the parameter and procedures) and eliminate the remap option 
> part of 'interpret'.


Georgios: Again the peak bandwidth in the Token Bucket parameter can be 
remapped to Bandwidth. Many people already agreed with this.
 Thus I do not see why you do not want to support this feature.
Probably, because in this way you will actually exclude the use and the
application 
of the RMD-QOSM work ???



> 4. Since remapping is eliminated, the concept of stacking 
> QSPECs is also eliminated.  Discussion of tunneling of QNSLP 
> messages is eliminated since there are no impacts on 
> tunneling required by QSPEC.

Georgios: 

QSPEC stacking is one of the significant concepts in the QoS based NSIS
solution.
This concept provides the additional flexibility that is missing from RSVP.
Furthermore, all the work that we have done until now is based on the 
concept of QSPEC stacking. Thus what you actually say is to just forget 
the work that we have done until now, in e.g.the RMD-QOSM draft, 
and start from the beginning.
The main problem that we have had until now was not the QSPEC stacking
issue, 
but it was the huge number of mandatory paremeters. Actualy I consider the 
issue of removing the QSPEC stacking concept a much more severe problem on 
achieving the flexibility that we wanted to achieve with NSIS than the huge 
number of mandatory parameters.
> 
> 5. Current QSPEC flags are modified as follows:
> A) The QNI sets the M flag for each QSPEC parameter it 
> populates that it intends must be interpreted or the 
> reservation fails.  Currently the M flag is statically set 
> for every QSPEC parameter.
> B) If the M flag is set, a downstream QNE must interpret the 
> parameter or the reservation fails.  If a QNE does not 
> support the parameter with the M flag set, it sets the N flag 
> and rejects the reservation.  If the QNE supports the 
> parameter but cannot meet the parameter, it sets the E flag 
> and rejects the reservation. 
> C) If the M flag is not set, a downstream QNE should 
> interpret the parameter.  If a QNE does not support the 
> parameter, it sets the N flag and can optionally accept or 
> reject the reservation.  If a QNE supports the parameter but 
> cannot meet the parameter, it sets the E flag and optional 
> accepts or rejects the reservation.
> D) The R (remapped parameter) flag and Q (non QOSM) flag are 
> eliminated.

Georgios: I do not agree to completely remove the remapping, see the reason
given on the previous points
 and I also disagree to remove the R and Q flags listed above!


 
> 6. Summary of proposed QSPEC parameters:
> 
> Source traffic Description:
>  TMOD-1 (mandatory to include)
>  TMOD-2 (optional to include)
> 
> Constraints (optional to include):
>  Path Latency
>  Path Jitter
>  Path PLR
>  Path PER
>  Y.1541 class (shorthand notation for above 4 parameters)  
> Slack term  Priority (Preemption, Defending, Admission, RPH 
> Priority)  DSTE class
> 
> Handling directives (optional to include):
>  excess treatment
> 
> Traffic Classifiers (optional to include):
>  PHB Class (note that PHB class is set by downstream QNE to 
> tell QNI how to mark traffic to ensure it will be treated as 
> committed by admission
> control)
> 
> Eliminated:
>  bandwidth
>  Ctot, Dtot, Csum, Dsum

Georgios: I do not agree, See above!

> 7. Editorial changes:
> A) rewrite description of read-only and read-write parameters 
> to make the rules more clear: 
> - which parameters can be ro/rw (function of object in which parameter
> included)
> - can only increase rw values and not delete, etc.
> B) eliminate the distinction of 'control' and 'qos description'
> parameters and replace with suggested taxonomy, as described above
> C) clarify discussion of packet outcomes from Y.1540 and 
> place in appropriate sections
> D) clarify which parameters can be included in <QoS Desired> <QoS
> Available>, <QoS Reserved>, or <QoS Minimum>

Georgios: I do not agree, See above!


> Please review and comment on the proposed changes.
> Thanks,
> Jerry, Dave, Cornelia


Best Regards,
Georgios




_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 10:01:00 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfe3D-0001sU-Lw; Thu, 02 Nov 2006 10:00:15 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfe35-0001lX-RU
	for nsis@ietf.org; Thu, 02 Nov 2006 10:00:07 -0500
Received: from mail.gmx.de ([213.165.64.20] helo=mail.gmx.net)
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gfe0n-0007uy-BR
	for nsis@ietf.org; Thu, 02 Nov 2006 09:57:49 -0500
Received: (qmail invoked by alias); 02 Nov 2006 14:57:43 -0000
Received: from socks1.netz.sbs.de (EHLO [192.35.17.26]) [192.35.17.26]
	by mail.gmx.net (mp037) with SMTP; 02 Nov 2006 15:57:43 +0100
X-Authenticated: #29516787
Message-ID: <454A0764.3070202@gmx.net>
Date: Thu, 02 Nov 2006 15:57:40 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.7 (Windows/20060909)
MIME-Version: 1.0
To: NSIS <nsis@ietf.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 7a6398bf8aaeabc7a7bb696b6b0a2aad
Subject: [NSIS] QSPEC Face-to-Face Chat Discussion
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all,

I wonder how many NSIS folks are available for a QSPEC face-to-face 
discussion Sunday morning?

Since many folks have to be at company internal meetings during lunch 
time and want to see some of the tutorials I would suggest 9am at the
Sheraton San Diego Hotel (= meeting venue) in front of the hotel reception.

Is this a useful idea?

Ciao
Hannes

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 10:12:28 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfeEE-0007Kd-Ss; Thu, 02 Nov 2006 10:11:38 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfeED-0007KY-Vq
	for nsis@ietf.org; Thu, 02 Nov 2006 10:11:37 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfeE7-0002fY-Hc
	for nsis@ietf.org; Thu, 02 Nov 2006 10:11:37 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA2FBJQD026522
	for <nsis@ietf.org>; Thu, 2 Nov 2006 16:11:25 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <nsis@ietf.org>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Thu, 2 Nov 2006 16:11:16 +0100
Message-ID: <004601c6fe91$27c60640$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: Acb+i4KAPH1aSyGtQcCzUG6Udgc0mwAAX0ag
In-Reply-To: <76BE4729-5601-4708-8139-334E96D9E84E@cisco.com>
X-Spam-Score: 0.25 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Thu, 02 Nov 2006 16:11:27 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4cbeb0f20efb229aa93fae1468d20275
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Dave

You mention that remapping of the Token bucket to bandwidth is not =
allowed
but it is allowed=20
to interpret the peak bandwidth parameter value.

Do you mean that a QNE can use the peak bandwidth parameter and ignore=20
all the other Token Bucket parameters?

Regarding the signifficance of the QSPEC stacking please refer to =
Section
3.2.8 in:
http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt

Best Regards,
Georgios


> -----Original Message-----
> From: David R Oran [mailto:oran@cisco.com]=20
> Sent: donderdag 2 november 2006 15:21
> To: Georgios Karagiannis
> Cc: 'Ash, Gerald R (Jerry), ALABS'; nsis@ietf.org; 'Cornelia=20
> Kappler'; 'Kappler,Cornelia'; 'Attila B=E1der (IJ/ETH)' ; 'Lars=20
> Westberg (KI/EAB)'; 'Magnus Westerlund'; 'Lars Eggert';=20
> john.loughney@nokia.com; stiemerling@netlab.nec.de
> Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
>=20
>=20
> On Nov 2, 2006, at 8:38 AM, Georgios Karagiannis wrote:
>=20
> > Hi Jerry, Dave and Cornelia
> >
> > Here are some more detail comments on your proposal!
> >
> > This proposal actually changes completely the QOS based concept of=20
> > NSIS and not only the QSPEC draft. Therefore, I do not=20
> agree with the=20
> > proposal.
> >
> > The problems that we have had until now, in the nsis mailing=20
> > discussions, were related to the huge number of mandatory=20
> parameters=20
> > that all nodes in the universe had to support.
> > This proposal proposes only one parameter Token Bucket,=20
> which is good,=20
> > but it removes the possibility from NSIS to use QoS signaling for=20
> > Diffserv domains, which is against the NSIS charter specification.
> >
> How do you reach this conclusion? Could you give an example=20
> of what doesn't work for diffserv?
>=20
> > Moreover, the remapping possibility from the peak bandwidth (Token
> > Bucket)
> > value to
> > the Bandwidth value that can be used in RMD-QOSM aware like=20
> nodes is=20
> > eliminated.
> Yes, this was intentional. There is no need to remap in the=20
> NSIS signaling. A given QOSM's RMF can simply interpret the=20
> peak bandwidth parameter as their=20
> admitted/shaped/policed/queued bandwidth. This behavior would=20
> be documented in the QOSM application document for RMD.
>=20
> > I do not understand why, since we have had a consensus on=20
> this point=20
> > in the nsis mailing list.
> What seems to have happened (at least from my vantage point=20
> as a long- time lurker and occasional contributor) is that as=20
> we got down to reach closure on issues around the mandatory=20
> nature of QSPEC parameters and why certain QOSMs wanted=20
> certain parameters for certain purposes, we got to a=20
> deadlock. Removing the deadlock simply by blowing away any=20
> common interoperable behavior for NSIS (by declaring=20
> everything optional) would likely have run straight into a=20
> brick wall in the IESG. This prompted my, Jerry and Cornelia=20
> to take a step back, put our heads together, and try to=20
> figure out what the underlying problems were. Our assessment,=20
> captured in the proposal we just made, was that in order to=20
> maintain flexibility but still have global interoperability=20
> we needed to step back and rethink some of the more complex=20
> and confusing parts of the QSPEC design. We did that and you=20
> see the result. It's a matter of point of view whether one=20
> considers this "big change" or "valuable simplification".
>=20
> > This proposal also removes the QSPEC stacking concept,=20
> which is one of=20
> > the most significant concepts in the QoS based NSIS solution.
> So, we clearly missed the significance, so perhaps you need=20
> to go back and re-educate us. No other QOSM I know of other=20
> than RMD claims to need this functionality.
>=20
> > This concept provides the additional flexibility that is=20
> missing from=20
> > RSVP.
> > Furthermore, all the work that we have done until now is=20
> based on the=20
> > concept of QSPEC stacking. Thus what you actually say is to just=20
> > forget the work that we have done until now, in e.g.the RMD-QOSM=20
> > draft, and start from the beginning.
> >
> Not necessarily. What we may need to do is rethink a bit how=20
> RMD integrates with NSIS as a QOSM. the other possibility is=20
> for the WG to reject our approach and stay with what's in the=20
> QPSEC draft today.
>=20
> > Furthermore, I wonder why Attila Bader has been left out of the=20
> > discussions that you have had to come up to this proposal.
> > From what I know Jerry, Cornelia and Attila were the editors of the=20
> > QSPEC template draft.
> >
> > More comments in line!
> >
> >
> >> -----Original Message-----
> >> From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
> >> Sent: woensdag 1 november 2006 23:13
> >> To: nsis@ietf.org
> >> Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran;=20
> >> Kappler,Cornelia
> >> Subject: [NSIS] Proposed Updates to the QSPEC Draft
> >>
> >> Hi All,
> >>
> >> Per John's suggestion below we've discussed Dave Oran's=20
> suggestions=20
> >> at length and have the following proposal.
> >>
> >> We think the proposed changes address many concerns raised on the=20
> >> list over the last few months and will require some rather basic=20
> >> changes in the QSPEC document=20
> >> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> >>
> >> We appreciate Dave's suggestions and careful review of the=20
> document.
> >>
> >> Here is a summary of the proposed changes:
> >>
> >> 1. The notion of QOSMs will be played down in the updated draft.
> >> Language to the effect that 'QNSLP/QSPEC can signal for different=20
> >> QOSMs across multiple domains' will be replaced by the notion that=20
> >> QNSLP/QSPEC allows different elements on the path to implement=20
> >> different data plane QoS mechanisms that meets the constraints=20
> >> included in the QSPEC.  A QOSM describes common capabilities among=20
> >> QNEs to act consistently when requested to admit traffic and in=20
> >> treating admitted traffic.  A QOSM ID need not be defined=20
> or signaled=20
> >> and a QNE need not support any particular QOSM.  A 'QOSM=20
> >> specification'
> >> would become a 'QNSLP/QSPEC applicability statement',=20
> which documents=20
> >> how a QNE/RMF interprets and treats the various elements=20
> in a QSPEC.
> >
> >
> > Georgios: This is a severe change in the NSIS QoS based=20
> concept, which=20
> > we have agreed
> > 4/5 years ago to use in our work.
> > The current agreed concept compared to RSVP provides a flexibility=20
> > advantage that is missing from RSVP.
> >
> It is a change/simplification. What I don't get (I won't=20
> speak for Jerry or Cornelia, but neither of them could come=20
> up with a succinct explanation when we discussed this) is=20
> exactly what this flexibility is and why it matters.
>=20
> >
> >
> >> An updated QOSM definition is as follows:
> >>
> >> QoS Model (QOSM): A method to achieve QoS for a traffic=20
> flow, e.g.,=20
> >> IntServ Controlled Load.  A QOSM specifies what sub-set of=20
> QSPEC QoS=20
> >> constraints and traffic handling directives a QNE=20
> implementing that=20
> >> QOSM is capable of supporting and how resources will be managed by=20
> >> the RMF in such a QNE.
> >>
> >> 2. The QSPEC1/QSPEC2 semantics will be replaced with=20
> Dave's a) --> d)
> >> semantics:
> >> A) The source traffic description is mandatory to include=20
> by the QNI=20
> >> and mandatory to interpret by downstream QNEs.  The traffic=20
> >> description is specified by a traffic model (TMOD) parameter=20
> >> consisting of 4
> >> sub-parameters:
> >> - rate (r)
> >> - bucket size (b)
> >> - peak rate (p)
> >> - minimum policed unit (m)
> >> The TMOD parameter is a mathematically complete way to describe a=20
> >> traffic source.  Examples will be given for setting TMOD to=20
> >> correspond to specific cases, e.g. where only bandwidth is=20
> available=20
> >> (r=3Dp; b and m set to default values) and other cases=20
> (e.g., TCP).  As=20
> >> such, the bandwidth parameter is eliminated.
> >
> > Georgios: What you actually now say is that a Diffserv=20
> domain cannot=20
> > use NSIS for QoS signaling. As you probably know the NSIS charter=20
> > specifies that NSIS should provide QoS signaling for=20
> Diffserv domains.
> > This is actually against the NSIS charter specification.
> >
> I'm lost. what leads you to this conclusion?
>=20
> >
> >> B) All other QSPEC parameters are optional to include by=20
> the QNI, and=20
> >> may be either mandatory or optional to interpret by=20
> downstream QNEs.
> >> That is, the QNI can explicitly specify for each=20
> constraint parameter=20
> >> whether it wishes admission control to succeed or fail if that=20
> >> constraint cannot be met.  A downstream QNE distinguishes=20
> whether it=20
> >> is failure to understand the parameter or failure to meet the=20
> >> constraint that causes the admission control failure (more=20
> detail on=20
> >> this below).
> >
> > Georgios: In other words, RMD-QOSM cannot any more use NSIS for QoS=20
> > signaling ???
> >
> >
> Possibly, since I'm not sure the goal of NSIS was ever to be=20
> just a schematic shell for inserting individual QOSM-specific=20
> signaling protocols. If it was, then clearly my understanding=20
> of the NSIS architecture and what I read in the NSIS charter=20
> does not correspond to the WG consensus our changes to the=20
> QSPEC draft would be ill-advised.
>=20
> >
> >> C) Additional QSPEC parameters can be defined in updates=20
> to the QSPEC=20
> >> document (or alternatively specified in QOSM applicability=20
> statement=20
> >> documents).  There is already an IANA registry established for=20
> >> additional QSPEC parameters.
> >>
> >> 3. The concept of remapping QSPEC parameters is eliminated.
> >> This concept is ill defined and has caused much confusion=20
> (there are=20
> >> few if any sensible remappings).  Rather we define 'interpret' a=20
> >> QSPEC parameter to only mean what we now call 'strictly interpret'=20
> >> (i.e., must conform to the RFCs defining the parameter and=20
> >> procedures) and eliminate the remap option part of 'interpret'.
> >
> >
> > Georgios: Again the peak bandwidth in the Token Bucket=20
> parameter can=20
> > be remapped to Bandwidth. Many people already agreed with this.
> If you mean "interpreted by the RMD QOSM as bandwidth" then=20
> we are in agreement. If you mean "modified in the NSLP=20
> signaling and QSPEC by replacing the <peak bandwidth=20
> parameter> with a <bandwidth> parameter, possibly stacking=20
> the new QPEC with <bandwidth> instead of <peak bandwidth>"=20
> then that's where we disagreed with the current approach in=20
> fact departed from the current spec.
>=20
> The problem with both remapping and stacking is that they do=20
> not constrain the semantics of the end-to-end signaling, and=20
> introduce considerable complexity into the protocol.
>=20
> >  Thus I do not see why you do not want to support this feature.
> > Probably, because in this way you will actually exclude the use and=20
> > the application of the RMD-QOSM work ???
> >
> >
> >
> >> 4. Since remapping is eliminated, the concept of stacking=20
> QSPECs is=20
> >> also eliminated.  Discussion of tunneling of QNSLP messages is=20
> >> eliminated since there are no impacts on tunneling=20
> required by QSPEC.
> >
> > Georgios:
> >
> > QSPEC stacking is one of the significant concepts in the QoS based=20
> > NSIS solution.
> > This concept provides the additional flexibility that is=20
> missing from=20
> > RSVP.
> > Furthermore, all the work that we have done until now is=20
> based on the=20
> > concept of QSPEC stacking. Thus what you actually say is to just=20
> > forget the work that we have done until now, in e.g.the RMD-QOSM=20
> > draft, and start from the beginning.
> > The main problem that we have had until now was not the=20
> QSPEC stacking=20
> > issue, but it was the huge number of mandatory paremeters.=20
> Actualy I=20
> > consider the issue of removing the QSPEC stacking concept a=20
> much more=20
> > severe problem on achieving the flexibility that we wanted=20
> to achieve=20
> > with NSIS than the huge number of mandatory parameters.
> >>
> >> 5. Current QSPEC flags are modified as follows:
> >> A) The QNI sets the M flag for each QSPEC parameter it=20
> populates that=20
> >> it intends must be interpreted or the reservation fails. =20
> Currently=20
> >> the M flag is statically set for every QSPEC parameter.
> >> B) If the M flag is set, a downstream QNE must interpret the=20
> >> parameter or the reservation fails.  If a QNE does not support the=20
> >> parameter with the M flag set, it sets the N flag and rejects the=20
> >> reservation.  If the QNE supports the parameter but cannot=20
> meet the=20
> >> parameter, it sets the E flag and rejects the reservation.
> >> C) If the M flag is not set, a downstream QNE should interpret the=20
> >> parameter.  If a QNE does not support the parameter, it sets the N=20
> >> flag and can optionally accept or reject the reservation. =20
> If a QNE=20
> >> supports the parameter but cannot meet the parameter, it=20
> sets the E=20
> >> flag and optional accepts or rejects the reservation.
> >> D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
> >> eliminated.
> >
> > Georgios: I do not agree to completely remove the=20
> remapping, see the=20
> > reason given on the previous points  and I also disagree to=20
> remove the=20
> > R and Q flags listed above!
> >
> >
> >
> >> 6. Summary of proposed QSPEC parameters:
> >>
> >> Source traffic Description:
> >>  TMOD-1 (mandatory to include)
> >>  TMOD-2 (optional to include)
> >>
> >> Constraints (optional to include):
> >>  Path Latency
> >>  Path Jitter
> >>  Path PLR
> >>  Path PER
> >>  Y.1541 class (shorthand notation for above 4 parameters)
> >> Slack term  Priority (Preemption, Defending, Admission, RPH
> >> Priority)  DSTE class
> >>
> >> Handling directives (optional to include):
> >>  excess treatment
> >>
> >> Traffic Classifiers (optional to include):
> >>  PHB Class (note that PHB class is set by downstream QNE to
> >> tell QNI how to mark traffic to ensure it will be treated as
> >> committed by admission
> >> control)
> >>
> >> Eliminated:
> >>  bandwidth
> >>  Ctot, Dtot, Csum, Dsum
> >
> > Georgios: I do not agree, See above!
> >
> >> 7. Editorial changes:
> >> A) rewrite description of read-only and read-write parameters
> >> to make the rules more clear:
> >> - which parameters can be ro/rw (function of object in which =20
> >> parameter
> >> included)
> >> - can only increase rw values and not delete, etc.
> >> B) eliminate the distinction of 'control' and 'qos description'
> >> parameters and replace with suggested taxonomy, as described above
> >> C) clarify discussion of packet outcomes from Y.1540 and
> >> place in appropriate sections
> >> D) clarify which parameters can be included in <QoS Desired> <QoS
> >> Available>, <QoS Reserved>, or <QoS Minimum>
> >
> > Georgios: I do not agree, See above!
> >
> >
> >> Please review and comment on the proposed changes.
> >> Thanks,
> >> Jerry, Dave, Cornelia
> >
> >
> > Best Regards,
> > Georgios
> >
>=20



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 10:32:30 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfeXG-00060p-Eq; Thu, 02 Nov 2006 10:31:18 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfeXF-0005zz-Fe
	for nsis@ietf.org; Thu, 02 Nov 2006 10:31:17 -0500
Received: from sj-iport-3-in.cisco.com ([171.71.176.72]
	helo=sj-iport-3.cisco.com) by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1GfeXB-0005q4-4a
	for nsis@ietf.org; Thu, 02 Nov 2006 10:31:17 -0500
Received: from sj-dkim-3.cisco.com ([171.71.179.195])
	by sj-iport-3.cisco.com with ESMTP; 02 Nov 2006 07:31:12 -0800
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: AgAAANhNSUWrR7PDh2dsb2JhbACMSgEBCQ4q
X-IronPort-AV: i="4.09,381,1157353200"; 
	d="scan'208"; a="447750968:sNHT30835264"
Received: from sj-core-2.cisco.com (sj-core-2.cisco.com [171.71.177.254])
	by sj-dkim-3.cisco.com (8.12.11.20060308/8.12.11) with ESMTP id
	kA2FVCpF032543; Thu, 2 Nov 2006 07:31:12 -0800
Received: from imail.cisco.com (sjc12-sbr-sw3-3f5.cisco.com [172.19.96.182])
	by sj-core-2.cisco.com (8.12.10/8.12.6) with ESMTP id kA2FVCin002916;
	Thu, 2 Nov 2006 07:31:12 -0800 (PST)
Received: from [10.32.245.152] (stealth-10-32-245-152.cisco.com
	[10.32.245.152])
	by imail.cisco.com (8.12.11/8.12.10) with ESMTP id kA2FIGUa013122;
	Thu, 2 Nov 2006 07:18:16 -0800
In-Reply-To: <454A0764.3070202@gmx.net>
References: <454A0764.3070202@gmx.net>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <C47B85AD-3740-4891-B23B-0D79FF1D6A3A@cisco.com>
Content-Transfer-Encoding: 7bit
From: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] QSPEC Face-to-Face Chat Discussion
Date: Thu, 2 Nov 2006 10:31:07 -0500
To: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
X-Mailer: Apple Mail (2.752.2)
Authentication-Results: sj-dkim-3.cisco.com; header.From=oran@cisco.com;
	dkim=pass ( sig from cisco.com verified; ); 
DKIM-Signature: a=rsa-sha1; q=dns; l=751; t=1162481472; x=1163345472;
	c=relaxed/simple; s=sjdkim3002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=oran@cisco.com;
	z=From:David=20R=20Oran=20<oran@cisco.com>
	|Subject:Re=3A=20[NSIS]=20QSPEC=20Face-to-Face=20Chat=20Discussion;
	X=v=3Dcisco.com=3B=20h=3DP4R2c9TCsl4TNvmRC3agWMPDnok=3D;
	b=P46wNT0XZ44gsgA0+/GARTry4VqxMDr77iIVOrpA8TSyST+mGdhQXeASBbIspuBaPeJFUc3M
	jlVNMl7wiHOiteMn1jxGyeNd1evBWXv/DTPEpiTM5hNM5XGidPVm/bsI;
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2409bba43e9c8d580670fda8b695204a
Cc: NSIS <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


On Nov 2, 2006, at 9:57 AM, Hannes Tschofenig wrote:

> Hi all,
>
> I wonder how many NSIS folks are available for a QSPEC face-to-face  
> discussion Sunday morning?
>
Not I - I have the IAB meeting followed by a joint IAB/IESG meeting.  
I get free at around 2:30PM.

> Since many folks have to be at company internal meetings during  
> lunch time and want to see some of the tutorials I would suggest  
> 9am at the
> Sheraton San Diego Hotel (= meeting venue) in front of the hotel  
> reception.
>
> Is this a useful idea?
>
Yes, but the timing is a bit problematic.
> Ciao
> Hannes
>
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 10:39:50 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfeen-0004S3-Hm; Thu, 02 Nov 2006 10:39:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfeem-0004Kx-Lw
	for nsis@ietf.org; Thu, 02 Nov 2006 10:39:04 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfeei-0006mL-AD
	for nsis@ietf.org; Thu, 02 Nov 2006 10:39:04 -0500
Received: from sbz-30.cs.helsinki.fi (sbz-30.cs.helsinki.fi [128.214.9.98])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Thu, 02 Nov 2006 17:38:59 +0200
	id 000AFD4F.454A1113.00002A42
Date: Thu, 2 Nov 2006 17:38:59 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
Subject: Re: [NSIS] QSPEC Face-to-Face Chat Discussion
In-Reply-To: <454A0764.3070202@gmx.net>
Message-ID: <Pine.LNX.4.64.0611021738210.11312@sbz-30.cs.Helsinki.FI>
References: <454A0764.3070202@gmx.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ea4ac80f790299f943f0a53be7e1a21a
Cc: NSIS <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi,

Yes, it is a good idea. I don't have anything particular on Sunday, so I 
can attend almost anytime.

Jukka

On Thu, 2 Nov 2006, Hannes Tschofenig wrote:

> Hi all,
> 
> I wonder how many NSIS folks are available for a QSPEC face-to-face
> discussion Sunday morning?
> 
> Since many folks have to be at company internal meetings during lunch time
> and want to see some of the tutorials I would suggest 9am at the
> Sheraton San Diego Hotel (= meeting venue) in front of the hotel reception.
> 
> Is this a useful idea?
> 
> Ciao
> Hannes
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 10:51:48 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfeqQ-0006mX-KU; Thu, 02 Nov 2006 10:51:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfeqP-0006jq-20
	for nsis@ietf.org; Thu, 02 Nov 2006 10:51:05 -0500
Received: from mail.gmx.net ([213.165.64.20])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfeqL-0007zC-Bm
	for nsis@ietf.org; Thu, 02 Nov 2006 10:51:05 -0500
Received: (qmail invoked by alias); 02 Nov 2006 15:51:00 -0000
Received: from socks1.netz.sbs.de (EHLO [192.35.17.26]) [192.35.17.26]
	by mail.gmx.net (mp016) with SMTP; 02 Nov 2006 16:51:00 +0100
X-Authenticated: #29516787
Message-ID: <454A13DF.2010104@gmx.net>
Date: Thu, 02 Nov 2006 16:50:55 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.7 (Windows/20060909)
MIME-Version: 1.0
To: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
Subject: Re: [NSIS] QSPEC Face-to-Face Chat Discussion
References: <454A0764.3070202@gmx.net>
	<Pine.LNX.4.64.0611021738210.11312@sbz-30.cs.Helsinki.FI>
In-Reply-To: <Pine.LNX.4.64.0611021738210.11312@sbz-30.cs.Helsinki.FI>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 39bd8f8cbb76cae18b7e23f7cf6b2b9f
Cc: NSIS <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

We should try to organize something on Sunday.
I will also give the QSPEC discussion another shot.

Ciao
hannes

Jukka MJ Manner wrote:
> Hi,
> 
> Yes, it is a good idea. I don't have anything particular on Sunday, so I 
> can attend almost anytime.
> 
> Jukka
> 
> On Thu, 2 Nov 2006, Hannes Tschofenig wrote:
> 
>> Hi all,
>>
>> I wonder how many NSIS folks are available for a QSPEC face-to-face
>> discussion Sunday morning?
>>
>> Since many folks have to be at company internal meetings during lunch time
>> and want to see some of the tutorials I would suggest 9am at the
>> Sheraton San Diego Hotel (= meeting venue) in front of the hotel reception.
>>
>> Is this a useful idea?
>>
>> Ciao
>> Hannes
>>
>> _______________________________________________
>> nsis mailing list
>> nsis@ietf.org
>> https://www1.ietf.org/mailman/listinfo/nsis
>>
>>


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 11:22:02 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GffJV-00037p-EY; Thu, 02 Nov 2006 11:21:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GffJT-00035P-EP
	for nsis@ietf.org; Thu, 02 Nov 2006 11:21:07 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GffJN-0002U8-66
	for nsis@ietf.org; Thu, 02 Nov 2006 11:21:07 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-8.tower-146.messagelabs.com!1162484460!7825275!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 1458 invoked from network); 2 Nov 2006 16:21:00 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-8.tower-146.messagelabs.com with SMTP;
	2 Nov 2006 16:21:00 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA2GKxLg005003; 
	Thu, 2 Nov 2006 11:21:00 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA2GKQgE004704; Thu, 2 Nov 2006 11:20:53 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] QSPEC Face-to-Face Chat Discussion
Date: Thu, 2 Nov 2006 10:20:51 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F14@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <C47B85AD-3740-4891-B23B-0D79FF1D6A3A@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] QSPEC Face-to-Face Chat Discussion
Thread-Index: Acb+lEDHPtGDTuI+Sm6KTCLRO5mzaQABj44g
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "David R Oran" <oran@cisco.com>,
	"Hannes Tschofenig" <Hannes.Tschofenig@gmx.net>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bb8f917bb6b8da28fc948aeffb74aa17
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>, NSIS <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

It is a useful idea.  Late afternoon would be OK for me (after 3 PM),
I'm not available in the morning.=20

Thanks,
Jerry

-----Original Message-----
From: David R Oran [mailto:oran@cisco.com]=20
Sent: Thursday, November 02, 2006 10:31 AM
To: Hannes Tschofenig
Cc: NSIS
Subject: Re: [NSIS] QSPEC Face-to-Face Chat Discussion


On Nov 2, 2006, at 9:57 AM, Hannes Tschofenig wrote:

> Hi all,
>
> I wonder how many NSIS folks are available for a QSPEC face-to-face =20
> discussion Sunday morning?
>
Not I - I have the IAB meeting followed by a joint IAB/IESG meeting. =20
I get free at around 2:30PM.

> Since many folks have to be at company internal meetings during =20
> lunch time and want to see some of the tutorials I would suggest =20
> 9am at the
> Sheraton San Diego Hotel (=3D meeting venue) in front of the hotel =20
> reception.
>
> Is this a useful idea?
>
Yes, but the timing is a bit problematic.
> Ciao
> Hannes

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 02 12:18:17 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfgBf-0007rP-Tg; Thu, 02 Nov 2006 12:17:07 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfgBe-0007qN-NS
	for nsis@ietf.org; Thu, 02 Nov 2006 12:17:06 -0500
Received: from rsys002x.roke.co.uk ([193.118.201.109])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfg9G-0008Al-60
	for nsis@ietf.org; Thu, 02 Nov 2006 12:15:29 -0500
Received: from rsys005a.comm.ad.roke.co.uk (rsys005a [193.118.193.85])
	by rsys002x.roke.co.uk (8.13.1/8.13.1) with ESMTP id kA2HERxo005959;
	Thu, 2 Nov 2006 17:14:28 GMT
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] question about GIST-Query message
Date: Thu, 2 Nov 2006 17:14:27 -0000
Message-ID: <A632AD91CF90F24A87C42F6B96ADE5C50157A988@rsys005a.comm.ad.roke.co.uk>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] question about GIST-Query message
Thread-Index: Acb+omZILDVvgDnnSW63gZQHRmHTGQ==
From: "Hancock, Robert" <robert.hancock@roke.co.uk>
To: =?iso-8859-1?B?sfOx8831?= <binbinwang118@gmail.com>, <nsis@ietf.org>
X-MailScanner-roke-co-uk: Found to be clean
X-MailScanner-roke-co-uk-SpamCheck: 
X-MailScanner-From: robert.hancock@roke.co.uk
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7268a2980febc47a9fa732aba2b737ba
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1849035636=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1849035636==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C6FEA2.5A304C86"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C6FEA2.5A304C86
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hi Wang,
=20
Let's assume you are talking about signalling for a data flow from A to =
D; this then uses the path-coupled MRM within GIST.
=20
The GIST-Query is then an IP datagram with the destination address D, =
and the source address A. Which path the datagram takes through the =
network depends on the normal network routing of IP datagrams with these =
addresses - GIST doesn't change this routing. In other words, if a =
normal datagram with a destination address of D is routed A>>C>>D, then =
the signalling will follow the same route (and these will be the nodes =
that create signalling state).
=20
Does this answer your question?
=20
yours,
=20
Robert H.

	-----Original Message-----
	From: =B1=F3=B1=F3=CD=F5 [mailto:binbinwang118@gmail.com]=20
	Sent: 02 November 2006 07:29
	To: nsis@ietf.org
	Subject: [NSIS] question about GIST-Query message
=09
=09
	Hi all,
	=20
	I have a question about GIST-Query message. Whether the GIST-Query =
message is group sending in networking?
	                         +--+            =20
	                          |B |
	                        .+--+.            =20
	                       .        .      =20
	                  +-+          . +--+
	                  |A |.            | D|=20
	                  +-+.          .+--+
	                        .        .     =20
	                         .+--+.       =20
	                          |C  |
	                          +--+=20
	Assume node A, node B, node C, and node D, they all support the same =
NSLP. Node A sends a GIST-Query message, namely a UDP datagram addressed =
to the node D and include an IP RAO. whether both node B and node C will =
receive the Query message? And which node that node A will establishes =
GIST state with it? Finally the signaling path is A >> B >> D, or A >> C =
>> D ?=20
	=20
	Bestwishes!
	=20
	Wang
	2006-11-2
	=20

	--=20
	skyWang.bupt=20


------_=_NextPart_001_01C6FEA2.5A304C86
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD><TITLE>Message</TITLE>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2900.2963" name=3DGENERATOR></HEAD>
<BODY>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =
size=3D2>Hi=20
Wang,</FONT></SPAN></DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =
size=3D2>Let's=20
assume you are talking about signalling for a data flow from A to D; =
this then=20
uses the path-coupled MRM within GIST.</FONT></SPAN></DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =
size=3D2>The=20
GIST-Query is then an IP datagram with the destination address D, and =
the source=20
address A. Which path the datagram takes through the network depends on =
the=20
normal network routing of IP datagrams with these addresses - GIST =
doesn't=20
change this routing. In other words, if a normal datagram with a =
destination=20
address of D is routed A&gt;&gt;C&gt;&gt;D, then the signalling will =
follow the=20
same route (and these will be the nodes that create signalling=20
state).</FONT></SPAN></DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =
size=3D2>Does=20
this answer your question?</FONT></SPAN></DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =

size=3D2>yours,</FONT></SPAN></DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =

size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV><SPAN class=3D361181017-02112006><FONT face=3DArial color=3D#0000ff =
size=3D2>Robert=20
H.</FONT></SPAN></DIV>
<BLOCKQUOTE dir=3Dltr=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
  <DIV></DIV>
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr =
align=3Dleft><FONT=20
  face=3DTahoma size=3D2>-----Original Message-----<BR><B>From:</B> =
=B1=F3=B1=F3=CD=F5=20
  [mailto:binbinwang118@gmail.com] <BR><B>Sent:</B> 02 November 2006=20
  07:29<BR><B>To:</B> nsis@ietf.org<BR><B>Subject:</B> [NSIS] question =
about=20
  GIST-Query message<BR><BR></FONT></DIV>
  <DIV>Hi all,</DIV>
  <DIV>&nbsp;</DIV>
  <DIV>I have a question about GIST-Query message. Whether =
the&nbsp;GIST-Query=20
  message is group sending in networking?</DIV>
  =
<DIV>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;=20
  =
+--+&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;=20
  =
<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  &nbsp; |B=20
  =
|<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  =
.+--+.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&=
nbsp;=20
  =
<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  .&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp; =
.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  =
<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  +-+&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; .=20
  =
+--+<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  |A =
|.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; | =
D|=20
  =
<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  +-+.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  =
.+--+<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  .&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  =
<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;=20
  .+--+.&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;=20
  =
<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs=
p;&nbsp;=20
  |C=20
  =
&nbsp;|<BR>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n=
bsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb=
sp;&nbsp;&nbsp;=20
  +--+&nbsp;</DIV>
  <DIV>Assume node A, node B, node C, and node D, they all support the =
same=20
  NSLP. Node A sends a GIST-Query message, namely a UDP datagram =
addressed to=20
  the node D and include an IP RAO. whether both node B and node C will =
receive=20
  the Query message? And which node that node A will establishes GIST =
state with=20
  it? Finally the signaling path is A &gt;&gt; B &gt;&gt; D, or A =
&gt;&gt; C=20
  &gt;&gt; D ? </DIV>
  <DIV>&nbsp;</DIV>
  <DIV>Bestwishes!</DIV>
  <DIV>&nbsp;</DIV>
  <DIV>Wang</DIV>
  <DIV>2006-11-2</DIV>
  <DIV>&nbsp;</DIV>
  <DIV> </DIV>
  <DIV><BR>-- <BR>skyWang.bupt </DIV></BLOCKQUOTE></BODY></HTML>

------_=_NextPart_001_01C6FEA2.5A304C86--


--===============1849035636==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1849035636==--




From nsis-bounces@ietf.org Fri Nov 03 01:28:04 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfsWO-0000cZ-AS; Fri, 03 Nov 2006 01:27:20 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfsWM-0000cP-Az
	for nsis@ietf.org; Fri, 03 Nov 2006 01:27:18 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfsWH-00005r-Qc
	for nsis@ietf.org; Fri, 03 Nov 2006 01:27:18 -0500
Received: from webmail.cs.utwente.nl (janus.ewi.utwente.nl [130.89.10.26])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with SMTP id kA36R1a1025209
	for <nsis@ietf.org>; Fri, 3 Nov 2006 07:27:06 +0100 (MET)
Received: from 84.82.109.231 (auth. user karagian@imap2.cs.utwente.nl)
	by webmail.cs.utwente.nl with HTTP; Fri, 03 Nov 2006 06:27:01 +0000
To: nsis@ietf.org
Date: Fri, 03 Nov 2006 06:27:01 +0000
X-Mailer: IlohaMail/0.8.13 (On: webmail.cs.utwente.nl)
Message-ID: <hvOvd46k.1162535221.3187200.karagian@ewi.utwente.nl>
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
Bounce-To: "Georgios Karagiannis" <karagian@cs.utwente.nl>
MIME-Version: 1.0 
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: 0 () 
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 07:27:06 +0100 (MET)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 8b431ad66d60be2d47c7bfeb879db82c
Subject: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all

After rereading the proposal of Jerry, Dave, Cornelia I have seen that the
main problem that we have had until now in the mailing discussions is not
solved. Mainly the number of mandatory parameters to interpret by all
the QNEs in the universe is not solved.

The proposal mentions in point 2) point B):

"All other QSPEC parameters are optional to include by the QNI, and may
be either mandatory or optional to interpret by downstream QNEs."

Meaning that if a QNI includes the Y.1541 class and wants that all
downstream QNEs interpret this parameter, then all downstream QNEs must
be able to strictly interpret the Y.1541 class.
Jerry, Dave, Cornelia is this true?

Meaning that all QNEs that want to use NSIS must be implemented in such a
way that if a QNI mandates the use of a parameter, e.g., Y.1541, then
the QNE must strictly interpret it, otherwise it has to reject the
parameter or the admission request.

But this is exactly the problem that we wanted to solve by having this
discussions until now. Thus the problem is not solved.

Additionaly, Jerry, Dave and Cornelia change the main concepts of QoS
based NSIS, please see my previous e-mails that I have sent yesterday on
this.

I have already proposed a solution to solve the problems that we have had
in the nsis mailing list, see a copy of that proposal below:
-------------------------

Another proposal for solving the inetroperability issue and the issue of
having too many parameters was to:
* use only one mandatory parameter, i.e., Token Bucket that could be
remapped into the peak Bandwidth parameter.
Thus except the Token Bucket parameter,
* change the QSPEC-1 parameters into QSPEC-2 parameters
* describe the remapping procedure of the Token Bucket into the peak
Bandwidth parameter in the QSPEC template draft.

After that we can submit the QSPEC template draft for last WG call. If
after the deployment phase we conclude that more mandatory parameters
are needed for interoperability then we could work on an interworking
draft.

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

Best Regards,
Georgios

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 03:37:35 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfuXt-0005mO-QY; Fri, 03 Nov 2006 03:37:01 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfuXs-0005mG-8J
	for nsis@ietf.org; Fri, 03 Nov 2006 03:37:00 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfuXl-0001sB-As
	for nsis@ietf.org; Fri, 03 Nov 2006 03:37:00 -0500
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id 1A3FD1635;
	Fri,  3 Nov 2006 09:35:49 +0100 (CET)
Received: from esealmw116.eemea.ericsson.se ([153.88.200.7]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 09:35:48 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 09:35:47 +0100
Message-ID: <05F34B6959F19F43BBD78FFCD4C15D055DFFD6@esealmw116.eemea.ericsson.se>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQAHs9PaA=
From: =?iso-8859-1?Q?Attila_B=E1der_=28IJ/ETH=29?= <attila.bader@ericsson.com>
To: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>,
	<nsis@ietf.org>
X-OriginalArrivalTime: 03 Nov 2006 08:35:48.0324 (UTC)
	FILETIME=[0FF43A40:01C6FF23]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 76c7db407a166e4c39f35d8215d8dd32
Cc: "Kappler, Cornelia" <cornelia.kappler@siemens.com>,
	David R Oran <oran@cisco.com>, Cornelia Kappler <cornelia.kappler@gmx.net>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi All, I disagree with the proposal below.=20

This would remove the flexibility to develop and define new QoS models. =
Actually the proposal supports only the Intserv model. Stacking and =
remapping have been fundamental features.=20

I think that the best solution would be not to mandate parameters and =
have only optional one. If we want to distinguish Intserv as e2e QoS =
model, it is OK to mandate token bucket parameters, but remapping it to =
other parameters and stacking it should be allowed.

Best regards, Attila


>-----Original Message-----
>From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]=20
>Sent: Wednesday, November 01, 2006 11:13 PM
>To: nsis@ietf.org
>Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R=20
>Oran; Kappler,Cornelia
>Subject: [NSIS] Proposed Updates to the QSPEC Draft
>
>Hi All,
>
>Per John's suggestion below we've discussed Dave Oran's=20
>suggestions at length and have the following proposal. =20
>
>We think the proposed changes address many concerns raised on=20
>the list over the last few months and will require some rather=20
>basic changes in the QSPEC document=20
>http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
>
>We appreciate Dave's suggestions and careful review of the document.
>
>Here is a summary of the proposed changes:
>
>1. The notion of QOSMs will be played down in the updated draft.
>Language to the effect that 'QNSLP/QSPEC can signal for=20
>different QOSMs across multiple domains' will be replaced by=20
>the notion that QNSLP/QSPEC allows different elements on the=20
>path to implement different data plane QoS mechanisms that=20
>meets the constraints included in the QSPEC.  A QOSM describes=20
>common capabilities among QNEs to act consistently when=20
>requested to admit traffic and in treating admitted traffic. =20
>A QOSM ID need not be defined or signaled and a QNE need not=20
>support any particular QOSM.  A 'QOSM specification' would=20
>become a 'QNSLP/QSPEC applicability statement', which=20
>documents how a QNE/RMF interprets and treats the various=20
>elements in a QSPEC. =20
>
>An updated QOSM definition is as follows:
>
>QoS Model (QOSM): A method to achieve QoS for a traffic flow,=20
>e.g., IntServ Controlled Load.  A QOSM specifies what sub-set=20
>of QSPEC QoS constraints and traffic handling directives a QNE=20
>implementing that QOSM is capable of supporting and how=20
>resources will be managed by the RMF in such a QNE.
>
>2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
>semantics:
>A) The source traffic description is mandatory to include by=20
>the QNI and mandatory to interpret by downstream QNEs.  The=20
>traffic description is specified by a traffic model (TMOD)=20
>parameter consisting of 4
>sub-parameters:
>- rate (r)
>- bucket size (b)
>- peak rate (p)
>- minimum policed unit (m)
>The TMOD parameter is a mathematically complete way to=20
>describe a traffic source.  Examples will be given for setting=20
>TMOD to correspond to specific cases, e.g. where only=20
>bandwidth is available (r=3Dp; b and m set to default values)=20
>and other cases (e.g., TCP).  As such, the bandwidth parameter=20
>is eliminated.
>B) All other QSPEC parameters are optional to include by the=20
>QNI, and may be either mandatory or optional to interpret by=20
>downstream QNEs.
>That is, the QNI can explicitly specify for each constraint=20
>parameter whether it wishes admission control to succeed or=20
>fail if that constraint cannot be met.  A downstream QNE=20
>distinguishes whether it is failure to understand the=20
>parameter or failure to meet the constraint that causes the=20
>admission control failure (more detail on this below).
>C) Additional QSPEC parameters can be defined in updates to=20
>the QSPEC document (or alternatively specified in QOSM=20
>applicability statement documents).  There is already an IANA=20
>registry established for additional QSPEC parameters.
>
>3. The concept of remapping QSPEC parameters is eliminated. =20
>This concept is ill defined and has caused much confusion=20
>(there are few if any sensible remappings).  Rather we define=20
>'interpret' a QSPEC parameter to only mean what we now call=20
>'strictly interpret' (i.e., must conform to the RFCs defining=20
>the parameter and procedures) and eliminate the remap option=20
>part of 'interpret'.
>
>4. Since remapping is eliminated, the concept of stacking=20
>QSPECs is also eliminated.  Discussion of tunneling of QNSLP=20
>messages is eliminated since there are no impacts on tunneling=20
>required by QSPEC.
>
>5. Current QSPEC flags are modified as follows:
>A) The QNI sets the M flag for each QSPEC parameter it=20
>populates that it intends must be interpreted or the=20
>reservation fails.  Currently the M flag is statically set for=20
>every QSPEC parameter.
>B) If the M flag is set, a downstream QNE must interpret the=20
>parameter or the reservation fails.  If a QNE does not support=20
>the parameter with the M flag set, it sets the N flag and=20
>rejects the reservation.  If the QNE supports the parameter=20
>but cannot meet the parameter, it sets the E flag and rejects=20
>the reservation.=20
>C) If the M flag is not set, a downstream QNE should interpret=20
>the parameter.  If a QNE does not support the parameter, it=20
>sets the N flag and can optionally accept or reject the=20
>reservation.  If a QNE supports the parameter but cannot meet=20
>the parameter, it sets the E flag and optional accepts or=20
>rejects the reservation.
>D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
>eliminated.
>
>6. Summary of proposed QSPEC parameters:
>
>Source traffic Description:
> TMOD-1 (mandatory to include)
> TMOD-2 (optional to include)
>
>Constraints (optional to include):
> Path Latency
> Path Jitter
> Path PLR
> Path PER
> Y.1541 class (shorthand notation for above 4 parameters) =20
>Slack term  Priority (Preemption, Defending, Admission, RPH=20
>Priority)  DSTE class
>
>Handling directives (optional to include):
> excess treatment
>
>Traffic Classifiers (optional to include):
> PHB Class (note that PHB class is set by downstream QNE to=20
>tell QNI how to mark traffic to ensure it will be treated as=20
>committed by admission
>control)
>
>Eliminated:
> bandwidth
> Ctot, Dtot, Csum, Dsum
>
>7. Editorial changes:
>A) rewrite description of read-only and read-write parameters=20
>to make the rules more clear:=20
>- which parameters can be ro/rw (function of object in which parameter
>included)
>- can only increase rw values and not delete, etc.
>B) eliminate the distinction of 'control' and 'qos description'
>parameters and replace with suggested taxonomy, as described above
>C) clarify discussion of packet outcomes from Y.1540 and place=20
>in appropriate sections
>D) clarify which parameters can be included in <QoS Desired> <QoS
>Available>, <QoS Reserved>, or <QoS Minimum>
>
>Please review and comment on the proposed changes.
>
>Thanks,
>Jerry, Dave, Cornelia
>
>
>-----Original Message-----
>From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
>Sent: Monday, October 23, 2006 2:59 AM
>To: nsis@ietf.org
>Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
>
>Hi all,
>
>We have had a very large thread on the current=20
>draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear=20
>consensus, so I'd like to try to summarize the discussions.
>
>Several people feel that the current draft contains too many=20
>mandatory parameters and mapping some QoS Models to others is=20
>mixing apples-to-oranges.  The feeling is that=20
>interoperability is hard to achieve and therefore we should=20
>not try to gain interoperability.  Jukka Manner's message on=20
>17th of October proposed a way forward.
>http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
>
>However, several (other) people feel that interoperability is=20
>important and is needed. Giving up on interoperability would=20
>not be a good thing, and that we are very close to a solution.=20
> Dave Oran proposed a mechanism to reach a solution on the=20
>19th October:
>http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
>
>I have asked Jerry and Dave to make a proposal along the lines=20
>of Dave's message above, and the working group can review=20
>this, to see if we can gain consensus on this.  I think it=20
>would be sufficient to send some rough text to the mailing=20
>list, and if this is OK, then we can revise the draft.
>
>One open question to Dave/Jerry about the proposal is, in the=20
>event that the source cannot be described a mathematically=20
>complete way, is there any value (or does it make any sense at=20
>all) in sending an incomplete qspec, or potentially=20
>"wildcarding" sending some of the mandatory parameters.  I'm=20
>wondering if some sources / deployments may have certain=20
>parameters that are relatively well defined, but not all are=20
>so well defined.
>
>thanks,
>John
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 03:56:10 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfupq-00062k-2z; Fri, 03 Nov 2006 03:55:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfupp-00062Z-2p
	for nsis@ietf.org; Fri, 03 Nov 2006 03:55:33 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfupm-0003e1-I6
	for nsis@ietf.org; Fri, 03 Nov 2006 03:55:33 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Fri, 03 Nov 2006 10:55:25 +0200
	id 000AFD99.454B03FD.000046E9
Date: Fri, 3 Nov 2006 10:55:25 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] jerry, dave,	cornelia's proposal very confusing and not
	solving problem
In-Reply-To: <hvOvd46k.1162535221.3187200.karagian@ewi.utwente.nl>
Message-ID: <Pine.LNX.4.64.0611031047160.7180@sbz-31.cs.Helsinki.FI>
References: <hvOvd46k.1162535221.3187200.karagian@ewi.utwente.nl>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6cca30437e2d04f45110f2ff8dc1b1d5
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi,

AFAICT, if the QNI _really_ wants to have an Y.1541 class supported 
(mandatory) and a QNE on the path does not support that, the reservation 
will be rejected. Tough luck for the QNI. This forces QNIs to be quite 
conservative in what they ask to be "mandatory" on the path, and suggests 
that all parameters in category B should in fact be optional.

If a QNI sets too many, and too complex mandatory objects, then it will 
not get an e2e reservation, simple.

The RMD, or any other QoS spec, does not need to support all wierd 
requests from QNIs.

Is that what the JCD-team are proposing?

Jukka

On Fri, 3 Nov 2006, Georgios Karagiannis wrote:

> Hi all
> 
> After rereading the proposal of Jerry, Dave, Cornelia I have seen that the
> main problem that we have had until now in the mailing discussions is not
> solved. Mainly the number of mandatory parameters to interpret by all
> the QNEs in the universe is not solved.
> 
> The proposal mentions in point 2) point B):
> 
> "All other QSPEC parameters are optional to include by the QNI, and may
> be either mandatory or optional to interpret by downstream QNEs."
> 
> Meaning that if a QNI includes the Y.1541 class and wants that all
> downstream QNEs interpret this parameter, then all downstream QNEs must
> be able to strictly interpret the Y.1541 class.
> Jerry, Dave, Cornelia is this true?
> 
> Meaning that all QNEs that want to use NSIS must be implemented in such a
> way that if a QNI mandates the use of a parameter, e.g., Y.1541, then
> the QNE must strictly interpret it, otherwise it has to reject the
> parameter or the admission request.
> 
> But this is exactly the problem that we wanted to solve by having this
> discussions until now. Thus the problem is not solved.
> 
> Additionaly, Jerry, Dave and Cornelia change the main concepts of QoS
> based NSIS, please see my previous e-mails that I have sent yesterday on
> this.
> 
> I have already proposed a solution to solve the problems that we have had
> in the nsis mailing list, see a copy of that proposal below:
> -------------------------
> 
> Another proposal for solving the inetroperability issue and the issue of
> having too many parameters was to:
> * use only one mandatory parameter, i.e., Token Bucket that could be
> remapped into the peak Bandwidth parameter.
> Thus except the Token Bucket parameter,
> * change the QSPEC-1 parameters into QSPEC-2 parameters
> * describe the remapping procedure of the Token Bucket into the peak
> Bandwidth parameter in the QSPEC template draft.
> 
> After that we can submit the QSPEC template draft for last WG call. If
> after the deployment phase we conclude that more mandatory parameters
> are needed for interoperability then we could work on an interworking
> draft.
> 
> --------------
> 
> Best Regards,
> Georgios
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 04:30:29 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfvMt-0003Ef-8e; Fri, 03 Nov 2006 04:29:43 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfvMq-0003Dt-6M
	for nsis@ietf.org; Fri, 03 Nov 2006 04:29:40 -0500
Received: from mail.gmx.net ([213.165.64.20])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfvMn-0007b0-OI
	for nsis@ietf.org; Fri, 03 Nov 2006 04:29:40 -0500
Received: (qmail invoked by alias); 03 Nov 2006 09:29:35 -0000
Received: from ip-80-226-11-234.vodafone-net.de (EHLO [139.7.169.112])
	[80.226.11.234]
	by mail.gmx.net (mp009) with SMTP; 03 Nov 2006 10:29:35 +0100
X-Authenticated: #29516787
Message-ID: <454B0BFE.3000708@gmx.net>
Date: Fri, 03 Nov 2006 10:29:34 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.7 (Windows/20060909)
MIME-Version: 1.0
To: NSIS <nsis@ietf.org>
References: <454A0764.3070202@gmx.net>
In-Reply-To: <454A0764.3070202@gmx.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8abaac9e10c826e8252866cbe6766464
Subject: [NSIS] QSPEC Face-to-Face Chat: Another Try
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all,

let me try it again.

What about Sunday, 3pm at the Sheraton San Diego Hotel (= meeting venue) 
in front of the hotel reception.

Would this work for someone?

Ciao
Hannes

Hannes Tschofenig wrote:
> Hi all,
> 
> I wonder how many NSIS folks are available for a QSPEC face-to-face 
> discussion Sunday morning?
> 
> Since many folks have to be at company internal meetings during lunch 
> time and want to see some of the tutorials I would suggest 9am at the
> Sheraton San Diego Hotel (= meeting venue) in front of the hotel reception.
> 
> Is this a useful idea?
> 
> Ciao
> Hannes
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 04:34:54 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfvRJ-0004wo-SN; Fri, 03 Nov 2006 04:34:17 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfvRI-0004wi-Cl
	for nsis@ietf.org; Fri, 03 Nov 2006 04:34:16 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfvRF-00082R-VX
	for nsis@ietf.org; Fri, 03 Nov 2006 04:34:16 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA39Y498000538; Fri, 3 Nov 2006 10:34:07 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Hannes Tschofenig'" <Hannes.Tschofenig@gmx.net>, "'NSIS'" <nsis@ietf.org>
Subject: RE: [NSIS] QSPEC Face-to-Face Chat: Another Try
Date: Fri, 3 Nov 2006 10:34:04 +0100
Message-ID: <000201c6ff2b$37153090$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <454B0BFE.3000708@gmx.net>
Thread-Index: Acb/KsenT+w7d/n0RVG8bI6p34xurwAAFiiQ
X-Spam-Score: 0.251 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 10:34:08 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 5a9a1bd6c2d06a21d748b7d0070ddcb8
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Hannes

At 3 p.m. will work for me!

Best Regards,
Georgios
 

> -----Original Message-----
> From: Hannes Tschofenig [mailto:Hannes.Tschofenig@gmx.net] 
> Sent: vrijdag 3 november 2006 10:30
> To: NSIS
> Subject: [NSIS] QSPEC Face-to-Face Chat: Another Try
> 
> Hi all,
> 
> let me try it again.
> 
> What about Sunday, 3pm at the Sheraton San Diego Hotel (= 
> meeting venue) in front of the hotel reception.
> 
> Would this work for someone?
> 
> Ciao
> Hannes
> 
> Hannes Tschofenig wrote:
> > Hi all,
> > 
> > I wonder how many NSIS folks are available for a QSPEC face-to-face 
> > discussion Sunday morning?
> > 
> > Since many folks have to be at company internal meetings 
> during lunch 
> > time and want to see some of the tutorials I would suggest 
> 9am at the 
> > Sheraton San Diego Hotel (= meeting venue) in front of the 
> hotel reception.
> > 
> > Is this a useful idea?
> > 
> > Ciao
> > Hannes
> > 
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 04:41:45 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfvXj-0004qv-Jy; Fri, 03 Nov 2006 04:40:55 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfvXi-0004qq-IE
	for nsis@ietf.org; Fri, 03 Nov 2006 04:40:54 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfvXg-0000Ng-7G
	for nsis@ietf.org; Fri, 03 Nov 2006 04:40:54 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Fri, 03 Nov 2006 11:40:51 +0200
	id 000AFD7C.454B0EA3.00002394
Date: Fri, 3 Nov 2006 11:40:49 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
Subject: Re: [NSIS] QSPEC Face-to-Face Chat: Another Try
In-Reply-To: <454B0BFE.3000708@gmx.net>
Message-ID: <Pine.LNX.4.64.0611031140410.7180@sbz-31.cs.Helsinki.FI>
References: <454A0764.3070202@gmx.net> <454B0BFE.3000708@gmx.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cab78e1e39c4b328567edb48482b6a69
Cc: NSIS <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Works for me.

Jukka

On Fri, 3 Nov 2006, Hannes Tschofenig wrote:

> Hi all,
> 
> let me try it again.
> 
> What about Sunday, 3pm at the Sheraton San Diego Hotel (= meeting venue) in
> front of the hotel reception.
> 
> Would this work for someone?
> 
> Ciao
> Hannes
> 
> Hannes Tschofenig wrote:
> > Hi all,
> > 
> > I wonder how many NSIS folks are available for a QSPEC face-to-face
> > discussion Sunday morning?
> > 
> > Since many folks have to be at company internal meetings during lunch time
> > and want to see some of the tutorials I would suggest 9am at the
> > Sheraton San Diego Hotel (= meeting venue) in front of the hotel reception.
> > 
> > Is this a useful idea?
> > 
> > Ciao
> > Hannes
> > 
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 04:49:05 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfveo-00080J-FE; Fri, 03 Nov 2006 04:48:14 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfvem-000805-Ow
	for nsis@ietf.org; Fri, 03 Nov 2006 04:48:12 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfvel-0001Ab-Ta
	for nsis@ietf.org; Fri, 03 Nov 2006 04:48:12 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Fri, 03 Nov 2006 11:48:11 +0200
	id 000AFD23.454B105B.0000561C
Date: Fri, 3 Nov 2006 11:48:10 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: "=?iso-8859-1?Q?Attila_B=E1der_=28IJ/ETH=29?=" <attila.bader@ericsson.com>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
In-Reply-To: <05F34B6959F19F43BBD78FFCD4C15D055DFFD6@esealmw116.eemea.ericsson.se>
Message-ID: <Pine.LNX.4.64.0611031144540.7180@sbz-31.cs.Helsinki.FI>
References: <05F34B6959F19F43BBD78FFCD4C15D055DFFD6@esealmw116.eemea.ericsson.se>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="=_courier-22048-1162547291-0001-2"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 162d87dc0b780d17da9b1934777fd451
Cc: "Ash, Gerald R \\\(Jerry\\\), ALABS" <gash@att.com>,
	Cornelia Kappler <cornelia.kappler@gmx.net>,
	David R Oran <oran@cisco.com>, nsis@ietf.org, "Kappler,
	Cornelia" <cornelia.kappler@siemens.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

This is a MIME-formatted message.  If you see this text it means that your
E-mail software does not support MIME-formatted messages.

--=_courier-22048-1162547291-0001-2
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable
X-Mime-Autoconverted: from 8bit to quoted-printable by courier 0.47


Hi all,

I tend to agree with Attila here that stacking is a new concept and is 
important to have. E.g., it allows a domain to add QoS information that =
is 
only relevant to the domain itself, an ingress QNE pushes a new QSPEC in=
to 
the NSLP message, and the egress QNE of the domain just pops it out. A 
domain isnt't required to do so, but allows new kinds of QoS management 
options.

Jukka

On Fri, 3 Nov 2006, Attila B=E1der (IJ/ETH) wrote:

> Hi All, I disagree with the proposal below. 
> 
> This would remove the flexibility to develop and define new QoS models=
. 
> Actually the proposal supports only the Intserv model. Stacking and 
> remapping have been fundamental features.
> 
> I think that the best solution would be not to mandate parameters and 
> have only optional one. If we want to distinguish Intserv as e2e QoS 
> model, it is OK to mandate token bucket parameters, but remapping it t=
o 
> other parameters and stacking it should be allowed.
> 
> Best regards, Attila
> 
> 
> >-----Original Message-----
> >From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com] 
> >Sent: Wednesday, November 01, 2006 11:13 PM
> >To: nsis@ietf.org
> >Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R 
> >Oran; Kappler,Cornelia
> >Subject: [NSIS] Proposed Updates to the QSPEC Draft
> >
> >Hi All,
> >
> >Per John's suggestion below we've discussed Dave Oran's 
> >suggestions at length and have the following proposal.  
> >
> >We think the proposed changes address many concerns raised on 
> >the list over the last few months and will require some rather 
> >basic changes in the QSPEC document 
> >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> >
> >We appreciate Dave's suggestions and careful review of the document.
> >
> >Here is a summary of the proposed changes:
> >
> >1. The notion of QOSMs will be played down in the updated draft.
> >Language to the effect that 'QNSLP/QSPEC can signal for 
> >different QOSMs across multiple domains' will be replaced by 
> >the notion that QNSLP/QSPEC allows different elements on the 
> >path to implement different data plane QoS mechanisms that 
> >meets the constraints included in the QSPEC.  A QOSM describes 
> >common capabilities among QNEs to act consistently when 
> >requested to admit traffic and in treating admitted traffic.  
> >A QOSM ID need not be defined or signaled and a QNE need not 
> >support any particular QOSM.  A 'QOSM specification' would 
> >become a 'QNSLP/QSPEC applicability statement', which 
> >documents how a QNE/RMF interprets and treats the various 
> >elements in a QSPEC.  
> >
> >An updated QOSM definition is as follows:
> >
> >QoS Model (QOSM): A method to achieve QoS for a traffic flow, 
> >e.g., IntServ Controlled Load.  A QOSM specifies what sub-set 
> >of QSPEC QoS constraints and traffic handling directives a QNE 
> >implementing that QOSM is capable of supporting and how 
> >resources will be managed by the RMF in such a QNE.
> >
> >2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
> >semantics:
> >A) The source traffic description is mandatory to include by 
> >the QNI and mandatory to interpret by downstream QNEs.  The 
> >traffic description is specified by a traffic model (TMOD) 
> >parameter consisting of 4
> >sub-parameters:
> >- rate (r)
> >- bucket size (b)
> >- peak rate (p)
> >- minimum policed unit (m)
> >The TMOD parameter is a mathematically complete way to 
> >describe a traffic source.  Examples will be given for setting 
> >TMOD to correspond to specific cases, e.g. where only 
> >bandwidth is available (r=3Dp; b and m set to default values) 
> >and other cases (e.g., TCP).  As such, the bandwidth parameter 
> >is eliminated.
> >B) All other QSPEC parameters are optional to include by the 
> >QNI, and may be either mandatory or optional to interpret by 
> >downstream QNEs.
> >That is, the QNI can explicitly specify for each constraint 
> >parameter whether it wishes admission control to succeed or 
> >fail if that constraint cannot be met.  A downstream QNE 
> >distinguishes whether it is failure to understand the 
> >parameter or failure to meet the constraint that causes the 
> >admission control failure (more detail on this below).
> >C) Additional QSPEC parameters can be defined in updates to 
> >the QSPEC document (or alternatively specified in QOSM 
> >applicability statement documents).  There is already an IANA 
> >registry established for additional QSPEC parameters.
> >
> >3. The concept of remapping QSPEC parameters is eliminated.  
> >This concept is ill defined and has caused much confusion 
> >(there are few if any sensible remappings).  Rather we define 
> >'interpret' a QSPEC parameter to only mean what we now call 
> >'strictly interpret' (i.e., must conform to the RFCs defining 
> >the parameter and procedures) and eliminate the remap option 
> >part of 'interpret'.
> >
> >4. Since remapping is eliminated, the concept of stacking 
> >QSPECs is also eliminated.  Discussion of tunneling of QNSLP 
> >messages is eliminated since there are no impacts on tunneling 
> >required by QSPEC.
> >
> >5. Current QSPEC flags are modified as follows:
> >A) The QNI sets the M flag for each QSPEC parameter it 
> >populates that it intends must be interpreted or the 
> >reservation fails.  Currently the M flag is statically set for 
> >every QSPEC parameter.
> >B) If the M flag is set, a downstream QNE must interpret the 
> >parameter or the reservation fails.  If a QNE does not support 
> >the parameter with the M flag set, it sets the N flag and 
> >rejects the reservation.  If the QNE supports the parameter 
> >but cannot meet the parameter, it sets the E flag and rejects 
> >the reservation. 
> >C) If the M flag is not set, a downstream QNE should interpret 
> >the parameter.  If a QNE does not support the parameter, it 
> >sets the N flag and can optionally accept or reject the 
> >reservation.  If a QNE supports the parameter but cannot meet 
> >the parameter, it sets the E flag and optional accepts or 
> >rejects the reservation.
> >D) The R (remapped parameter) flag and Q (non QOSM) flag are 
> >eliminated.
> >
> >6. Summary of proposed QSPEC parameters:
> >
> >Source traffic Description:
> > TMOD-1 (mandatory to include)
> > TMOD-2 (optional to include)
> >
> >Constraints (optional to include):
> > Path Latency
> > Path Jitter
> > Path PLR
> > Path PER
> > Y.1541 class (shorthand notation for above 4 parameters)  
> >Slack term  Priority (Preemption, Defending, Admission, RPH 
> >Priority)  DSTE class
> >
> >Handling directives (optional to include):
> > excess treatment
> >
> >Traffic Classifiers (optional to include):
> > PHB Class (note that PHB class is set by downstream QNE to 
> >tell QNI how to mark traffic to ensure it will be treated as 
> >committed by admission
> >control)
> >
> >Eliminated:
> > bandwidth
> > Ctot, Dtot, Csum, Dsum
> >
> >7. Editorial changes:
> >A) rewrite description of read-only and read-write parameters 
> >to make the rules more clear: 
> >- which parameters can be ro/rw (function of object in which paramete=
r
> >included)
> >- can only increase rw values and not delete, etc.
> >B) eliminate the distinction of 'control' and 'qos description'
> >parameters and replace with suggested taxonomy, as described above
> >C) clarify discussion of packet outcomes from Y.1540 and place 
> >in appropriate sections
> >D) clarify which parameters can be included in <QoS Desired> <QoS
> >Available>, <QoS Reserved>, or <QoS Minimum>
> >
> >Please review and comment on the proposed changes.
> >
> >Thanks,
> >Jerry, Dave, Cornelia
> >
> >
> >-----Original Message-----
> >From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> >Sent: Monday, October 23, 2006 2:59 AM
> >To: nsis@ietf.org
> >Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
> >
> >Hi all,
> >
> >We have had a very large thread on the current 
> >draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear 
> >consensus, so I'd like to try to summarize the discussions.
> >
> >Several people feel that the current draft contains too many 
> >mandatory parameters and mapping some QoS Models to others is 
> >mixing apples-to-oranges.  The feeling is that 
> >interoperability is hard to achieve and therefore we should 
> >not try to gain interoperability.  Jukka Manner's message on 
> >17th of October proposed a way forward.
> >http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> >
> >However, several (other) people feel that interoperability is 
> >important and is needed. Giving up on interoperability would 
> >not be a good thing, and that we are very close to a solution. 
> > Dave Oran proposed a mechanism to reach a solution on the 
> >19th October:
> >http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> >
> >I have asked Jerry and Dave to make a proposal along the lines 
> >of Dave's message above, and the working group can review 
> >this, to see if we can gain consensus on this.  I think it 
> >would be sufficient to send some rough text to the mailing 
> >list, and if this is OK, then we can revise the draft.
> >
> >One open question to Dave/Jerry about the proposal is, in the 
> >event that the source cannot be described a mathematically 
> >complete way, is there any value (or does it make any sense at 
> >all) in sending an incomplete qspec, or potentially 
> >"wildcarding" sending some of the mandatory parameters.  I'm 
> >wondering if some sources / deployments may have certain 
> >parameters that are relatively well defined, but not all are 
> >so well defined.
> >
> >thanks,
> >John
> >
> >_______________________________________________
> >nsis mailing list
> >nsis@ietf.org
> >https://www1.ietf.org/mailman/listinfo/nsis
> >
> >_______________________________________________
> >nsis mailing list
> >nsis@ietf.org
> >https://www1.ietf.org/mailman/listinfo/nsis
> >
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 
--=_courier-22048-1162547291-0001-2
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--=_courier-22048-1162547291-0001-2--




From nsis-bounces@ietf.org Fri Nov 03 04:58:25 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfvnh-0007vP-Mw; Fri, 03 Nov 2006 04:57:25 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfvnd-0007ta-9l
	for nsis@ietf.org; Fri, 03 Nov 2006 04:57:21 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfvnR-0002A5-Dg
	for nsis@ietf.org; Fri, 03 Nov 2006 04:57:21 -0500
Received: from esealmw128.eemea.ericsson.se (unknown [153.88.254.121])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id 82B035F2; 
	Fri,  3 Nov 2006 10:56:56 +0100 (CET)
Received: from esealmw128.eemea.ericsson.se ([153.88.254.176]) by
	esealmw128.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 10:56:56 +0100
Received: from ericsson.com ([147.214.26.58]) by esealmw128.eemea.ericsson.se
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 10:56:56 +0100
Message-ID: <454B1268.4030405@ericsson.com>
Date: Fri, 03 Nov 2006 10:56:56 +0100
From: Lars Westberg <Lars.westberg@ericsson.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.4) Gecko/20030624 Netscape/7.1 (ax)
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>
	<F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>
In-Reply-To: <F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>
X-OriginalArrivalTime: 03 Nov 2006 09:56:56.0094 (UTC)
	FILETIME=[655EFBE0:01C6FF2E]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.6 (/)
X-Scan-Signature: a233275913d1d34f257cf9b4f3544846
Cc: john.loughney@nokia.com, Georgios Karagiannis <karagian@cs.utwente.nl>,
	=?ISO-8859-1?Q?=22Attila_B=E1?=, nsis@ietf.org,
	=?ISO-8859-1?Q?der_=28IJ/ETH=29=22?= <attila.bader@ericsson.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0146235622=="
Errors-To: nsis-bounces@ietf.org


This is a multi-part message in MIME format.
--===============0146235622==
Content-Type: multipart/alternative;
	boundary="------------010506000106020009040200"


This is a multi-part message in MIME format.
--------------010506000106020009040200
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit


Let's start with the historic aspect of QoS. Since ATM was defined 
during beginning of 90, the discussion about traffic managment (QOS) 
started to explode. Since then,  all standardization organization has 
tried to defined QOS-parameters based on their own scenario. The same 
happends in IETF with RSVP and Int.serv. We have today parameters and 
signalling sequence diagram defined differently in each standards-body 
especially for different access networks.


Another problem that we experiencing is that the application programmer 
do not understand QoS-API. They actually do not know how to set the 
parameters.
The basic problem is that the QNI will not generate appropriate 
parameter set given the functionality in the sub-path the signalling is 
traversing.
so, the reason for QoS-model stacking is to allow a network entity to 
introduce more advanced QoS-functionality that is required for the 
specific sub-path.

However, the specificly added parameters may also be needed between two 
sub-paths i.e. different access networks with a backbones (Internet?) in 
between.
   
so I recommend that the stacking feature should be a part of the 
QoS-spec template so we can allow for new usage scenario of the NSIS. If 
it is taken away, I do not se any difference between RSVP and NSIS.

I do not understand how the process works. Should we now start a 
blank-sheet argue functionality from beginning ?????

the main reason for bringing the discussion of parameters has nothing to 
with RMD. It is an arhitectural limitation. Just compared it to RSVP. 
RSVP do not have any hardcoded QoS-paraneters into the core part of the 
protocol specification, while NSIS has????
Given the present proposal in the QoS-spec, the architecture is much 
less open for new use than RSVP. Should NSIS make the same mistake as RSVP ?

comment about RMD:

the design target for RMD is twofold:
    a) support for PCN-like bandwidth management.
    b) scalable per-flow reservation.

during many years of discussion, we can conclude that additional 
functionality like RSVP-aggregation have ben design to cope with the 
state-explosion for per-flow reservation. RMD is developed to solve it 
inother approach: scalable per-flow reseravtion with the simillar 
resource guarantee as RSVP.

Our target was actually to improve RSVP more improved scalabillity to 
solve a problem within the mobile access network. But, NSIS becomes the 
answer to the standardization discission so we introduced it in this 
context instead.

regards Lasse

David R Oran wrote:

> I think some detailed discussion of this is going to be needed, 
> because after considerable pondering the three of us could not at all 
> figure out what either remapping or QSPEC stacking was good for. It 
> certainly introduces a vast amount of complexity and confusion as 
> currently formulated.
>
> It appears at least to me that one of three situation obtains:
> (a) we (well, at least I) don't understand RMD
> (b) there is some general QoS concept that we are missing - we went 
> through some complex scenarios like reservation aggregation any 
> multiple QOSMs chained together and still could not figure out how 
> one needed either stacking or remapping of QSPEC parameters
> (c) RMD is really a completely different QoS signaling architecture 
> from the rest of NSIS and is being shoehorned in by complicating 
> QNSLP and QSPEC to allow multiple QoS signaling schemes to masquerade 
> as a single architecture.
>
> I really hope it's (a) or (b) since then we can probably resolve this 
> by either putting back the missing functionality or figuring out 
> another simple way to accommodate RMD. What I'm most afraid of is 
> that this is really (c).
>
>  From the little I understand RMD, it's really like mapping onto a 
> custom L2 QoS scheme like ATM or DOCSIS, and that rather than doing 
> the mapping the way any conventional L3 RMF would do it, you co-
> mingle RMD's model with NSIS's model.
>
> But, I could of course be wildly wrong here.
>
> Dave.
>
>
> On Nov 2, 2006, at 6:00 AM, Georgios Karagiannis wrote:
>
> > Hi John
> >
> > I have just come back from a trip and I have just look into point 4):
> >>> 4. Since remapping is eliminated, the concept of stacking QSPECs is
> >>> also eliminated.  Discussion of tunneling of QNSLP messages is
> >>> eliminated since there are no impacts on tunneling required by 
> >>> QSPEC.
> >
> > QSPEC stacking is one of the significant concepts in the QoS based 
> > NSIS
> > solution.
> > This concept provides the additional flexibility that is missing 
> > from RSVP.
> > Furthermore, all the work that we have done until now is based on
> > the concept of QSPEC stacking. Thus what you actually say is to just
> > forget the work that we have done
> > until now, in e.g.the RMD-QOSM draft, and start from the beginning.
> >
> >
> > The main problem that we have had until now was not the QSPEC stacking
> > issue,
> > but it was the huge number of mandatory paremeters. Actualy I 
> > consider the
> > issue of removing the
> > QSPEC stacking concept a much more severe problem on achieving the
> > flexibility
> > that we wanted to achieve with NSIS than the huge number of mandatory
> > parameters.
> >
> > Therefore, I am severely against removing the QSPEC stacking concept!
> >
> > Best Regards,
> > Georgios
> >
> >
> >
> >
> >
> >> -----Original Message-----
> >> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> >> Sent: donderdag 2 november 2006 8:35
> >> To: gash@att.com; nsis@ietf.org
> >> Cc: cornelia.kappler@siemens.com; oran@cisco.com;
> >> cornelia.kappler@gmx.net
> >> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> >>
> >> Jerry, Dave and Cornelia;
> >>
> >> This looks like a very good simplification of the QSPEC, and
> >> I think that is a good idea.  I think that there will be some
> >> implications to the QoSM drafts - I think QoS Model drafts
> >> will need to contain some more applicability statement type
> >> of text, along with deployment considerations.
> >>
> >> A plus side of this is that I think that the QoS template
> >> will become more broadly applicable to other uses than just
> >> NSIS QoS Signaling, so I think that is a gain as well.
> >>
> >> I'd really like people to review this and send comments on it
> >> so that we could really focus the discussion next week and
> >> get this work finally resolved.
> >>
> >> Thanks a lot for working through this, I think this is a good
> >> basis for a proposal.
> >>
> >> thanks,
> >> John
> >>
> >>> -----Original Message-----
> >>> From: ext Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
> >>> Sent: 02 November, 2006 00:13
> >>> To: nsis@ietf.org
> >>> Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran;
> >>> Kappler,Cornelia
> >>> Subject: [NSIS] Proposed Updates to the QSPEC Draft
> >>>
> >>> Hi All,
> >>>
> >>> Per John's suggestion below we've discussed Dave Oran's
> >> suggestions at
> >>> length and have the following proposal.
> >>>
> >>> We think the proposed changes address many concerns raised
> >> on the list
> >>> over the last few months and will require some rather basic
> >> changes in
> >>> the QSPEC document
> >>> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> >>>
> >>> We appreciate Dave's suggestions and careful review of the document.
> >>>
> >>> Here is a summary of the proposed changes:
> >>>
> >>> 1. The notion of QOSMs will be played down in the updated draft.
> >>> Language to the effect that 'QNSLP/QSPEC can signal for
> >> different QOSMs
> >>> across multiple domains' will be replaced by the notion that
> >>> QNSLP/QSPEC allows different elements on the path to implement
> >>> different data plane QoS mechanisms that meets the
> >> constraints included
> >>> in the QSPEC.  A QOSM describes common capabilities among
> >> QNEs to act
> >>> consistently when requested to admit traffic and in treating
> >> admitted
> >>> traffic.
> >>> A QOSM ID need not be defined or signaled and a QNE need not support
> >>> any particular QOSM.  A 'QOSM specification' would become a
> >>> 'QNSLP/QSPEC applicability statement', which documents how a QNE/RMF
> >>> interprets and treats the various elements in a QSPEC.
> >>>
> >>> An updated QOSM definition is as follows:
> >>>
> >>> QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
> >>> IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
> >>> constraints and traffic handling directives a QNE implementing that
> >>> QOSM is capable of supporting and how resources will be
> >> managed by the
> >>> RMF in such a QNE.
> >>>
> >>> 2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --
> >>> > d)
> >>> semantics:
> >>> A) The source traffic description is mandatory to include by the QNI
> >>> and mandatory to interpret by downstream QNEs.  The traffic
> >> description
> >>> is specified by a traffic model (TMOD) parameter consisting of 4
> >>> sub-parameters:
> >>> - rate (r)
> >>> - bucket size (b)
> >>> - peak rate (p)
> >>> - minimum policed unit (m)
> >>> The TMOD parameter is a mathematically complete way to describe a
> >>> traffic source.  Examples will be given for setting TMOD to
> >> correspond
> >>> to specific cases, e.g. where only bandwidth is available
> >> (r=p; b and m
> >>> set to default values) and other cases (e.g., TCP).  As such, the
> >>> bandwidth parameter is eliminated.
> >>> B) All other QSPEC parameters are optional to include by the
> >> QNI, and
> >>> may be either mandatory or optional to interpret by downstream QNEs.
> >>> That is, the QNI can explicitly specify for each constraint
> >> parameter
> >>> whether it wishes admission control to succeed or fail if that
> >>> constraint cannot be met.  A downstream QNE distinguishes
> >> whether it is
> >>> failure to understand the parameter or failure to meet the
> >> constraint
> >>> that causes the admission control failure (more detail on
> >> this below).
> >>> C) Additional QSPEC parameters can be defined in updates to
> >> the QSPEC
> >>> document (or alternatively specified in QOSM applicability statement
> >>> documents).  There is already an IANA registry established for
> >>> additional QSPEC parameters.
> >>>
> >>> 3. The concept of remapping QSPEC parameters is eliminated.
> >>> This concept is ill defined and has caused much confusion (there are
> >>> few if any sensible remappings).  Rather we define
> >> 'interpret' a QSPEC
> >>> parameter to only mean what we now call 'strictly interpret' (i.e.,
> >>> must conform to the RFCs defining the parameter and procedures) and
> >>> eliminate the remap option part of 'interpret'.
> >>>
> >>> 4. Since remapping is eliminated, the concept of stacking QSPECs is
> >>> also eliminated.  Discussion of tunneling of QNSLP messages is
> >>> eliminated since there are no impacts on tunneling required by 
> >>> QSPEC.
> >>>
> >>> 5. Current QSPEC flags are modified as follows:
> >>> A) The QNI sets the M flag for each QSPEC parameter it
> >> populates that
> >>> it intends must be interpreted or the reservation fails.
> >> Currently the
> >>> M flag is statically set for every QSPEC parameter.
> >>> B) If the M flag is set, a downstream QNE must interpret the
> >> parameter
> >>> or the reservation fails.  If a QNE does not support the
> >> parameter with
> >>> the M flag set, it sets the N flag and rejects the
> >> reservation.  If the
> >>> QNE supports the parameter but cannot meet the parameter, it
> >> sets the E
> >>> flag and rejects the reservation.
> >>> C) If the M flag is not set, a downstream QNE should interpret the
> >>> parameter.  If a QNE does not support the parameter, it sets
> >> the N flag
> >>> and can optionally accept or reject the reservation.  If a
> >> QNE supports
> >>> the parameter but cannot meet the parameter, it sets the E flag and
> >>> optional accepts or rejects the reservation.
> >>> D) The R (remapped parameter) flag and Q (non QOSM) flag are
> >>> eliminated.
> >>>
> >>> 6. Summary of proposed QSPEC parameters:
> >>>
> >>> Source traffic Description:
> >>> TMOD-1 (mandatory to include)
> >>> TMOD-2 (optional to include)
> >>>
> >>> Constraints (optional to include):
> >>> Path Latency
> >>> Path Jitter
> >>> Path PLR
> >>> Path PER
> >>> Y.1541 class (shorthand notation for above 4 parameters)
> >> Slack term
> >>> Priority (Preemption, Defending, Admission, RPH
> >>> Priority)  DSTE class
> >>>
> >>> Handling directives (optional to include):
> >>> excess treatment
> >>>
> >>> Traffic Classifiers (optional to include):
> >>> PHB Class (note that PHB class is set by downstream QNE to tell QNI
> >>> how to mark traffic to ensure it will be treated as committed by
> >>> admission
> >>> control)
> >>>
> >>> Eliminated:
> >>> bandwidth
> >>> Ctot, Dtot, Csum, Dsum
> >>>
> >>> 7. Editorial changes:
> >>> A) rewrite description of read-only and read-write
> >> parameters to make
> >>> the rules more clear:
> >>> - which parameters can be ro/rw (function of object in which
> >> parameter
> >>> included)
> >>> - can only increase rw values and not delete, etc.
> >>> B) eliminate the distinction of 'control' and 'qos description'
> >>> parameters and replace with suggested taxonomy, as described above
> >>> C) clarify discussion of packet outcomes from Y.1540 and place in
> >>> appropriate sections
> >>> D) clarify which parameters can be included in <QoS Desired> <QoS
> >>> Available>, <QoS Reserved>, or <QoS Minimum>
> >>>
> >>> Please review and comment on the proposed changes.
> >>>
> >>> Thanks,
> >>> Jerry, Dave, Cornelia
> >>>
> >>>
> >>> -----Original Message-----
> >>> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> >>> Sent: Monday, October 23, 2006 2:59 AM
> >>> To: nsis@ietf.org
> >>> Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
> >>>
> >>> Hi all,
> >>>
> >>> We have had a very large thread on the current
> >>> draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear
> >> consensus,
> >>> so I'd like to try to summarize the discussions.
> >>>
> >>> Several people feel that the current draft contains too many
> >> mandatory
> >>> parameters and mapping some QoS Models to others is mixing
> >>> apples-to-oranges.  The feeling is that interoperability is hard to
> >>> achieve and therefore we should not try to gain interoperability.
> >>> Jukka Manner's message on 17th of October proposed a way forward.
> >>> http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> >>>
> >>> However, several (other) people feel that interoperability
> >> is important
> >>> and is needed. Giving up on interoperability would not be a
> >> good thing,
> >>> and that we are very close to a solution.
> >>> Dave Oran proposed a mechanism to reach a solution on the 19th
> >>> October:
> >>> http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> >>>
> >>> I have asked Jerry and Dave to make a proposal along the lines of
> >>> Dave's message above, and the working group can review this,
> >> to see if
> >>> we can gain consensus on this.  I think it would be
> >> sufficient to send
> >>> some rough text to the mailing list, and if this is OK, then we can
> >>> revise the draft.
> >>>
> >>> One open question to Dave/Jerry about the proposal is, in the event
> >>> that the source cannot be described a mathematically
> >> complete way, is
> >>> there any value (or does it make any sense at
> >>> all) in sending an incomplete qspec, or potentially "wildcarding"
> >>> sending some of the mandatory parameters.  I'm wondering if some
> >>> sources / deployments may have certain parameters that are
> >> relatively
> >>> well defined, but not all are so well defined.
> >>>
> >>> thanks,
> >>> John
> >>>
> >>> _______________________________________________
> >>> nsis mailing list
> >>> nsis@ietf.org
> >>> https://www1.ietf.org/mailman/listinfo/nsis
> >>>
> >>> _______________________________________________
> >>> nsis mailing list
> >>> nsis@ietf.org
> >>> https://www1.ietf.org/mailman/listinfo/nsis
> >>>
> >>
> >> _______________________________________________
> >> nsis mailing list
> >> nsis@ietf.org
> >> https://www1.ietf.org/mailman/listinfo/nsis
> >>
> >
> >
> >
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
>

--------------010506000106020009040200
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta http-equiv="Content-Type" content="text/html;charset=ISO-8859-1">
  <title></title>
</head>
<body text="#000000" bgcolor="#ffffff">
<br>
Let's start with the historic aspect of QoS. Since ATM was defined
during beginning of 90, the discussion about traffic managment (QOS)
started to explode. Since then,&nbsp; all standardization organization has
tried to defined QOS-parameters based on their own scenario. The same
happends in IETF with RSVP and Int.serv. We have today parameters and
signalling sequence diagram defined differently in each standards-body
especially for different access networks. <br>
<br>
<br>
Another problem that we experiencing is that the application programmer
do not understand QoS-API. They actually do not know how to set the
parameters.<br>
The basic problem is that the QNI will not generate appropriate
parameter set given the functionality in the sub-path the signalling is
traversing. <br>
so, the reason for QoS-model stacking is to allow a network entity to
introduce more advanced QoS-functionality that is required for the
specific sub-path. <br>
<br>
However, the specificly added parameters may also be needed between two
sub-paths i.e. different access networks with a backbones (Internet?)
in between.<br>
&nbsp;&nbsp;&nbsp; <br>
so I recommend that the stacking feature should be a part of the
QoS-spec template so we can allow for new usage scenario of the NSIS.
If it is taken away, I do not se any difference between RSVP and NSIS.<br>
<br>
I do not understand how the process works. Should we now start a
blank-sheet argue functionality from beginning ????? <br>
<br>
the main reason for bringing the discussion of parameters has nothing
to with RMD. It is an arhitectural limitation. Just compared it to
RSVP. RSVP do not have any hardcoded QoS-paraneters into the core part
of the protocol specification, while NSIS has????<br>
Given the present proposal in the QoS-spec, the architecture is much
less open for new use than RSVP. Should NSIS make the same mistake as
RSVP ?<br>
<br>
comment about RMD:<br>
<br>
the design target for RMD is twofold: <br>
&nbsp;&nbsp;&nbsp; a) support for PCN-like bandwidth management.<br>
&nbsp;&nbsp;&nbsp; b) scalable per-flow reservation.<br>
<br>
during many years of discussion, we can conclude that additional
functionality like RSVP-aggregation have ben design to cope with the
state-explosion for per-flow reservation. RMD is developed to solve it
inother approach: scalable per-flow reseravtion with the simillar
resource guarantee as RSVP. <br>
<br>
Our target was actually to improve RSVP more improved scalabillity to
solve a problem within the mobile access network. But, NSIS becomes the
answer to the standardization discission so we introduced it in this
context instead. <br>
<br>
regards Lasse<br>
<br>
David R Oran wrote:<br>
<blockquote type="cite"
 cite="midF0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com">
  <meta http-equiv="Content-Type" content="text/html; ">
  <meta name="Generator"
 content="MS Exchange Server version 6.5.7650.28">
  <title>Re: [NSIS] Proposed Updates to the QSPEC Draft</title>
<!-- Converted from text/plain format -->
  <p><font size="2">I think some detailed discussion of this is going
to be needed,&nbsp;<br>
because after considerable pondering the three of us could not at all&nbsp;<br>
figure out what either remapping or QSPEC stacking was good for. It&nbsp;<br>
certainly introduces a vast amount of complexity and confusion as&nbsp;<br>
currently formulated.<br>
  <br>
It appears at least to me that one of three situation obtains:<br>
(a) we (well, at least I) don't understand RMD<br>
(b) there is some general QoS concept that we are missing - we went&nbsp;<br>
through some complex scenarios like reservation aggregation any&nbsp;<br>
multiple QOSMs chained together and still could not figure out how&nbsp;<br>
one needed either stacking or remapping of QSPEC parameters<br>
(c) RMD is really a completely different QoS signaling architecture&nbsp;<br>
from the rest of NSIS and is being shoehorned in by complicating&nbsp;<br>
QNSLP and QSPEC to allow multiple QoS signaling schemes to masquerade&nbsp;<br>
as a single architecture.<br>
  <br>
I really hope it's (a) or (b) since then we can probably resolve this&nbsp;<br>
by either putting back the missing functionality or figuring out&nbsp;<br>
another simple way to accommodate RMD. What I'm most afraid of is&nbsp;<br>
that this is really (c).<br>
  <br>
&nbsp;From the little I understand RMD, it's really like mapping onto a&nbsp;<br>
custom L2 QoS scheme like ATM or DOCSIS, and that rather than doing&nbsp;<br>
the mapping the way any conventional L3 RMF would do it, you co-<br>
mingle RMD's model with NSIS's model.<br>
  <br>
But, I could of course be wildly wrong here.<br>
  <br>
Dave.<br>
  <br>
  <br>
On Nov 2, 2006, at 6:00 AM, Georgios Karagiannis wrote:<br>
  <br>
&gt; Hi John<br>
&gt;<br>
&gt; I have just come back from a trip and I have just look into point
4):<br>
&gt;&gt;&gt; 4. Since remapping is eliminated, the concept of stacking
QSPECs is<br>
&gt;&gt;&gt; also eliminated.&nbsp; Discussion of tunneling of QNSLP
messages is<br>
&gt;&gt;&gt; eliminated since there are no impacts on tunneling
required by&nbsp;<br>
&gt;&gt;&gt; QSPEC.<br>
&gt;<br>
&gt; QSPEC stacking is one of the significant concepts in the QoS based&nbsp;<br>
&gt; NSIS<br>
&gt; solution.<br>
&gt; This concept provides the additional flexibility that is missing&nbsp;<br>
&gt; from RSVP.<br>
&gt; Furthermore, all the work that we have done until now is based on<br>
&gt; the concept of QSPEC stacking. Thus what you actually say is to
just<br>
&gt; forget the work that we have done<br>
&gt; until now, in e.g.the RMD-QOSM draft, and start from the beginning.<br>
&gt;<br>
&gt;<br>
&gt; The main problem that we have had until now was not the QSPEC
stacking<br>
&gt; issue,<br>
&gt; but it was the huge number of mandatory paremeters. Actualy I&nbsp;<br>
&gt; consider the<br>
&gt; issue of removing the<br>
&gt; QSPEC stacking concept a much more severe problem on achieving the<br>
&gt; flexibility<br>
&gt; that we wanted to achieve with NSIS than the huge number of
mandatory<br>
&gt; parameters.<br>
&gt;<br>
&gt; Therefore, I am severely against removing the QSPEC stacking
concept!<br>
&gt;<br>
&gt; Best Regards,<br>
&gt; Georgios<br>
&gt;<br>
&gt;<br>
&gt;<br>
&gt;<br>
&gt;<br>
&gt;&gt; -----Original Message-----<br>
&gt;&gt; From: <a class="moz-txt-link-abbreviated" href="mailto:john.loughney@nokia.com">john.loughney@nokia.com</a> [<a
 href="mailto:john.loughney@nokia.com">mailto:john.loughney@nokia.com</a>]<br>
&gt;&gt; Sent: donderdag 2 november 2006 8:35<br>
&gt;&gt; To: <a class="moz-txt-link-abbreviated" href="mailto:gash@att.com">gash@att.com</a>; <a class="moz-txt-link-abbreviated" href="mailto:nsis@ietf.org">nsis@ietf.org</a><br>
&gt;&gt; Cc: <a class="moz-txt-link-abbreviated" href="mailto:cornelia.kappler@siemens.com">cornelia.kappler@siemens.com</a>; <a class="moz-txt-link-abbreviated" href="mailto:oran@cisco.com">oran@cisco.com</a>;<br>
&gt;&gt; <a class="moz-txt-link-abbreviated" href="mailto:cornelia.kappler@gmx.net">cornelia.kappler@gmx.net</a><br>
&gt;&gt; Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft<br>
&gt;&gt;<br>
&gt;&gt; Jerry, Dave and Cornelia;<br>
&gt;&gt;<br>
&gt;&gt; This looks like a very good simplification of the QSPEC, and<br>
&gt;&gt; I think that is a good idea.&nbsp; I think that there will be some<br>
&gt;&gt; implications to the QoSM drafts - I think QoS Model drafts<br>
&gt;&gt; will need to contain some more applicability statement type<br>
&gt;&gt; of text, along with deployment considerations.<br>
&gt;&gt;<br>
&gt;&gt; A plus side of this is that I think that the QoS template<br>
&gt;&gt; will become more broadly applicable to other uses than just<br>
&gt;&gt; NSIS QoS Signaling, so I think that is a gain as well.<br>
&gt;&gt;<br>
&gt;&gt; I'd really like people to review this and send comments on it<br>
&gt;&gt; so that we could really focus the discussion next week and<br>
&gt;&gt; get this work finally resolved.<br>
&gt;&gt;<br>
&gt;&gt; Thanks a lot for working through this, I think this is a good<br>
&gt;&gt; basis for a proposal.<br>
&gt;&gt;<br>
&gt;&gt; thanks,<br>
&gt;&gt; John<br>
&gt;&gt;<br>
&gt;&gt;&gt; -----Original Message-----<br>
&gt;&gt;&gt; From: ext Ash, Gerald R (Jerry), ALABS [<a
 href="mailto:gash@att.com">mailto:gash@att.com</a>]<br>
&gt;&gt;&gt; Sent: 02 November, 2006 00:13<br>
&gt;&gt;&gt; To: <a class="moz-txt-link-abbreviated" href="mailto:nsis@ietf.org">nsis@ietf.org</a><br>
&gt;&gt;&gt; Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David
R Oran;<br>
&gt;&gt;&gt; Kappler,Cornelia<br>
&gt;&gt;&gt; Subject: [NSIS] Proposed Updates to the QSPEC Draft<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Hi All,<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Per John's suggestion below we've discussed Dave Oran's<br>
&gt;&gt; suggestions at<br>
&gt;&gt;&gt; length and have the following proposal.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; We think the proposed changes address many concerns raised<br>
&gt;&gt; on the list<br>
&gt;&gt;&gt; over the last few months and will require some rather basic<br>
&gt;&gt; changes in<br>
&gt;&gt;&gt; the QSPEC document<br>
&gt;&gt;&gt; <a
 href="http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt">http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt</a>.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; We appreciate Dave's suggestions and careful review of the
document.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Here is a summary of the proposed changes:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 1. The notion of QOSMs will be played down in the updated
draft.<br>
&gt;&gt;&gt; Language to the effect that 'QNSLP/QSPEC can signal for<br>
&gt;&gt; different QOSMs<br>
&gt;&gt;&gt; across multiple domains' will be replaced by the notion
that<br>
&gt;&gt;&gt; QNSLP/QSPEC allows different elements on the path to
implement<br>
&gt;&gt;&gt; different data plane QoS mechanisms that meets the<br>
&gt;&gt; constraints included<br>
&gt;&gt;&gt; in the QSPEC.&nbsp; A QOSM describes common capabilities among<br>
&gt;&gt; QNEs to act<br>
&gt;&gt;&gt; consistently when requested to admit traffic and in
treating<br>
&gt;&gt; admitted<br>
&gt;&gt;&gt; traffic.<br>
&gt;&gt;&gt; A QOSM ID need not be defined or signaled and a QNE need
not support<br>
&gt;&gt;&gt; any particular QOSM.&nbsp; A 'QOSM specification' would become a<br>
&gt;&gt;&gt; 'QNSLP/QSPEC applicability statement', which documents how
a QNE/RMF<br>
&gt;&gt;&gt; interprets and treats the various elements in a QSPEC.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; An updated QOSM definition is as follows:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; QoS Model (QOSM): A method to achieve QoS for a traffic
flow, e.g.,<br>
&gt;&gt;&gt; IntServ Controlled Load.&nbsp; A QOSM specifies what sub-set of
QSPEC QoS<br>
&gt;&gt;&gt; constraints and traffic handling directives a QNE
implementing that<br>
&gt;&gt;&gt; QOSM is capable of supporting and how resources will be<br>
&gt;&gt; managed by the<br>
&gt;&gt;&gt; RMF in such a QNE.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 2. The QSPEC1/QSPEC2 semantics will be replaced with
Dave's a) --<br>
&gt;&gt;&gt; &gt; d)<br>
&gt;&gt;&gt; semantics:<br>
&gt;&gt;&gt; A) The source traffic description is mandatory to include
by the QNI<br>
&gt;&gt;&gt; and mandatory to interpret by downstream QNEs.&nbsp; The traffic<br>
&gt;&gt; description<br>
&gt;&gt;&gt; is specified by a traffic model (TMOD) parameter
consisting of 4<br>
&gt;&gt;&gt; sub-parameters:<br>
&gt;&gt;&gt; - rate (r)<br>
&gt;&gt;&gt; - bucket size (b)<br>
&gt;&gt;&gt; - peak rate (p)<br>
&gt;&gt;&gt; - minimum policed unit (m)<br>
&gt;&gt;&gt; The TMOD parameter is a mathematically complete way to
describe a<br>
&gt;&gt;&gt; traffic source.&nbsp; Examples will be given for setting TMOD to<br>
&gt;&gt; correspond<br>
&gt;&gt;&gt; to specific cases, e.g. where only bandwidth is available<br>
&gt;&gt; (r=p; b and m<br>
&gt;&gt;&gt; set to default values) and other cases (e.g., TCP).&nbsp; As
such, the<br>
&gt;&gt;&gt; bandwidth parameter is eliminated.<br>
&gt;&gt;&gt; B) All other QSPEC parameters are optional to include by
the<br>
&gt;&gt; QNI, and<br>
&gt;&gt;&gt; may be either mandatory or optional to interpret by
downstream QNEs.<br>
&gt;&gt;&gt; That is, the QNI can explicitly specify for each constraint<br>
&gt;&gt; parameter<br>
&gt;&gt;&gt; whether it wishes admission control to succeed or fail if
that<br>
&gt;&gt;&gt; constraint cannot be met.&nbsp; A downstream QNE distinguishes<br>
&gt;&gt; whether it is<br>
&gt;&gt;&gt; failure to understand the parameter or failure to meet the<br>
&gt;&gt; constraint<br>
&gt;&gt;&gt; that causes the admission control failure (more detail on<br>
&gt;&gt; this below).<br>
&gt;&gt;&gt; C) Additional QSPEC parameters can be defined in updates to<br>
&gt;&gt; the QSPEC<br>
&gt;&gt;&gt; document (or alternatively specified in QOSM applicability
statement<br>
&gt;&gt;&gt; documents).&nbsp; There is already an IANA registry established
for<br>
&gt;&gt;&gt; additional QSPEC parameters.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 3. The concept of remapping QSPEC parameters is eliminated.<br>
&gt;&gt;&gt; This concept is ill defined and has caused much confusion
(there are<br>
&gt;&gt;&gt; few if any sensible remappings).&nbsp; Rather we define<br>
&gt;&gt; 'interpret' a QSPEC<br>
&gt;&gt;&gt; parameter to only mean what we now call 'strictly
interpret' (i.e.,<br>
&gt;&gt;&gt; must conform to the RFCs defining the parameter and
procedures) and<br>
&gt;&gt;&gt; eliminate the remap option part of 'interpret'.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 4. Since remapping is eliminated, the concept of stacking
QSPECs is<br>
&gt;&gt;&gt; also eliminated.&nbsp; Discussion of tunneling of QNSLP
messages is<br>
&gt;&gt;&gt; eliminated since there are no impacts on tunneling
required by&nbsp;<br>
&gt;&gt;&gt; QSPEC.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 5. Current QSPEC flags are modified as follows:<br>
&gt;&gt;&gt; A) The QNI sets the M flag for each QSPEC parameter it<br>
&gt;&gt; populates that<br>
&gt;&gt;&gt; it intends must be interpreted or the reservation fails.<br>
&gt;&gt; Currently the<br>
&gt;&gt;&gt; M flag is statically set for every QSPEC parameter.<br>
&gt;&gt;&gt; B) If the M flag is set, a downstream QNE must interpret
the<br>
&gt;&gt; parameter<br>
&gt;&gt;&gt; or the reservation fails.&nbsp; If a QNE does not support the<br>
&gt;&gt; parameter with<br>
&gt;&gt;&gt; the M flag set, it sets the N flag and rejects the<br>
&gt;&gt; reservation.&nbsp; If the<br>
&gt;&gt;&gt; QNE supports the parameter but cannot meet the parameter,
it<br>
&gt;&gt; sets the E<br>
&gt;&gt;&gt; flag and rejects the reservation.<br>
&gt;&gt;&gt; C) If the M flag is not set, a downstream QNE should
interpret the<br>
&gt;&gt;&gt; parameter.&nbsp; If a QNE does not support the parameter, it
sets<br>
&gt;&gt; the N flag<br>
&gt;&gt;&gt; and can optionally accept or reject the reservation.&nbsp; If a<br>
&gt;&gt; QNE supports<br>
&gt;&gt;&gt; the parameter but cannot meet the parameter, it sets the E
flag and<br>
&gt;&gt;&gt; optional accepts or rejects the reservation.<br>
&gt;&gt;&gt; D) The R (remapped parameter) flag and Q (non QOSM) flag
are<br>
&gt;&gt;&gt; eliminated.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 6. Summary of proposed QSPEC parameters:<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Source traffic Description:<br>
&gt;&gt;&gt; TMOD-1 (mandatory to include)<br>
&gt;&gt;&gt; TMOD-2 (optional to include)<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Constraints (optional to include):<br>
&gt;&gt;&gt; Path Latency<br>
&gt;&gt;&gt; Path Jitter<br>
&gt;&gt;&gt; Path PLR<br>
&gt;&gt;&gt; Path PER<br>
&gt;&gt;&gt; Y.1541 class (shorthand notation for above 4 parameters)<br>
&gt;&gt; Slack term<br>
&gt;&gt;&gt; Priority (Preemption, Defending, Admission, RPH<br>
&gt;&gt;&gt; Priority)&nbsp; DSTE class<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Handling directives (optional to include):<br>
&gt;&gt;&gt; excess treatment<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Traffic Classifiers (optional to include):<br>
&gt;&gt;&gt; PHB Class (note that PHB class is set by downstream QNE to
tell QNI<br>
&gt;&gt;&gt; how to mark traffic to ensure it will be treated as
committed by<br>
&gt;&gt;&gt; admission<br>
&gt;&gt;&gt; control)<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Eliminated:<br>
&gt;&gt;&gt; bandwidth<br>
&gt;&gt;&gt; Ctot, Dtot, Csum, Dsum<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; 7. Editorial changes:<br>
&gt;&gt;&gt; A) rewrite description of read-only and read-write<br>
&gt;&gt; parameters to make<br>
&gt;&gt;&gt; the rules more clear:<br>
&gt;&gt;&gt; - which parameters can be ro/rw (function of object in
which<br>
&gt;&gt; parameter<br>
&gt;&gt;&gt; included)<br>
&gt;&gt;&gt; - can only increase rw values and not delete, etc.<br>
&gt;&gt;&gt; B) eliminate the distinction of 'control' and 'qos
description'<br>
&gt;&gt;&gt; parameters and replace with suggested taxonomy, as
described above<br>
&gt;&gt;&gt; C) clarify discussion of packet outcomes from Y.1540 and
place in<br>
&gt;&gt;&gt; appropriate sections<br>
&gt;&gt;&gt; D) clarify which parameters can be included in &lt;QoS
Desired&gt; &lt;QoS<br>
&gt;&gt;&gt; Available&gt;, &lt;QoS Reserved&gt;, or &lt;QoS Minimum&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Please review and comment on the proposed changes.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Thanks,<br>
&gt;&gt;&gt; Jerry, Dave, Cornelia<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; -----Original Message-----<br>
&gt;&gt;&gt; From: <a class="moz-txt-link-abbreviated" href="mailto:john.loughney@nokia.com">john.loughney@nokia.com</a> [<a
 href="mailto:john.loughney@nokia.com">mailto:john.loughney@nokia.com</a>]<br>
&gt;&gt;&gt; Sent: Monday, October 23, 2006 2:59 AM<br>
&gt;&gt;&gt; To: <a class="moz-txt-link-abbreviated" href="mailto:nsis@ietf.org">nsis@ietf.org</a><br>
&gt;&gt;&gt; Subject: [NSIS] Where we stand with the
draft-ietf-nsis-qspec draft<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Hi all,<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; We have had a very large thread on the current<br>
&gt;&gt;&gt; draft-ietf-nsis-qspec-12.txt draft.&nbsp; We don't have a clear<br>
&gt;&gt; consensus,<br>
&gt;&gt;&gt; so I'd like to try to summarize the discussions.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; Several people feel that the current draft contains too
many<br>
&gt;&gt; mandatory<br>
&gt;&gt;&gt; parameters and mapping some QoS Models to others is mixing<br>
&gt;&gt;&gt; apples-to-oranges.&nbsp; The feeling is that interoperability
is hard to<br>
&gt;&gt;&gt; achieve and therefore we should not try to gain
interoperability.<br>
&gt;&gt;&gt; Jukka Manner's message on 17th of October proposed a way
forward.<br>
&gt;&gt;&gt; <a
 href="http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html">http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; However, several (other) people feel that interoperability<br>
&gt;&gt; is important<br>
&gt;&gt;&gt; and is needed. Giving up on interoperability would not be a<br>
&gt;&gt; good thing,<br>
&gt;&gt;&gt; and that we are very close to a solution.<br>
&gt;&gt;&gt; Dave Oran proposed a mechanism to reach a solution on the
19th<br>
&gt;&gt;&gt; October:<br>
&gt;&gt;&gt; <a
 href="http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html">http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; I have asked Jerry and Dave to make a proposal along the
lines of<br>
&gt;&gt;&gt; Dave's message above, and the working group can review
this,<br>
&gt;&gt; to see if<br>
&gt;&gt;&gt; we can gain consensus on this.&nbsp; I think it would be<br>
&gt;&gt; sufficient to send<br>
&gt;&gt;&gt; some rough text to the mailing list, and if this is OK,
then we can<br>
&gt;&gt;&gt; revise the draft.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; One open question to Dave/Jerry about the proposal is, in
the event<br>
&gt;&gt;&gt; that the source cannot be described a mathematically<br>
&gt;&gt; complete way, is<br>
&gt;&gt;&gt; there any value (or does it make any sense at<br>
&gt;&gt;&gt; all) in sending an incomplete qspec, or potentially
"wildcarding"<br>
&gt;&gt;&gt; sending some of the mandatory parameters.&nbsp; I'm wondering
if some<br>
&gt;&gt;&gt; sources / deployments may have certain parameters that are<br>
&gt;&gt; relatively<br>
&gt;&gt;&gt; well defined, but not all are so well defined.<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; thanks,<br>
&gt;&gt;&gt; John<br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; _______________________________________________<br>
&gt;&gt;&gt; nsis mailing list<br>
&gt;&gt;&gt; <a class="moz-txt-link-abbreviated" href="mailto:nsis@ietf.org">nsis@ietf.org</a><br>
&gt;&gt;&gt; <a href="https://www1.ietf.org/mailman/listinfo/nsis">https://www1.ietf.org/mailman/listinfo/nsis</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;&gt; _______________________________________________<br>
&gt;&gt;&gt; nsis mailing list<br>
&gt;&gt;&gt; <a class="moz-txt-link-abbreviated" href="mailto:nsis@ietf.org">nsis@ietf.org</a><br>
&gt;&gt;&gt; <a href="https://www1.ietf.org/mailman/listinfo/nsis">https://www1.ietf.org/mailman/listinfo/nsis</a><br>
&gt;&gt;&gt;<br>
&gt;&gt;<br>
&gt;&gt; _______________________________________________<br>
&gt;&gt; nsis mailing list<br>
&gt;&gt; <a class="moz-txt-link-abbreviated" href="mailto:nsis@ietf.org">nsis@ietf.org</a><br>
&gt;&gt; <a href="https://www1.ietf.org/mailman/listinfo/nsis">https://www1.ietf.org/mailman/listinfo/nsis</a><br>
&gt;&gt;<br>
&gt;<br>
&gt;<br>
&gt;<br>
&gt; _______________________________________________<br>
&gt; nsis mailing list<br>
&gt; <a class="moz-txt-link-abbreviated" href="mailto:nsis@ietf.org">nsis@ietf.org</a><br>
&gt; <a href="https://www1.ietf.org/mailman/listinfo/nsis">https://www1.ietf.org/mailman/listinfo/nsis</a><br>
  </font>
  </p>
</blockquote>
</body>
</html>

--------------010506000106020009040200--



--===============0146235622==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============0146235622==--





From nsis-bounces@ietf.org Fri Nov 03 05:35:49 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwO0-0003dK-VD; Fri, 03 Nov 2006 05:34:56 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwNz-0003dC-S8
	for nsis@ietf.org; Fri, 03 Nov 2006 05:34:55 -0500
Received: from deprox.docomolab-euro.com ([212.119.9.186])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfwNy-0007GS-7z
	for nsis@ietf.org; Fri, 03 Nov 2006 05:34:55 -0500
Received: from 172.27.10.3 by deprox.docomolab-euro.com (InterScan E-Mail
	VirusWall NT); Fri, 03 Nov 2006 11:34:28 +0100
Received: from [172.27.100.96] ([172.27.100.96]) by deex.docomolab-euro.com
	with Microsoft SMTPSVC(5.0.2195.6713); 
	Fri, 3 Nov 2006 11:34:27 +0100
Message-ID: <454B1B33.1060801@docomolab-euro.com>
Date: Fri, 03 Nov 2006 11:34:27 +0100
From: mendes <mendes@docomolab-euro.com>
User-Agent: Thunderbird 1.5.0.7 (Windows/20060909)
MIME-Version: 1.0
To: jian.zhang@dcs.qmul.ac.uk
Subject: Re: [NSIS] The version 3 of the InterDomain-QOSM Draft: the NSIS
	QOS Model to Fulfill the E2E QoS Control in the ITU-T RACF Functional
	Architecture
References: <BAA65A575825454CBB0103267553FCCCED0466@esebe199.NOE.Nokia.com>
	<4549CACD.2000209@dcs.qmul.ac.uk>
In-Reply-To: <4549CACD.2000209@dcs.qmul.ac.uk>
X-Enigmail-Version: 0.94.0.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 03 Nov 2006 10:34:27.0149 (UTC)
	FILETIME=[A31AB7D0:01C6FF33]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a2c12dacc0736f14d6b540e805505a86
Cc: john.loughney@nokia.com, edmundo@dei.uc.pt, karagian@cs.utwente.nl,
	nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi,

i'm not directly involved in the ITU-T RACF activities.

Paulo

Jian Zhang wrote:
> Hello John,
>
> Currently I have not involved in the ITU-T RACF activities and I am
> not aware if Paulo and Jorge are involving with these activities
> or not, but I believe the EU IST FP6 EuQoS project, as a whole, is
> involved in the discussions (Edmundo knows better about that).
>
> Meanwhile, Georgios will discuss with the ITU-T people about these
> issues during the coming IETF meeting.
>
> Cheers,
>
> Jian
>
> john.loughney@nokia.com wrote:
>> Jian,
>>
>> Are any of the authors participating in the ITU-T RACF discussions?
>>
>> thanks,
>> John
>>  
>>> -----Original Message-----
>>> From: ext Jian Zhang [mailto:jian.zhang@dcs.qmul.ac.uk] Sent: 31
>>> October, 2006 17:18
>>> To: nsis@ietf.org
>>> Cc: edmundo@dei.uc.pt; karagian@cs.utwente.nl;
>>> mendes@docomolab-euro.com
>>> Subject: [NSIS] The version 3 of the InterDomain-QOSM Draft: the
>>> NSIS QOS Model to Fulfill the E2E QoS Control in the ITU-T RACF
>>> Functional Architecture
>>>
>>> Dear all,
>>>
>>> Please kindly read the version 3 of the InterDomain-QOSM Draft: our
>>> effort to define a NSIS QOS Model to fulfill the E2E QoS control in
>>> the ITU-T RACF functional architecture, which has been submitted to
>>> the NSIS WG for discussion at the coming IETF meeting.
>>>
>>> And if you have any questions or suggestions about this draft, we
>>> would like to answer or discuss with you in this mailing list. The
>>> abstract of this draft is attached below and the link which holds
>>> the whole contents of the draft is also provided below.
>>>
>>> Cheers,
>>>
>>> Jian Zhang
>>>
>>> Dept. of Computer Science
>>> Queen Mary, Univ. of London
>>>
>>> ---------------------------------------
>>>
>>> Abstract
>>>
>>>   This document has three goals. First of all, it presents our analysis
>>>   of how to use the NSIS signaling (inter-domain QOSM and intra-domain
>>>   QOSM) to fulfill the QoS control in accord with the ITU-T RACF
>>>   functional architecture. For this goal, we discuss how the ITU-T RACF
>>>   entities in the ITU-T RACF functional architecture can be mapped to
>>>   the NSIS entities and how the RACF reference points can be
>>>   implemented by using the NSIS protocol suites and QOSMs. Secondly, we
>>>   aim at specifying an NSIS Inter-domain QOSM for E2E QoS control
>>>   across heterogeneous IP networks and applying this Inter-domain QOSM
>>>   to the e2e QoS control in the ITU-T RACF functional architecture
>>>   based on the above ITU-T RACF analysis. The detailed description of
>>>   the NSIS Inter-domain QOSM are given and the e2e QoS control
>>>   scenarios in the ITU-T RACF architecture (including RACF Push and
>>>   Pull resource control modes), which will be covered by the NSIS
>>>   Inter-domain QOSM are described and implemented. Thirdly, we specify
>>>   and implement those QSPECs that will be used by the Inter-domain QOSM
>>>   for the e2e QoS control in the ITU-T RACF architecture.
>>>
>>> The link is
>>> http://ietfreport.isoc.org/ids/draft-zhang-nsis-interdomain-qos
>>> m-03.txt .
>>>
>>>
>>> _______________________________________________
>>> nsis mailing list
>>> nsis@ietf.org
>>> https://www1.ietf.org/mailman/listinfo/nsis
>>>
>>>     
>
>


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 05:39:40 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwRz-0003wy-Ck; Fri, 03 Nov 2006 05:39:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwRy-0003ws-Pk
	for nsis@ietf.org; Fri, 03 Nov 2006 05:39:02 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfwRw-0007wf-Ec
	for nsis@ietf.org; Fri, 03 Nov 2006 05:39:02 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-3.tower-146.messagelabs.com!1162550339!5960043!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 10074 invoked from network); 3 Nov 2006 10:38:59 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-3.tower-146.messagelabs.com with SMTP;
	3 Nov 2006 10:38:59 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA3AcxlS017194; 
	Fri, 3 Nov 2006 05:38:59 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA3AcsKT017173; Fri, 3 Nov 2006 05:38:55 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 04:38:54 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F1D@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <hvOvd46k.1162535221.3187200.karagian@ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQ
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "Georgios Karagiannis" <karagian@cs.utwente.nl>, <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d2e37451f7f22841e3b6f40c67db0f
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios,=20

> -----Original Message-----
> From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]=20
> Sent: Friday, November 03, 2006 1:27 AM
> To: nsis@ietf.org
> Subject: [NSIS] jerry, dave,cornelia's proposal very=20
> confusing and not solving problem
>=20
> Hi all
>=20
> After rereading the proposal of Jerry, Dave, Cornelia I have=20
> seen that the
> main problem that we have had until now in the mailing=20
> discussions is not
> solved. Mainly the number of mandatory parameters to interpret by all
> the QNEs in the universe is not solved.

It is solved.  There is only one mandatory parameter that must be
interpreted by all QNE/RMFs (traffic description =3D TMOD).  All other
parameters are optional to interpret and can be ignored by a QNE/RMF.
I.e., there is only one 'qspec1' parameter (although qspec1/qspec2
terminology is no longer used).
=20
> The proposal mentions in point 2) point B):
>=20
> "All other QSPEC parameters are optional to include by the=20
> QNI, and may
> be either mandatory or optional to interpret by downstream QNEs."
>=20
> Meaning that if a QNI includes the Y.1541 class and wants that all
> downstream QNEs interpret this parameter, then all downstream=20
> QNEs must
> be able to strictly interpret the Y.1541 class.
> Jerry, Dave, Cornelia is this true?

No.  If the Y.1541 parameter is populated by the QNI and the QNE does
not support the Y.1541 parameter then:
A) if the M flag is set by the QNI for the Y.1541 parameter, the QNE
sets the N flag for the Y.1541 parameter and rejects the reservation
B) if the M flag is not set by the QNI for the Y.1541 parameter, the QNE
sets the N flag for the Y.1541 parameter and optionally either accepts
or rejects the reservation (i.e., the QNE can simply ignore the Y.1541
parameter).
=20
> Meaning that all QNEs that want to use NSIS must be=20
> implemented in such a
> way that if a QNI mandates the use of a parameter, e.g., Y.1541, then
> the QNE must strictly interpret it, otherwise it has to reject the
> parameter or the admission request.
>=20
> But this is exactly the problem that we wanted to solve by having this
> discussions until now. Thus the problem is not solved.

I disagree.  All parameters except TMOD can be *not-supported* by a QNE.
If a QNI truly wants according to its QoS procedure being implemented
that a parameter must be interpreted or rejected, then no other outcome
makes sense other than to reject the reservation if a parameter is not
supported.  The QNI governs the QoS outcome of the reservation.

The proposal is also similar to an earlier suggestion (by Roland?) that
a QNI can decide which parameters are mandatory to interpret.
=20
> Additionally, Jerry, Dave and Cornelia change the main concepts of QoS
> based NSIS, please see my previous e-mails that I have sent=20
> yesterday on
> this.
>=20
> I have already proposed a solution to solve the problems that=20
> we have had
> in the nsis mailing list, see a copy of that proposal below:
> -------------------------
>=20
> Another proposal for solving the interoperability issue and=20
> the issue of
> having too many parameters was to:
> * use only one mandatory parameter, i.e., Token Bucket that could be
> remapped into the peak Bandwidth parameter.
> Thus except the Token Bucket parameter,
> * change the QSPEC-1 parameters into QSPEC-2 parameters

It is equivalent.  If you change Y.1541 to qspec2 and a QNE doesn't
support Y.1541, it sets the N flag and ignores the parameter.  This is
the same as the proposal as long as the QNI doesn't care whether Y.1541
is supported.  However, if the QNI does care that Y.1541 is supported,
it will reject/tear the reservation if the N flag is set.  The result is
the same as the proposal!

Jerry

> * describe the remapping procedure of the Token Bucket into the peak
> Bandwidth parameter in the QSPEC template draft.
>=20
> After that we can submit the QSPEC template draft for last WG call. If
> after the deployment phase we conclude that more mandatory parameters
> are needed for interoperability then we could work on an interworking
> draft.
>=20
> --------------
>=20
> Best Regards,
> Georgios
>=20
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 05:41:35 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwTk-0004IC-BL; Fri, 03 Nov 2006 05:40:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwTi-0004Hi-Nd
	for nsis@ietf.org; Fri, 03 Nov 2006 05:40:50 -0500
Received: from mail121.messagelabs.com ([216.82.241.195])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfwTh-0008Ga-D8
	for nsis@ietf.org; Fri, 03 Nov 2006 05:40:50 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-8.tower-121.messagelabs.com!1162550448!10532183!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 29339 invoked from network); 3 Nov 2006 10:40:48 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-8.tower-121.messagelabs.com with SMTP;
	3 Nov 2006 10:40:48 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA3Aelb6017559; 
	Fri, 3 Nov 2006 05:40:48 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA3Aehdo017538; Fri, 3 Nov 2006 05:40:43 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 04:40:43 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F1E@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <Pine.LNX.4.64.0611031047160.7180@sbz-31.cs.Helsinki.FI>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Thread-Index: Acb/JhvMK3QwbBPGRXOnpodalRM/qQACsQQA
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "Jukka MJ Manner" <jmanner@cs.Helsinki.FI>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 5d7a7e767f20255fce80fa0b77fb2433
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Jukka,=20

> -----Original Message-----
> From: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI]=20
> Sent: Friday, November 03, 2006 3:55 AM
> To: Georgios Karagiannis
> Cc: nsis@ietf.org
> Subject: Re: [NSIS] jerry, dave, cornelia's proposal very=20
> confusing and not solving problem
>=20
>=20
> Hi,
>=20
> AFAICT, if the QNI _really_ wants to have an Y.1541 class supported=20
> (mandatory) and a QNE on the path does not support that, the=20
> reservation=20
> will be rejected. Tough luck for the QNI. This forces QNIs to=20
> be quite=20
> conservative in what they ask to be "mandatory" on the path,=20
> and suggests=20
> that all parameters in category B should in fact be optional.
>=20
> If a QNI sets too many, and too complex mandatory objects,=20
> then it will=20
> not get an e2e reservation, simple.
>=20
> The RMD, or any other QoS spec, does not need to support all weird=20
> requests from QNIs.
>=20
> Is that what the JCD-team are proposing?

Yes, exactly, well phrased.  There is only one mandatory parameter that
all QNE/RMFs must support and are required to interpret (traffic
description =3D TMOD).

Thanks,
Regards,
Jerry

>=20
> Jukka
>=20
> On Fri, 3 Nov 2006, Georgios Karagiannis wrote:
>=20
> > Hi all
> >=20
> > After rereading the proposal of Jerry, Dave, Cornelia I=20
> have seen that the
> > main problem that we have had until now in the mailing=20
> discussions is not
> > solved. Mainly the number of mandatory parameters to=20
> interpret by all
> > the QNEs in the universe is not solved.
> >=20
> > The proposal mentions in point 2) point B):
> >=20
> > "All other QSPEC parameters are optional to include by the=20
> QNI, and may
> > be either mandatory or optional to interpret by downstream QNEs."
> >=20
> > Meaning that if a QNI includes the Y.1541 class and wants that all
> > downstream QNEs interpret this parameter, then all=20
> downstream QNEs must
> > be able to strictly interpret the Y.1541 class.
> > Jerry, Dave, Cornelia is this true?
> >=20
> > Meaning that all QNEs that want to use NSIS must be=20
> implemented in such a
> > way that if a QNI mandates the use of a parameter, e.g.,=20
> Y.1541, then
> > the QNE must strictly interpret it, otherwise it has to reject the
> > parameter or the admission request.
> >=20
> > But this is exactly the problem that we wanted to solve by=20
> having this
> > discussions until now. Thus the problem is not solved.
> >=20
> > Additionally, Jerry, Dave and Cornelia change the main=20
> concepts of QoS
> > based NSIS, please see my previous e-mails that I have sent=20
> yesterday on
> > this.
> >=20
> > I have already proposed a solution to solve the problems=20
> that we have had
> > in the nsis mailing list, see a copy of that proposal below:
> > -------------------------
> >=20
> > Another proposal for solving the interoperability issue and=20
> the issue of
> > having too many parameters was to:
> > * use only one mandatory parameter, i.e., Token Bucket that could be
> > remapped into the peak Bandwidth parameter.
> > Thus except the Token Bucket parameter,
> > * change the QSPEC-1 parameters into QSPEC-2 parameters
> > * describe the remapping procedure of the Token Bucket into the peak
> > Bandwidth parameter in the QSPEC template draft.
> >=20
> > After that we can submit the QSPEC template draft for last=20
> WG call. If
> > after the deployment phase we conclude that more mandatory=20
> parameters
> > are needed for interoperability then we could work on an=20
> interworking
> > draft.
> >=20
> > --------------
> >=20
> > Best Regards,
> > Georgios
> >=20
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> >=20
>=20
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 05:42:48 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwUz-0004bs-JS; Fri, 03 Nov 2006 05:42:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwUy-0004bn-OQ
	for nsis@ietf.org; Fri, 03 Nov 2006 05:42:08 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfwUy-0008Va-87
	for nsis@ietf.org; Fri, 03 Nov 2006 05:42:08 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-8.tower-146.messagelabs.com!1162550527!7928661!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 30150 invoked from network); 3 Nov 2006 10:42:07 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-8.tower-146.messagelabs.com with SMTP;
	3 Nov 2006 10:42:07 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA3Ag7Ol017762; 
	Fri, 3 Nov 2006 05:42:07 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA3Ag1iN017729; Fri, 3 Nov 2006 05:42:01 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 04:42:00 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F1F@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <Pine.LNX.4.64.0611031144540.7180@sbz-31.cs.Helsinki.FI>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1A
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "Jukka MJ Manner" <jmanner@cs.Helsinki.FI>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2d133cc328f58695161c98bb4f4dc213
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Jukka,=20

> -----Original Message-----
> From: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI]=20
> Sent: Friday, November 03, 2006 4:48 AM
> To: Attila B=E1der (IJ/ETH)
> Cc: Ash, Gerald R (Jerry), NPE; nsis@ietf.org; Kappler,=20
> Cornelia; David R Oran; Cornelia Kappler
> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
>=20
>=20
> Hi all,
>=20
> I tend to agree with Attila here that stacking is a new=20
> concept and is=20
> important to have. E.g., it allows a domain to add QoS=20
> information that is=20
> only relevant to the domain itself, an ingress QNE pushes a=20
> new QSPEC into=20
> the NSLP message, and the egress QNE of the domain just pops=20
> it out. A=20
> domain isn't required to do so, but allows new kinds of QoS=20
> management options.

Without remapping of parameters being allowed, I don't know what a =
stacked QSPEC would do?  Let's take an example, e.g., the QNI sets
TMOD (M flag set)
Parm A (M flag set)
Parm B (M flag not set)

In the local domain the edge QNEs must interpret TMOD and must interpret =
Parm B, otherwise the reservation is rejected.  Parm B can either be =
supported or ignored by the edge QNEs.

What might a stacked qspec accomplish differently in the local domain?  =
Perhaps a new Parm C is pushed onto the local QSPEC, or what?  And if =
so, what would that accomplish?

IMO we need a concrete example of how stacking would be useful if =
remapping is not allowed.  All our prior examples of stacking involved =
some kind of remapping of parameters from the initiator qspec to a local =
qspec.  Without that, I don't see what a stacked qspec accomplishes.

Dave also asked for a concrete example of how stacking is useful (when =
remapping is not possible) in one of his posts yesterday.

Thanks,
Jerry

P.S. Note that the only example of remapping that we ever discussed was =
token bucket to bandwidth.  All other examples of remapping were trying =
to map apples to oranges, and causing great confusion.  The concept of =
remapping is rightfully rejected.

>=20
> Jukka
>=20
> On Fri, 3 Nov 2006, Attila B=E1der (IJ/ETH) wrote:
>=20
> > Hi All, I disagree with the proposal below.=20
> >=20
> > This would remove the flexibility to develop and define new=20
> QoS models.=20
> > Actually the proposal supports only the Intserv model. Stacking and=20
> > remapping have been fundamental features.
> >=20
> > I think that the best solution would be not to mandate=20
> parameters and=20
> > have only optional one. If we want to distinguish Intserv=20
> as e2e QoS=20
> > model, it is OK to mandate token bucket parameters, but=20
> remapping it to=20
> > other parameters and stacking it should be allowed.
> >=20
> > Best regards, Attila
> >=20
> >=20
> > >-----Original Message-----
> > >From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]=20
> > >Sent: Wednesday, November 01, 2006 11:13 PM
> > >To: nsis@ietf.org
> > >Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R=20
> > >Oran; Kappler,Cornelia
> > >Subject: [NSIS] Proposed Updates to the QSPEC Draft
> > >
> > >Hi All,
> > >
> > >Per John's suggestion below we've discussed Dave Oran's=20
> > >suggestions at length and have the following proposal. =20
> > >
> > >We think the proposed changes address many concerns raised on=20
> > >the list over the last few months and will require some rather=20
> > >basic changes in the QSPEC document=20
> > >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> > >
> > >We appreciate Dave's suggestions and careful review of the=20
> document.
> > >
> > >Here is a summary of the proposed changes:
> > >
> > >1. The notion of QOSMs will be played down in the updated draft.
> > >Language to the effect that 'QNSLP/QSPEC can signal for=20
> > >different QOSMs across multiple domains' will be replaced by=20
> > >the notion that QNSLP/QSPEC allows different elements on the=20
> > >path to implement different data plane QoS mechanisms that=20
> > >meets the constraints included in the QSPEC.  A QOSM describes=20
> > >common capabilities among QNEs to act consistently when=20
> > >requested to admit traffic and in treating admitted traffic. =20
> > >A QOSM ID need not be defined or signaled and a QNE need not=20
> > >support any particular QOSM.  A 'QOSM specification' would=20
> > >become a 'QNSLP/QSPEC applicability statement', which=20
> > >documents how a QNE/RMF interprets and treats the various=20
> > >elements in a QSPEC. =20
> > >
> > >An updated QOSM definition is as follows:
> > >
> > >QoS Model (QOSM): A method to achieve QoS for a traffic flow,=20
> > >e.g., IntServ Controlled Load.  A QOSM specifies what sub-set=20
> > >of QSPEC QoS constraints and traffic handling directives a QNE=20
> > >implementing that QOSM is capable of supporting and how=20
> > >resources will be managed by the RMF in such a QNE.
> > >
> > >2. The QSPEC1/QSPEC2 semantics will be replaced with=20
> Dave's a) --> d)
> > >semantics:
> > >A) The source traffic description is mandatory to include by=20
> > >the QNI and mandatory to interpret by downstream QNEs.  The=20
> > >traffic description is specified by a traffic model (TMOD)=20
> > >parameter consisting of 4
> > >sub-parameters:
> > >- rate (r)
> > >- bucket size (b)
> > >- peak rate (p)
> > >- minimum policed unit (m)
> > >The TMOD parameter is a mathematically complete way to=20
> > >describe a traffic source.  Examples will be given for setting=20
> > >TMOD to correspond to specific cases, e.g. where only=20
> > >bandwidth is available (r=3Dp; b and m set to default values)=20
> > >and other cases (e.g., TCP).  As such, the bandwidth parameter=20
> > >is eliminated.
> > >B) All other QSPEC parameters are optional to include by the=20
> > >QNI, and may be either mandatory or optional to interpret by=20
> > >downstream QNEs.
> > >That is, the QNI can explicitly specify for each constraint=20
> > >parameter whether it wishes admission control to succeed or=20
> > >fail if that constraint cannot be met.  A downstream QNE=20
> > >distinguishes whether it is failure to understand the=20
> > >parameter or failure to meet the constraint that causes the=20
> > >admission control failure (more detail on this below).
> > >C) Additional QSPEC parameters can be defined in updates to=20
> > >the QSPEC document (or alternatively specified in QOSM=20
> > >applicability statement documents).  There is already an IANA=20
> > >registry established for additional QSPEC parameters.
> > >
> > >3. The concept of remapping QSPEC parameters is eliminated. =20
> > >This concept is ill defined and has caused much confusion=20
> > >(there are few if any sensible remappings).  Rather we define=20
> > >'interpret' a QSPEC parameter to only mean what we now call=20
> > >'strictly interpret' (i.e., must conform to the RFCs defining=20
> > >the parameter and procedures) and eliminate the remap option=20
> > >part of 'interpret'.
> > >
> > >4. Since remapping is eliminated, the concept of stacking=20
> > >QSPECs is also eliminated.  Discussion of tunneling of QNSLP=20
> > >messages is eliminated since there are no impacts on tunneling=20
> > >required by QSPEC.
> > >
> > >5. Current QSPEC flags are modified as follows:
> > >A) The QNI sets the M flag for each QSPEC parameter it=20
> > >populates that it intends must be interpreted or the=20
> > >reservation fails.  Currently the M flag is statically set for=20
> > >every QSPEC parameter.
> > >B) If the M flag is set, a downstream QNE must interpret the=20
> > >parameter or the reservation fails.  If a QNE does not support=20
> > >the parameter with the M flag set, it sets the N flag and=20
> > >rejects the reservation.  If the QNE supports the parameter=20
> > >but cannot meet the parameter, it sets the E flag and rejects=20
> > >the reservation.=20
> > >C) If the M flag is not set, a downstream QNE should interpret=20
> > >the parameter.  If a QNE does not support the parameter, it=20
> > >sets the N flag and can optionally accept or reject the=20
> > >reservation.  If a QNE supports the parameter but cannot meet=20
> > >the parameter, it sets the E flag and optional accepts or=20
> > >rejects the reservation.
> > >D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
> > >eliminated.
> > >
> > >6. Summary of proposed QSPEC parameters:
> > >
> > >Source traffic Description:
> > > TMOD-1 (mandatory to include)
> > > TMOD-2 (optional to include)
> > >
> > >Constraints (optional to include):
> > > Path Latency
> > > Path Jitter
> > > Path PLR
> > > Path PER
> > > Y.1541 class (shorthand notation for above 4 parameters) =20
> > >Slack term  Priority (Preemption, Defending, Admission, RPH=20
> > >Priority)  DSTE class
> > >
> > >Handling directives (optional to include):
> > > excess treatment
> > >
> > >Traffic Classifiers (optional to include):
> > > PHB Class (note that PHB class is set by downstream QNE to=20
> > >tell QNI how to mark traffic to ensure it will be treated as=20
> > >committed by admission
> > >control)
> > >
> > >Eliminated:
> > > bandwidth
> > > Ctot, Dtot, Csum, Dsum
> > >
> > >7. Editorial changes:
> > >A) rewrite description of read-only and read-write parameters=20
> > >to make the rules more clear:=20
> > >- which parameters can be ro/rw (function of object in=20
> which parameter
> > >included)
> > >- can only increase rw values and not delete, etc.
> > >B) eliminate the distinction of 'control' and 'qos description'
> > >parameters and replace with suggested taxonomy, as described above
> > >C) clarify discussion of packet outcomes from Y.1540 and place=20
> > >in appropriate sections
> > >D) clarify which parameters can be included in <QoS Desired> <QoS
> > >Available>, <QoS Reserved>, or <QoS Minimum>
> > >
> > >Please review and comment on the proposed changes.
> > >
> > >Thanks,
> > >Jerry, Dave, Cornelia
> > >
> > >
> > >-----Original Message-----
> > >From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> > >Sent: Monday, October 23, 2006 2:59 AM
> > >To: nsis@ietf.org
> > >Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
> > >
> > >Hi all,
> > >
> > >We have had a very large thread on the current=20
> > >draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear=20
> > >consensus, so I'd like to try to summarize the discussions.
> > >
> > >Several people feel that the current draft contains too many=20
> > >mandatory parameters and mapping some QoS Models to others is=20
> > >mixing apples-to-oranges.  The feeling is that=20
> > >interoperability is hard to achieve and therefore we should=20
> > >not try to gain interoperability.  Jukka Manner's message on=20
> > >17th of October proposed a way forward.
> > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> > >
> > >However, several (other) people feel that interoperability is=20
> > >important and is needed. Giving up on interoperability would=20
> > >not be a good thing, and that we are very close to a solution.=20
> > > Dave Oran proposed a mechanism to reach a solution on the=20
> > >19th October:
> > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> > >
> > >I have asked Jerry and Dave to make a proposal along the lines=20
> > >of Dave's message above, and the working group can review=20
> > >this, to see if we can gain consensus on this.  I think it=20
> > >would be sufficient to send some rough text to the mailing=20
> > >list, and if this is OK, then we can revise the draft.
> > >
> > >One open question to Dave/Jerry about the proposal is, in the=20
> > >event that the source cannot be described a mathematically=20
> > >complete way, is there any value (or does it make any sense at=20
> > >all) in sending an incomplete qspec, or potentially=20
> > >"wildcarding" sending some of the mandatory parameters.  I'm=20
> > >wondering if some sources / deployments may have certain=20
> > >parameters that are relatively well defined, but not all are=20
> > >so well defined.
> > >
> > >thanks,
> > >John
> > >
> > >_______________________________________________
> > >nsis mailing list
> > >nsis@ietf.org
> > >https://www1.ietf.org/mailman/listinfo/nsis
> > >
> > >_______________________________________________
> > >nsis mailing list
> > >nsis@ietf.org
> > >https://www1.ietf.org/mailman/listinfo/nsis
> > >
> >=20
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> >=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 05:43:30 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwVh-0004s8-6h; Fri, 03 Nov 2006 05:42:53 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwVf-0004s1-LN
	for nsis@ietf.org; Fri, 03 Nov 2006 05:42:51 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfwVe-00008a-Di
	for nsis@ietf.org; Fri, 03 Nov 2006 05:42:51 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-11.tower-146.messagelabs.com!1162550569!4050804!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 30982 invoked from network); 3 Nov 2006 10:42:49 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-11.tower-146.messagelabs.com with SMTP;
	3 Nov 2006 10:42:49 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA3AgnBA017893; 
	Fri, 3 Nov 2006 05:42:49 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA3AglM0017882; Fri, 3 Nov 2006 05:42:47 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] QSPEC Face-to-Face Chat: Another Try
Date: Fri, 3 Nov 2006 04:42:46 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F20@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <454B0BFE.3000708@gmx.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] QSPEC Face-to-Face Chat: Another Try
Thread-Index: Acb/Kw1VSYnxttyUQJ6bf9sfhbfvUAABVZ4Q
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "Hannes Tschofenig" <Hannes.Tschofenig@gmx.net>, "NSIS" <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 30ac594df0e66ffa5a93eb4c48bcb014
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

> What about Sunday, 3pm at the Sheraton San Diego Hotel (=3D=20
> meeting venue) in front of the hotel reception.

OK for me.  But please clarify which reception we should meet at (there
are 2), there is the Annex reception where most people are staying and
the main hotel reception (where the conference rooms are), which is it?

Thanks,
Jerry

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:02:43 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwoM-0006Pt-9F; Fri, 03 Nov 2006 06:02:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwoL-0006Pj-AJ
	for nsis@ietf.org; Fri, 03 Nov 2006 06:02:09 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfwoK-0003Aa-J8
	for nsis@ietf.org; Fri, 03 Nov 2006 06:02:09 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3B1Sc2003596; Fri, 3 Nov 2006 12:01:33 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>,
	"'Jukka MJ Manner'" <jmanner@cs.Helsinki.FI>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 12:01:26 +0100
Message-ID: <001d01c6ff37$6b50cf20$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F1F@KCCLUST06EVS1.ugd.att.com>
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1AAAElaWA=
X-Spam-Score: 0.262 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 12:01:35 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b058151374d77ee76edaac850f7449fb
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

QSPEC Stacking is not needed  for remapping.
For more details on QSPEC stacking please check Section 3.2.8 in:
http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt


Please note that QSEPC stacking is needed for stacking=20
local objects needed by a local (intra-domain)=20
QoS model that has different characteristics=20
than an end-to-end QoS model together with the end-to-end objects=20
needed by the end-to-end QoS model (see also requirement 5.4.2 in =
RFC3762).
Examples of such intra-domain QOS models are models such as Diffserv =
based,=20
MEF (Metro Ethernet Forum) based and 3GPP based.=20
An example of an end-to-end QoS model is the Intserv model.

Please also note that the remapping ptoblem from Token Bucket to=20
Bandwidth is not yet solved by your proposal.
=20
Best regards,
Georgios

> -----Original Message-----
> From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com]=20
> Sent: vrijdag 3 november 2006 11:42
> To: Jukka MJ Manner
> Cc: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
>=20
> Jukka,=20
>=20
> > -----Original Message-----
> > From: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI]
> > Sent: Friday, November 03, 2006 4:48 AM
> > To: Attila B=E1der (IJ/ETH)
> > Cc: Ash, Gerald R (Jerry), NPE; nsis@ietf.org; Kappler, Cornelia;=20
> > David R Oran; Cornelia Kappler
> > Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> >=20
> >=20
> > Hi all,
> >=20
> > I tend to agree with Attila here that stacking is a new=20
> concept and is=20
> > important to have. E.g., it allows a domain to add QoS information=20
> > that is only relevant to the domain itself, an ingress QNE pushes a=20
> > new QSPEC into the NSLP message, and the egress QNE of the=20
> domain just=20
> > pops it out. A domain isn't required to do so, but allows=20
> new kinds of=20
> > QoS management options.
>=20
> Without remapping of parameters being allowed, I don't know=20
> what a stacked QSPEC would do?  Let's take an example, e.g.,=20
> the QNI sets TMOD (M flag set) Parm A (M flag set) Parm B (M=20
> flag not set)
>=20
> In the local domain the edge QNEs must interpret TMOD and=20
> must interpret Parm B, otherwise the reservation is rejected.=20
>  Parm B can either be supported or ignored by the edge QNEs.
>=20
> What might a stacked qspec accomplish differently in the=20
> local domain?  Perhaps a new Parm C is pushed onto the local=20
> QSPEC, or what?  And if so, what would that accomplish?
>=20
> IMO we need a concrete example of how stacking would be=20
> useful if remapping is not allowed.  All our prior examples=20
> of stacking involved some kind of remapping of parameters=20
> from the initiator qspec to a local qspec.  Without that, I=20
> don't see what a stacked qspec accomplishes.
>=20
> Dave also asked for a concrete example of how stacking is=20
> useful (when remapping is not possible) in one of his posts yesterday.
>=20
> Thanks,
> Jerry
>=20
> P.S. Note that the only example of remapping that we ever=20
> discussed was token bucket to bandwidth.  All other examples=20
> of remapping were trying to map apples to oranges, and=20
> causing great confusion.  The concept of remapping is=20
> rightfully rejected.
>=20
> >=20
> > Jukka
> >=20
> > On Fri, 3 Nov 2006, Attila B=E1der (IJ/ETH) wrote:
> >=20
> > > Hi All, I disagree with the proposal below.=20
> > >=20
> > > This would remove the flexibility to develop and define new
> > QoS models.=20
> > > Actually the proposal supports only the Intserv model.=20
> Stacking and=20
> > > remapping have been fundamental features.
> > >=20
> > > I think that the best solution would be not to mandate
> > parameters and
> > > have only optional one. If we want to distinguish Intserv
> > as e2e QoS
> > > model, it is OK to mandate token bucket parameters, but
> > remapping it to
> > > other parameters and stacking it should be allowed.
> > >=20
> > > Best regards, Attila
> > >=20
> > >=20
> > > >-----Original Message-----
> > > >From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
> > > >Sent: Wednesday, November 01, 2006 11:13 PM
> > > >To: nsis@ietf.org
> > > >Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler;=20
> David R Oran;=20
> > > >Kappler,Cornelia
> > > >Subject: [NSIS] Proposed Updates to the QSPEC Draft
> > > >
> > > >Hi All,
> > > >
> > > >Per John's suggestion below we've discussed Dave Oran's=20
> suggestions=20
> > > >at length and have the following proposal.
> > > >
> > > >We think the proposed changes address many concerns=20
> raised on the=20
> > > >list over the last few months and will require some rather basic=20
> > > >changes in the QSPEC document=20
> > > >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> > > >
> > > >We appreciate Dave's suggestions and careful review of the
> > document.
> > > >
> > > >Here is a summary of the proposed changes:
> > > >
> > > >1. The notion of QOSMs will be played down in the updated draft.
> > > >Language to the effect that 'QNSLP/QSPEC can signal for=20
> different=20
> > > >QOSMs across multiple domains' will be replaced by the=20
> notion that=20
> > > >QNSLP/QSPEC allows different elements on the path to implement=20
> > > >different data plane QoS mechanisms that meets the constraints=20
> > > >included in the QSPEC.  A QOSM describes common=20
> capabilities among=20
> > > >QNEs to act consistently when requested to admit traffic and in=20
> > > >treating admitted traffic.
> > > >A QOSM ID need not be defined or signaled and a QNE need not=20
> > > >support any particular QOSM.  A 'QOSM specification'=20
> would become a=20
> > > >'QNSLP/QSPEC applicability statement', which documents how a=20
> > > >QNE/RMF interprets and treats the various elements in a QSPEC.
> > > >
> > > >An updated QOSM definition is as follows:
> > > >
> > > >QoS Model (QOSM): A method to achieve QoS for a traffic=20
> flow, e.g.,=20
> > > >IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC=20
> > > >QoS constraints and traffic handling directives a QNE=20
> implementing=20
> > > >that QOSM is capable of supporting and how resources will be=20
> > > >managed by the RMF in such a QNE.
> > > >
> > > >2. The QSPEC1/QSPEC2 semantics will be replaced with
> > Dave's a) --> d)
> > > >semantics:
> > > >A) The source traffic description is mandatory to include by the=20
> > > >QNI and mandatory to interpret by downstream QNEs.  The traffic=20
> > > >description is specified by a traffic model (TMOD) parameter=20
> > > >consisting of 4
> > > >sub-parameters:
> > > >- rate (r)
> > > >- bucket size (b)
> > > >- peak rate (p)
> > > >- minimum policed unit (m)
> > > >The TMOD parameter is a mathematically complete way to=20
> describe a=20
> > > >traffic source.  Examples will be given for setting TMOD to=20
> > > >correspond to specific cases, e.g. where only bandwidth is=20
> > > >available (r=3Dp; b and m set to default values) and other cases=20
> > > >(e.g., TCP).  As such, the bandwidth parameter is eliminated.
> > > >B) All other QSPEC parameters are optional to include by=20
> the QNI,=20
> > > >and may be either mandatory or optional to interpret by=20
> downstream=20
> > > >QNEs.
> > > >That is, the QNI can explicitly specify for each constraint=20
> > > >parameter whether it wishes admission control to succeed=20
> or fail if=20
> > > >that constraint cannot be met.  A downstream QNE distinguishes=20
> > > >whether it is failure to understand the parameter or failure to=20
> > > >meet the constraint that causes the admission control=20
> failure (more=20
> > > >detail on this below).
> > > >C) Additional QSPEC parameters can be defined in updates to the=20
> > > >QSPEC document (or alternatively specified in QOSM applicability=20
> > > >statement documents).  There is already an IANA registry=20
> > > >established for additional QSPEC parameters.
> > > >
> > > >3. The concept of remapping QSPEC parameters is eliminated. =20
> > > >This concept is ill defined and has caused much confusion (there=20
> > > >are few if any sensible remappings).  Rather we define=20
> 'interpret'=20
> > > >a QSPEC parameter to only mean what we now call 'strictly=20
> > > >interpret' (i.e., must conform to the RFCs defining the=20
> parameter=20
> > > >and procedures) and eliminate the remap option part of=20
> 'interpret'.
> > > >
> > > >4. Since remapping is eliminated, the concept of=20
> stacking QSPECs is=20
> > > >also eliminated.  Discussion of tunneling of QNSLP messages is=20
> > > >eliminated since there are no impacts on tunneling required by=20
> > > >QSPEC.
> > > >
> > > >5. Current QSPEC flags are modified as follows:
> > > >A) The QNI sets the M flag for each QSPEC parameter it populates=20
> > > >that it intends must be interpreted or the reservation fails. =20
> > > >Currently the M flag is statically set for every QSPEC parameter.
> > > >B) If the M flag is set, a downstream QNE must interpret the=20
> > > >parameter or the reservation fails.  If a QNE does not=20
> support the=20
> > > >parameter with the M flag set, it sets the N flag and=20
> rejects the=20
> > > >reservation.  If the QNE supports the parameter but=20
> cannot meet the=20
> > > >parameter, it sets the E flag and rejects the reservation.
> > > >C) If the M flag is not set, a downstream QNE should=20
> interpret the=20
> > > >parameter.  If a QNE does not support the parameter, it=20
> sets the N=20
> > > >flag and can optionally accept or reject the=20
> reservation.  If a QNE=20
> > > >supports the parameter but cannot meet the parameter, it=20
> sets the E=20
> > > >flag and optional accepts or rejects the reservation.
> > > >D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
> > > >eliminated.
> > > >
> > > >6. Summary of proposed QSPEC parameters:
> > > >
> > > >Source traffic Description:
> > > > TMOD-1 (mandatory to include)
> > > > TMOD-2 (optional to include)
> > > >
> > > >Constraints (optional to include):
> > > > Path Latency
> > > > Path Jitter
> > > > Path PLR
> > > > Path PER
> > > > Y.1541 class (shorthand notation for above 4 parameters) Slack=20
> > > >term  Priority (Preemption, Defending, Admission, RPH
> > > >Priority)  DSTE class
> > > >
> > > >Handling directives (optional to include):
> > > > excess treatment
> > > >
> > > >Traffic Classifiers (optional to include):
> > > > PHB Class (note that PHB class is set by downstream QNE to tell=20
> > > >QNI how to mark traffic to ensure it will be treated as=20
> committed=20
> > > >by admission
> > > >control)
> > > >
> > > >Eliminated:
> > > > bandwidth
> > > > Ctot, Dtot, Csum, Dsum
> > > >
> > > >7. Editorial changes:
> > > >A) rewrite description of read-only and read-write parameters to=20
> > > >make the rules more clear:
> > > >- which parameters can be ro/rw (function of object in
> > which parameter
> > > >included)
> > > >- can only increase rw values and not delete, etc.
> > > >B) eliminate the distinction of 'control' and 'qos description'
> > > >parameters and replace with suggested taxonomy, as=20
> described above
> > > >C) clarify discussion of packet outcomes from Y.1540 and=20
> place in=20
> > > >appropriate sections
> > > >D) clarify which parameters can be included in <QoS Desired> <QoS
> > > >Available>, <QoS Reserved>, or <QoS Minimum>
> > > >
> > > >Please review and comment on the proposed changes.
> > > >
> > > >Thanks,
> > > >Jerry, Dave, Cornelia
> > > >
> > > >
> > > >-----Original Message-----
> > > >From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> > > >Sent: Monday, October 23, 2006 2:59 AM
> > > >To: nsis@ietf.org
> > > >Subject: [NSIS] Where we stand with the=20
> draft-ietf-nsis-qspec draft
> > > >
> > > >Hi all,
> > > >
> > > >We have had a very large thread on the current=20
> > > >draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear=20
> > > >consensus, so I'd like to try to summarize the discussions.
> > > >
> > > >Several people feel that the current draft contains too many=20
> > > >mandatory parameters and mapping some QoS Models to others is=20
> > > >mixing apples-to-oranges.  The feeling is that=20
> interoperability is=20
> > > >hard to achieve and therefore we should not try to gain=20
> > > >interoperability.  Jukka Manner's message on 17th of October=20
> > > >proposed a way forward.
> > > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> > > >
> > > >However, several (other) people feel that interoperability is=20
> > > >important and is needed. Giving up on interoperability=20
> would not be=20
> > > >a good thing, and that we are very close to a solution.
> > > > Dave Oran proposed a mechanism to reach a solution on the 19th=20
> > > >October:
> > > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> > > >
> > > >I have asked Jerry and Dave to make a proposal along the=20
> lines of=20
> > > >Dave's message above, and the working group can review=20
> this, to see=20
> > > >if we can gain consensus on this.  I think it would be=20
> sufficient=20
> > > >to send some rough text to the mailing list, and if this is OK,=20
> > > >then we can revise the draft.
> > > >
> > > >One open question to Dave/Jerry about the proposal is,=20
> in the event=20
> > > >that the source cannot be described a mathematically=20
> complete way,=20
> > > >is there any value (or does it make any sense at
> > > >all) in sending an incomplete qspec, or potentially=20
> "wildcarding"=20
> > > >sending some of the mandatory parameters.  I'm wondering if some=20
> > > >sources / deployments may have certain parameters that are=20
> > > >relatively well defined, but not all are so well defined.
> > > >
> > > >thanks,
> > > >John
> > > >
> > > >_______________________________________________
> > > >nsis mailing list
> > > >nsis@ietf.org
> > > >https://www1.ietf.org/mailman/listinfo/nsis
> > > >
> > > >_______________________________________________
> > > >nsis mailing list
> > > >nsis@ietf.org
> > > >https://www1.ietf.org/mailman/listinfo/nsis
> > > >
> > >=20
> > > _______________________________________________
> > > nsis mailing list
> > > nsis@ietf.org
> > > https://www1.ietf.org/mailman/listinfo/nsis
> > >=20
> >=20
>=20
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>=20



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:11:06 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwwG-0007hq-Ij; Fri, 03 Nov 2006 06:10:20 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwwF-0007go-I4
	for nsis@ietf.org; Fri, 03 Nov 2006 06:10:19 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfwwD-0004Uf-0n
	for nsis@ietf.org; Fri, 03 Nov 2006 06:10:19 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3B9tXj003875; Fri, 3 Nov 2006 12:09:59 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, <nsis@ietf.org>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 12:09:53 +0100
Message-ID: <001e01c6ff38$995c43d0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F1D@KCCLUST06EVS1.ugd.att.com>
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQAAIJ7gA=
X-Spam-Score: 0.413 () AWL,J_CHICKENPOX_48
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 12:10:00 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0e9ebc0cbd700a87c0637ad0e2c91610
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

Thank you for the clarification!
Regarding the interpretation of the token bucket parameter 
I have another problem on understanding your proposal.

In your proposal you say:
>The TMOD parameter is a mathematically complete way to describe a 
>traffic source.  Examples will be given for setting TMOD to correspond 
>to specific cases, e.g. where only bandwidth is available (r=p; b and m 
>set to default values) and other cases (e.g., TCP).  As such, the 
>bandwidth parameter is eliminated.

Does the above mean that in specific cases a QNE can interpret only the
Token Bucket 
peak bandwdith parameter and ignore all the other Token Bucket parameters?

1) If the above is true, then how a QNI could know that a QNE used only the
Token Bucket 
Peak bandwidth parameter and another QNE used all the Token Bucket
paremeters?

2) If the above is not true then it means that a QoS model that needs to use
only 
the Token Bucket peak bandwidth parameter it will not be able to use 
the token bucket specification, thus it will not be able to use NSIS.


Which of the two alternatives is the right one?


Best Regards,
Georgios



 

> -----Original Message-----
> From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com] 
> Sent: vrijdag 3 november 2006 11:39
> To: Georgios Karagiannis; nsis@ietf.org
> Cc: Ash, Gerald R (Jerry), NPE
> Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing and not solving problem
> 
> Georgios, 
> 
> > -----Original Message-----
> > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > Sent: Friday, November 03, 2006 1:27 AM
> > To: nsis@ietf.org
> > Subject: [NSIS] jerry, dave,cornelia's proposal very 
> confusing and not 
> > solving problem
> > 
> > Hi all
> > 
> > After rereading the proposal of Jerry, Dave, Cornelia I 
> have seen that 
> > the main problem that we have had until now in the mailing 
> discussions 
> > is not solved. Mainly the number of mandatory parameters to 
> interpret 
> > by all the QNEs in the universe is not solved.
> 
> It is solved.  There is only one mandatory parameter that 
> must be interpreted by all QNE/RMFs (traffic description = 
> TMOD).  All other parameters are optional to interpret and 
> can be ignored by a QNE/RMF.
> I.e., there is only one 'qspec1' parameter (although 
> qspec1/qspec2 terminology is no longer used).
>  
> > The proposal mentions in point 2) point B):
> > 
> > "All other QSPEC parameters are optional to include by the QNI, and 
> > may be either mandatory or optional to interpret by 
> downstream QNEs."
> > 
> > Meaning that if a QNI includes the Y.1541 class and wants that all 
> > downstream QNEs interpret this parameter, then all downstream QNEs 
> > must be able to strictly interpret the Y.1541 class.
> > Jerry, Dave, Cornelia is this true?
> 
> No.  If the Y.1541 parameter is populated by the QNI and the 
> QNE does not support the Y.1541 parameter then:
> A) if the M flag is set by the QNI for the Y.1541 parameter, 
> the QNE sets the N flag for the Y.1541 parameter and rejects 
> the reservation
> B) if the M flag is not set by the QNI for the Y.1541 
> parameter, the QNE sets the N flag for the Y.1541 parameter 
> and optionally either accepts or rejects the reservation 
> (i.e., the QNE can simply ignore the Y.1541 parameter).
>  
> > Meaning that all QNEs that want to use NSIS must be implemented in 
> > such a way that if a QNI mandates the use of a parameter, e.g., 
> > Y.1541, then the QNE must strictly interpret it, otherwise 
> it has to 
> > reject the parameter or the admission request.
> > 
> > But this is exactly the problem that we wanted to solve by 
> having this 
> > discussions until now. Thus the problem is not solved.
> 
> I disagree.  All parameters except TMOD can be 
> *not-supported* by a QNE.
> If a QNI truly wants according to its QoS procedure being 
> implemented that a parameter must be interpreted or rejected, 
> then no other outcome makes sense other than to reject the 
> reservation if a parameter is not supported.  The QNI governs 
> the QoS outcome of the reservation.
> 
> The proposal is also similar to an earlier suggestion (by 
> Roland?) that a QNI can decide which parameters are mandatory 
> to interpret.
>  
> > Additionally, Jerry, Dave and Cornelia change the main 
> concepts of QoS 
> > based NSIS, please see my previous e-mails that I have sent 
> yesterday 
> > on this.
> > 
> > I have already proposed a solution to solve the problems 
> that we have 
> > had in the nsis mailing list, see a copy of that proposal below:
> > -------------------------
> > 
> > Another proposal for solving the interoperability issue and 
> the issue 
> > of having too many parameters was to:
> > * use only one mandatory parameter, i.e., Token Bucket that 
> could be 
> > remapped into the peak Bandwidth parameter.
> > Thus except the Token Bucket parameter,
> > * change the QSPEC-1 parameters into QSPEC-2 parameters
> 
> It is equivalent.  If you change Y.1541 to qspec2 and a QNE 
> doesn't support Y.1541, it sets the N flag and ignores the 
> parameter.  This is the same as the proposal as long as the 
> QNI doesn't care whether Y.1541 is supported.  However, if 
> the QNI does care that Y.1541 is supported, it will 
> reject/tear the reservation if the N flag is set.  The result 
> is the same as the proposal!
> 
> Jerry
> 
> > * describe the remapping procedure of the Token Bucket into 
> the peak 
> > Bandwidth parameter in the QSPEC template draft.
> > 
> > After that we can submit the QSPEC template draft for last 
> WG call. If 
> > after the deployment phase we conclude that more mandatory 
> parameters 
> > are needed for interoperability then we could work on an 
> interworking 
> > draft.
> > 
> > --------------
> > 
> > Best Regards,
> > Georgios
> > 
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> > 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:13:19 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfwyW-0002ad-GZ; Fri, 03 Nov 2006 06:12:40 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfwyV-0002Wv-Hg
	for nsis@ietf.org; Fri, 03 Nov 2006 06:12:39 -0500
Received: from mgw-ext13.nokia.com ([131.228.20.172])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfwyU-0004vQ-13
	for nsis@ietf.org; Fri, 03 Nov 2006 06:12:39 -0500
Received: from esebh105.NOE.Nokia.com (esebh105.ntc.nokia.com [172.21.138.211])
	by mgw-ext13.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA3BCJJQ009716; Fri, 3 Nov 2006 13:12:29 +0200
Received: from esebh101.NOE.Nokia.com ([172.21.138.177]) by
	esebh105.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 13:12:04 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh101.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 13:12:04 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] QSPEC Face-to-Face Chat: Another Try
Date: Fri, 3 Nov 2006 13:11:42 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED04EF@esebe199.NOE.Nokia.com>
In-Reply-To: <454B0BFE.3000708@gmx.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] QSPEC Face-to-Face Chat: Another Try
Thread-Index: Acb/K27F+x6cYeUBQL2zDMlAeIMRIAADU/bw
From: <john.loughney@nokia.com>
To: <Hannes.Tschofenig@gmx.net>, <nsis@ietf.org>
X-OriginalArrivalTime: 03 Nov 2006 11:12:04.0139 (UTC)
	FILETIME=[E4600FB0:01C6FF38]
X-eXpurgate-Category: 1/0
X-eXpurgate-ID: 149371::061103131229-24DFCBB0-67BD8251/0-0/0-0
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 538aad3a3c4f01d8b6a6477ca4248793
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

I changed my flight, should arrive Sunday at 2:10 PM, so I might be able
to
make this, barring flight delays.

John=20

>-----Original Message-----
>From: ext Hannes Tschofenig [mailto:Hannes.Tschofenig@gmx.net]=20
>Sent: 03 November, 2006 11:30
>To: NSIS
>Subject: [NSIS] QSPEC Face-to-Face Chat: Another Try
>
>Hi all,
>
>let me try it again.
>
>What about Sunday, 3pm at the Sheraton San Diego Hotel (=3D=20
>meeting venue) in front of the hotel reception.
>
>Would this work for someone?
>
>Ciao
>Hannes
>
>Hannes Tschofenig wrote:
>> Hi all,
>>=20
>> I wonder how many NSIS folks are available for a QSPEC face-to-face=20
>> discussion Sunday morning?
>>=20
>> Since many folks have to be at company internal meetings=20
>during lunch=20
>> time and want to see some of the tutorials I would suggest=20
>9am at the=20
>> Sheraton San Diego Hotel (=3D meeting venue) in front of the=20
>hotel reception.
>>=20
>> Is this a useful idea?
>>=20
>> Ciao
>> Hannes
>>=20
>> _______________________________________________
>> nsis mailing list
>> nsis@ietf.org
>> https://www1.ietf.org/mailman/listinfo/nsis
>
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:19:53 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfx4h-0006zd-46; Fri, 03 Nov 2006 06:19:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfx4f-0006ww-5b
	for nsis@ietf.org; Fri, 03 Nov 2006 06:19:01 -0500
Received: from tidos.tid.es ([193.145.240.2] helo=correo.tid.es)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfx4b-00063r-F6
	for nsis@ietf.org; Fri, 03 Nov 2006 06:19:01 -0500
Received: from tid (filvit [192.168.48.202])
	by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8
	2006))
	with ESMTP id <0J8500CTNIS646@tid.hi.inet> for nsis@ietf.org; Fri,
	03 Nov 2006 12:04:55 +0100 (MET)
Received: from [10.95.10.249] (herreria2.hi.inet [10.95.10.249])
	by tid.hi.inet (iPlanet Messaging Server 5.2 HotFix 2.14 (built Aug 8
	2006))
	with ESMTPA id <0J8500DYJIS61W@tid.hi.inet> for nsis@ietf.org; Fri,
	03 Nov 2006 12:04:54 +0100 (MET)
Date: Fri, 03 Nov 2006 12:08:09 +0100
From: =?ISO-8859-1?Q?Jorge_Andr=E9s_Col=E1s?= <jorgeac@tid.es>
Subject: Re: [NSIS] The version 3 of the InterDomain-QOSM Draft: the	NSIS QOS
	Model to Fulfill the E2E QoS Control in the ITU-T
	RACF	Functional	Architecture
In-reply-to: <454B1B33.1060801@docomolab-euro.com>
To: mendes <mendes@docomolab-euro.com>
Message-id: <454B2319.3010205@tid.es>
MIME-version: 1.0
Content-type: text/plain; format=flowed; charset=ISO-8859-1
Content-transfer-encoding: QUOTED-PRINTABLE
User-Agent: Thunderbird 1.5.0.7 (Windows/20060909)
References: <BAA65A575825454CBB0103267553FCCCED0466@esebe199.NOE.Nokia.com>
	<4549CACD.2000209@dcs.qmul.ac.uk> <454B1B33.1060801@docomolab-euro.com>
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 200d029292fbb60d25b263122ced50fc
Cc: john.loughney@nokia.com, edmundo@dei.uc.pt, karagian@cs.utwente.nl,
	nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hello,

I am not directly involved in the ITU-T RACF activities, I only have=
=20
access to the working documents.

Regards,
Jorge

mendes escribi=F3:
> Hi,
>
> i'm not directly involved in the ITU-T RACF activities.
>
> Paulo
>
> Jian Zhang wrote:
>  =20
>> Hello John,
>>
>> Currently I have not involved in the ITU-T RACF activities and I a=
m
>> not aware if Paulo and Jorge are involving with these activities
>> or not, but I believe the EU IST FP6 EuQoS project, as a whole, is
>> involved in the discussions (Edmundo knows better about that).
>>
>> Meanwhile, Georgios will discuss with the ITU-T people about these
>> issues during the coming IETF meeting.
>>
>> Cheers,
>>
>> Jian
>>
>> john.loughney@nokia.com wrote:
>>    =20
>>> Jian,
>>>
>>> Are any of the authors participating in the ITU-T RACF discussion=
s?
>>>
>>> thanks,
>>> John
>>> =20
>>>      =20
>>>> -----Original Message-----
>>>> From: ext Jian Zhang [mailto:jian.zhang@dcs.qmul.ac.uk] Sent: 31
>>>> October, 2006 17:18
>>>> To: nsis@ietf.org
>>>> Cc: edmundo@dei.uc.pt; karagian@cs.utwente.nl;
>>>> mendes@docomolab-euro.com
>>>> Subject: [NSIS] The version 3 of the InterDomain-QOSM Draft: the
>>>> NSIS QOS Model to Fulfill the E2E QoS Control in the ITU-T RACF
>>>> Functional Architecture
>>>>
>>>> Dear all,
>>>>
>>>> Please kindly read the version 3 of the InterDomain-QOSM Draft: =
our
>>>> effort to define a NSIS QOS Model to fulfill the E2E QoS control=
 in
>>>> the ITU-T RACF functional architecture, which has been submitted=
 to
>>>> the NSIS WG for discussion at the coming IETF meeting.
>>>>
>>>> And if you have any questions or suggestions about this draft, w=
e
>>>> would like to answer or discuss with you in this mailing list. T=
he
>>>> abstract of this draft is attached below and the link which hold=
s
>>>> the whole contents of the draft is also provided below.
>>>>
>>>> Cheers,
>>>>
>>>> Jian Zhang
>>>>
>>>> Dept. of Computer Science
>>>> Queen Mary, Univ. of London
>>>>
>>>> ---------------------------------------
>>>>
>>>> Abstract
>>>>
>>>>   This document has three goals. First of all, it presents our a=
nalysis
>>>>   of how to use the NSIS signaling (inter-domain QOSM and intra-=
domain
>>>>   QOSM) to fulfill the QoS control in accord with the ITU-T RACF
>>>>   functional architecture. For this goal, we discuss how the ITU=
-T RACF
>>>>   entities in the ITU-T RACF functional architecture can be mapp=
ed to
>>>>   the NSIS entities and how the RACF reference points can be
>>>>   implemented by using the NSIS protocol suites and QOSMs. Secon=
dly, we
>>>>   aim at specifying an NSIS Inter-domain QOSM for E2E QoS contro=
l
>>>>   across heterogeneous IP networks and applying this Inter-domai=
n QOSM
>>>>   to the e2e QoS control in the ITU-T RACF functional architectu=
re
>>>>   based on the above ITU-T RACF analysis. The detailed descripti=
on of
>>>>   the NSIS Inter-domain QOSM are given and the e2e QoS control
>>>>   scenarios in the ITU-T RACF architecture (including RACF Push =
and
>>>>   Pull resource control modes), which will be covered by the NSI=
S
>>>>   Inter-domain QOSM are described and implemented. Thirdly, we s=
pecify
>>>>   and implement those QSPECs that will be used by the Inter-doma=
in QOSM
>>>>   for the e2e QoS control in the ITU-T RACF architecture.
>>>>
>>>> The link is
>>>> http://ietfreport.isoc.org/ids/draft-zhang-nsis-interdomain-qos
>>>> m-03.txt .
>>>>
>>>>
>>>> _______________________________________________
>>>> nsis mailing list
>>>> nsis@ietf.org
>>>> https://www1.ietf.org/mailman/listinfo/nsis
>>>>
>>>>    =20
>>>>        =20
>>    =20
>
>
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>
>  =20

--=20
Jorge Andr=E9s Col=E1s
Telef=F3nica I+D
Emilio Vargas, 6
28043 - Madrid (Espa=F1a)
=20
E-mail: jorgeac@tid.es
Tel. +34 91 337 39 28
Fax. +34 91 510 33 08



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:33:51 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfxIN-0002En-Kn; Fri, 03 Nov 2006 06:33:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfxIM-0002Ei-He
	for nsis@ietf.org; Fri, 03 Nov 2006 06:33:10 -0500
Received: from goliath.siemens.de ([192.35.17.28])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfxIG-0008Kt-G1
	for nsis@ietf.org; Fri, 03 Nov 2006 06:33:10 -0500
Received: from mail3.siemens.de (localhost [127.0.0.1])
	by goliath.siemens.de (8.12.6/8.12.6) with ESMTP id kA3BWtFO001198;
	Fri, 3 Nov 2006 12:32:55 +0100
Received: from mchp771a.ww002.siemens.net (mchp771a.ww002.siemens.net
	[139.25.131.189])
	by mail3.siemens.de (8.12.6/8.12.6) with ESMTP id kA3BWseC023793;
	Fri, 3 Nov 2006 12:32:54 +0100
Received: from blns00fa.ww002.siemens.net ([147.54.45.102]) by
	mchp771a.ww002.siemens.net with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 12:32:54 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: AW: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 12:32:51 +0100
Message-ID: <DAA882AD01598448BB6FAB8C4D96F9DB02A3A4E3@blns00fa.ww002.siemens.net>
In-Reply-To: <Pine.LNX.4.64.0611031144540.7180@sbz-31.cs.Helsinki.FI>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb/Ow4aul89VaQ0R9CmhW0SoW+DwQAACBWw
From: "Kappler, Cornelia" <cornelia.kappler@siemens.com>
To: "Jukka MJ Manner" <jmanner@cs.Helsinki.FI>,
	=?iso-8859-1?Q?Attila_B=E1der_=28IJ/ETH=29?= <attila.bader@ericsson.com>
X-OriginalArrivalTime: 03 Nov 2006 11:32:54.0318 (UTC)
	FILETIME=[CD8A3CE0:01C6FF3B]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 10d2fdecab7a7fa796e06e001d026c91
Cc: "Ash, Gerald R \\\(Jerry\\\), ALABS" <gash@att.com>,
	David R Oran <oran@cisco.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jukka,=20

when you say "new information", would this be information not included =
in the original reservation? I.e. would the stacked QSPEC then be able =
to reserve something "different" from the initiator QSPEC? E.g. the =
original QSPEC just contains the TMOD parameter, and the stacked QSPEC =
additionally contains delay. If yes then we just need to agree whether =
we want this.=20

Couldnt stacking anyways always be replaced by tunnelling through a =
domain - i.e. the ingress QNE tunnels the incoming message to the end of =
the domain and generates a new message through the domain. So no real =
need to have the feature!?=20

Cornelia

> -----Urspr=FCngliche Nachricht-----
> Von: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI]=20
> Gesendet: Freitag, 3. November 2006 10:48
> An: Attila B=E1der (IJ/ETH)
> Cc: Ash, Gerald R \(Jerry\), ALABS; nsis@ietf.org; Kappler,=20
> Cornelia; David R Oran; Cornelia Kappler
> Betreff: RE: [NSIS] Proposed Updates to the QSPEC Draft
>=20
>=20
> Hi all,
>=20
> I tend to agree with Attila here that stacking is a new=20
> concept and is=20
> important to have. E.g., it allows a domain to add QoS=20
> information that is=20
> only relevant to the domain itself, an ingress QNE pushes a=20
> new QSPEC into=20
> the NSLP message, and the egress QNE of the domain just pops=20
> it out. A=20
> domain isnt't required to do so, but allows new kinds of QoS=20
> management=20
> options.
>=20
> Jukka
>=20
> On Fri, 3 Nov 2006, Attila B=E1der (IJ/ETH) wrote:
>=20
> > Hi All, I disagree with the proposal below.=20
> >=20
> > This would remove the flexibility to develop and define new=20
> QoS models.=20
> > Actually the proposal supports only the Intserv model. Stacking and=20
> > remapping have been fundamental features.
> >=20
> > I think that the best solution would be not to mandate=20
> parameters and=20
> > have only optional one. If we want to distinguish Intserv=20
> as e2e QoS=20
> > model, it is OK to mandate token bucket parameters, but=20
> remapping it to=20
> > other parameters and stacking it should be allowed.
> >=20
> > Best regards, Attila
> >=20
> >=20
> > >-----Original Message-----
> > >From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]=20
> > >Sent: Wednesday, November 01, 2006 11:13 PM
> > >To: nsis@ietf.org
> > >Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R=20
> > >Oran; Kappler,Cornelia
> > >Subject: [NSIS] Proposed Updates to the QSPEC Draft
> > >
> > >Hi All,
> > >
> > >Per John's suggestion below we've discussed Dave Oran's=20
> > >suggestions at length and have the following proposal. =20
> > >
> > >We think the proposed changes address many concerns raised on=20
> > >the list over the last few months and will require some rather=20
> > >basic changes in the QSPEC document=20
> > >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> > >
> > >We appreciate Dave's suggestions and careful review of the=20
> document.
> > >
> > >Here is a summary of the proposed changes:
> > >
> > >1. The notion of QOSMs will be played down in the updated draft.
> > >Language to the effect that 'QNSLP/QSPEC can signal for=20
> > >different QOSMs across multiple domains' will be replaced by=20
> > >the notion that QNSLP/QSPEC allows different elements on the=20
> > >path to implement different data plane QoS mechanisms that=20
> > >meets the constraints included in the QSPEC.  A QOSM describes=20
> > >common capabilities among QNEs to act consistently when=20
> > >requested to admit traffic and in treating admitted traffic. =20
> > >A QOSM ID need not be defined or signaled and a QNE need not=20
> > >support any particular QOSM.  A 'QOSM specification' would=20
> > >become a 'QNSLP/QSPEC applicability statement', which=20
> > >documents how a QNE/RMF interprets and treats the various=20
> > >elements in a QSPEC. =20
> > >
> > >An updated QOSM definition is as follows:
> > >
> > >QoS Model (QOSM): A method to achieve QoS for a traffic flow,=20
> > >e.g., IntServ Controlled Load.  A QOSM specifies what sub-set=20
> > >of QSPEC QoS constraints and traffic handling directives a QNE=20
> > >implementing that QOSM is capable of supporting and how=20
> > >resources will be managed by the RMF in such a QNE.
> > >
> > >2. The QSPEC1/QSPEC2 semantics will be replaced with=20
> Dave's a) --> d)
> > >semantics:
> > >A) The source traffic description is mandatory to include by=20
> > >the QNI and mandatory to interpret by downstream QNEs.  The=20
> > >traffic description is specified by a traffic model (TMOD)=20
> > >parameter consisting of 4
> > >sub-parameters:
> > >- rate (r)
> > >- bucket size (b)
> > >- peak rate (p)
> > >- minimum policed unit (m)
> > >The TMOD parameter is a mathematically complete way to=20
> > >describe a traffic source.  Examples will be given for setting=20
> > >TMOD to correspond to specific cases, e.g. where only=20
> > >bandwidth is available (r=3Dp; b and m set to default values)=20
> > >and other cases (e.g., TCP).  As such, the bandwidth parameter=20
> > >is eliminated.
> > >B) All other QSPEC parameters are optional to include by the=20
> > >QNI, and may be either mandatory or optional to interpret by=20
> > >downstream QNEs.
> > >That is, the QNI can explicitly specify for each constraint=20
> > >parameter whether it wishes admission control to succeed or=20
> > >fail if that constraint cannot be met.  A downstream QNE=20
> > >distinguishes whether it is failure to understand the=20
> > >parameter or failure to meet the constraint that causes the=20
> > >admission control failure (more detail on this below).
> > >C) Additional QSPEC parameters can be defined in updates to=20
> > >the QSPEC document (or alternatively specified in QOSM=20
> > >applicability statement documents).  There is already an IANA=20
> > >registry established for additional QSPEC parameters.
> > >
> > >3. The concept of remapping QSPEC parameters is eliminated. =20
> > >This concept is ill defined and has caused much confusion=20
> > >(there are few if any sensible remappings).  Rather we define=20
> > >'interpret' a QSPEC parameter to only mean what we now call=20
> > >'strictly interpret' (i.e., must conform to the RFCs defining=20
> > >the parameter and procedures) and eliminate the remap option=20
> > >part of 'interpret'.
> > >
> > >4. Since remapping is eliminated, the concept of stacking=20
> > >QSPECs is also eliminated.  Discussion of tunneling of QNSLP=20
> > >messages is eliminated since there are no impacts on tunneling=20
> > >required by QSPEC.
> > >
> > >5. Current QSPEC flags are modified as follows:
> > >A) The QNI sets the M flag for each QSPEC parameter it=20
> > >populates that it intends must be interpreted or the=20
> > >reservation fails.  Currently the M flag is statically set for=20
> > >every QSPEC parameter.
> > >B) If the M flag is set, a downstream QNE must interpret the=20
> > >parameter or the reservation fails.  If a QNE does not support=20
> > >the parameter with the M flag set, it sets the N flag and=20
> > >rejects the reservation.  If the QNE supports the parameter=20
> > >but cannot meet the parameter, it sets the E flag and rejects=20
> > >the reservation.=20
> > >C) If the M flag is not set, a downstream QNE should interpret=20
> > >the parameter.  If a QNE does not support the parameter, it=20
> > >sets the N flag and can optionally accept or reject the=20
> > >reservation.  If a QNE supports the parameter but cannot meet=20
> > >the parameter, it sets the E flag and optional accepts or=20
> > >rejects the reservation.
> > >D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
> > >eliminated.
> > >
> > >6. Summary of proposed QSPEC parameters:
> > >
> > >Source traffic Description:
> > > TMOD-1 (mandatory to include)
> > > TMOD-2 (optional to include)
> > >
> > >Constraints (optional to include):
> > > Path Latency
> > > Path Jitter
> > > Path PLR
> > > Path PER
> > > Y.1541 class (shorthand notation for above 4 parameters) =20
> > >Slack term  Priority (Preemption, Defending, Admission, RPH=20
> > >Priority)  DSTE class
> > >
> > >Handling directives (optional to include):
> > > excess treatment
> > >
> > >Traffic Classifiers (optional to include):
> > > PHB Class (note that PHB class is set by downstream QNE to=20
> > >tell QNI how to mark traffic to ensure it will be treated as=20
> > >committed by admission
> > >control)
> > >
> > >Eliminated:
> > > bandwidth
> > > Ctot, Dtot, Csum, Dsum
> > >
> > >7. Editorial changes:
> > >A) rewrite description of read-only and read-write parameters=20
> > >to make the rules more clear:=20
> > >- which parameters can be ro/rw (function of object in=20
> which parameter
> > >included)
> > >- can only increase rw values and not delete, etc.
> > >B) eliminate the distinction of 'control' and 'qos description'
> > >parameters and replace with suggested taxonomy, as described above
> > >C) clarify discussion of packet outcomes from Y.1540 and place=20
> > >in appropriate sections
> > >D) clarify which parameters can be included in <QoS Desired> <QoS
> > >Available>, <QoS Reserved>, or <QoS Minimum>
> > >
> > >Please review and comment on the proposed changes.
> > >
> > >Thanks,
> > >Jerry, Dave, Cornelia
> > >
> > >
> > >-----Original Message-----
> > >From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> > >Sent: Monday, October 23, 2006 2:59 AM
> > >To: nsis@ietf.org
> > >Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft
> > >
> > >Hi all,
> > >
> > >We have had a very large thread on the current=20
> > >draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear=20
> > >consensus, so I'd like to try to summarize the discussions.
> > >
> > >Several people feel that the current draft contains too many=20
> > >mandatory parameters and mapping some QoS Models to others is=20
> > >mixing apples-to-oranges.  The feeling is that=20
> > >interoperability is hard to achieve and therefore we should=20
> > >not try to gain interoperability.  Jukka Manner's message on=20
> > >17th of October proposed a way forward.
> > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> > >
> > >However, several (other) people feel that interoperability is=20
> > >important and is needed. Giving up on interoperability would=20
> > >not be a good thing, and that we are very close to a solution.=20
> > > Dave Oran proposed a mechanism to reach a solution on the=20
> > >19th October:
> > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> > >
> > >I have asked Jerry and Dave to make a proposal along the lines=20
> > >of Dave's message above, and the working group can review=20
> > >this, to see if we can gain consensus on this.  I think it=20
> > >would be sufficient to send some rough text to the mailing=20
> > >list, and if this is OK, then we can revise the draft.
> > >
> > >One open question to Dave/Jerry about the proposal is, in the=20
> > >event that the source cannot be described a mathematically=20
> > >complete way, is there any value (or does it make any sense at=20
> > >all) in sending an incomplete qspec, or potentially=20
> > >"wildcarding" sending some of the mandatory parameters.  I'm=20
> > >wondering if some sources / deployments may have certain=20
> > >parameters that are relatively well defined, but not all are=20
> > >so well defined.
> > >
> > >thanks,
> > >John
> > >
> > >_______________________________________________
> > >nsis mailing list
> > >nsis@ietf.org
> > >https://www1.ietf.org/mailman/listinfo/nsis
> > >
> > >_______________________________________________
> > >nsis mailing list
> > >nsis@ietf.org
> > >https://www1.ietf.org/mailman/listinfo/nsis
> > >
> >=20
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> >=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:36:57 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfxL2-0004wT-0e; Fri, 03 Nov 2006 06:35:56 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfxL0-0004wN-Uk
	for nsis@ietf.org; Fri, 03 Nov 2006 06:35:54 -0500
Received: from mgw-ext14.nokia.com ([131.228.20.173])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfxKu-0000Je-S8
	for nsis@ietf.org; Fri, 03 Nov 2006 06:35:54 -0500
Received: from esebh108.NOE.Nokia.com (esebh108.ntc.nokia.com [172.21.143.145])
	by mgw-ext14.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA3BZcrG018017; Fri, 3 Nov 2006 13:35:40 +0200
Received: from esebh101.NOE.Nokia.com ([172.21.138.177]) by
	esebh108.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 13:35:38 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh101.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 13:35:38 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 13:35:32 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED04F5@esebe199.NOE.Nokia.com>
In-Reply-To: <001d01c6ff37$6b50cf20$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1AAAElaWAAAXRn4A==
From: <john.loughney@nokia.com>
To: <karagian@cs.utwente.nl>, <gash@att.com>, <jmanner@cs.Helsinki.FI>
X-OriginalArrivalTime: 03 Nov 2006 11:35:38.0324 (UTC)
	FILETIME=[2F4B9140:01C6FF3C]
X-eXpurgate-Category: 1/0
X-eXpurgate-ID: 149371::061103133540-44D30BB0-3E196BDD/0-0/0-0
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 9466e0365fc95844abaf7c3f15a05c7d
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios

>QSPEC Stacking is not needed  for remapping.
>For more details on QSPEC stacking please check Section 3.2.8 in:
>http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt

The text in that section says the following:

   When a domain wants to apply a certain QoS model to an incoming per-
   flow reservation request, each edge of the domain is configured to
   map the incoming QSPEC object to a local QSPEC object and push that
   object onto the stack of QSPEC objects. QNEs inside the domain look
   at the top of the QSPEC object stack to determine which QoS model to

Manner et al.              Expires March 2007                  [Page 15]
=0C
Internet-Draft                  QoS NSLP                    October 2006

   apply for the reservation.

=20
This is clearly remapping from the e2e QSPEC to the local QSPEC, right?
since this is the case, could not the local QoS domain just re-use the
existing (e2e) QSPEC - or does the local QoS domain need to add
additional
parameters?  I feel like I'm missing something from you description.

thanks,
John

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:40:13 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfxOa-0002K3-LC; Fri, 03 Nov 2006 06:39:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfxOZ-0002Jy-L6
	for nsis@ietf.org; Fri, 03 Nov 2006 06:39:35 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfxOY-0000ti-VI
	for nsis@ietf.org; Fri, 03 Nov 2006 06:39:35 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3Bcesp004913; Fri, 3 Nov 2006 12:38:45 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Kappler, Cornelia'" <cornelia.kappler@siemens.com>,
	"'Jukka MJ Manner'" <jmanner@cs.Helsinki.FI>,
	"=?iso-8859-1?Q?'Attila_B=E1der_=28IJ/ETH=29'?="
	<attila.bader@ericsson.com>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 12:38:37 +0100
Message-ID: <001f01c6ff3c$9d6d1450$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <DAA882AD01598448BB6FAB8C4D96F9DB02A3A4E3@blns00fa.ww002.siemens.net>
Thread-Index: Acb/Ow4aul89VaQ0R9CmhW0SoW+DwQAACBWwAABGd2A=
X-Spam-Score: 0.265 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 12:38:47 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: fca741f5016e6ff607eaed2fd431d10d
Cc: "'Ash, Gerald R \\\(Jerry\\\), ALABS'" <gash@att.com>,
	'David R Oran' <oran@cisco.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Cornelia

Please note that QSEPC stacking is needed for stacking local=20
objects needed by a local (intra-domain) QoS model that=20
has different characteristics than an end-to-end QoS=20
model together with the end-to-end objects needed by the=20
end-to-end QoS model (see also requirement 5.4.2 in RFC3762).
Examples of such intra-domain QOS models are models such as=20
Diffserv based, MEF (Metro Ethernet Forum) based and 3GPP based.=20
An example of an end-to-end QoS model is the Intserv model.=20

Thus these stacking issues are already approved by the NSIS WG and IESG, =

before RFC3762 had been published.

Best Regards,
Georgios


> -----Original Message-----
> From: Kappler, Cornelia [mailto:cornelia.kappler@siemens.com]=20
> Sent: vrijdag 3 november 2006 12:33
> To: Jukka MJ Manner; Attila B=E1der (IJ/ETH)
> Cc: Ash, Gerald R \(Jerry\), ALABS; David R Oran; nsis@ietf.org
> Subject: AW: [NSIS] Proposed Updates to the QSPEC Draft
>=20
> Hi Jukka,=20
>=20
> when you say "new information", would this be information not=20
> included in the original reservation? I.e. would the stacked=20
> QSPEC then be able to reserve something "different" from the=20
> initiator QSPEC? E.g. the original QSPEC just contains the=20
> TMOD parameter, and the stacked QSPEC additionally contains=20
> delay. If yes then we just need to agree whether we want this.=20
>=20
> Couldnt stacking anyways always be replaced by tunnelling=20
> through a domain - i.e. the ingress QNE tunnels the incoming=20
> message to the end of the domain and generates a new message=20
> through the domain. So no real need to have the feature!?=20
>=20
> Cornelia
>=20
> > -----Urspr=FCngliche Nachricht-----
> > Von: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI]
> > Gesendet: Freitag, 3. November 2006 10:48
> > An: Attila B=E1der (IJ/ETH)
> > Cc: Ash, Gerald R \(Jerry\), ALABS; nsis@ietf.org; Kappler,=20
> Cornelia;=20
> > David R Oran; Cornelia Kappler
> > Betreff: RE: [NSIS] Proposed Updates to the QSPEC Draft
> >=20
> >=20
> > Hi all,
> >=20
> > I tend to agree with Attila here that stacking is a new=20
> concept and is=20
> > important to have. E.g., it allows a domain to add QoS information=20
> > that is only relevant to the domain itself, an ingress QNE pushes a=20
> > new QSPEC into the NSLP message, and the egress QNE of the=20
> domain just=20
> > pops it out. A domain isnt't required to do so, but allows=20
> new kinds=20
> > of QoS management options.
> >=20
> > Jukka
> >=20
> > On Fri, 3 Nov 2006, Attila B=E1der (IJ/ETH) wrote:
> >=20
> > > Hi All, I disagree with the proposal below.=20
> > >=20
> > > This would remove the flexibility to develop and define new
> > QoS models.=20
> > > Actually the proposal supports only the Intserv model.=20
> Stacking and=20
> > > remapping have been fundamental features.
> > >=20
> > > I think that the best solution would be not to mandate
> > parameters and
> > > have only optional one. If we want to distinguish Intserv
> > as e2e QoS
> > > model, it is OK to mandate token bucket parameters, but
> > remapping it to
> > > other parameters and stacking it should be allowed.
> > >=20
> > > Best regards, Attila
> > >=20
> > >=20
> > > >-----Original Message-----
> > > >From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
> > > >Sent: Wednesday, November 01, 2006 11:13 PM
> > > >To: nsis@ietf.org
> > > >Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler;=20
> David R Oran;=20
> > > >Kappler,Cornelia
> > > >Subject: [NSIS] Proposed Updates to the QSPEC Draft
> > > >
> > > >Hi All,
> > > >
> > > >Per John's suggestion below we've discussed Dave Oran's=20
> suggestions=20
> > > >at length and have the following proposal.
> > > >
> > > >We think the proposed changes address many concerns=20
> raised on the=20
> > > >list over the last few months and will require some rather basic=20
> > > >changes in the QSPEC document=20
> > > >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
> > > >
> > > >We appreciate Dave's suggestions and careful review of the
> > document.
> > > >
> > > >Here is a summary of the proposed changes:
> > > >
> > > >1. The notion of QOSMs will be played down in the updated draft.
> > > >Language to the effect that 'QNSLP/QSPEC can signal for=20
> different=20
> > > >QOSMs across multiple domains' will be replaced by the=20
> notion that=20
> > > >QNSLP/QSPEC allows different elements on the path to implement=20
> > > >different data plane QoS mechanisms that meets the constraints=20
> > > >included in the QSPEC.  A QOSM describes common=20
> capabilities among=20
> > > >QNEs to act consistently when requested to admit traffic and in=20
> > > >treating admitted traffic.
> > > >A QOSM ID need not be defined or signaled and a QNE need not=20
> > > >support any particular QOSM.  A 'QOSM specification'=20
> would become a=20
> > > >'QNSLP/QSPEC applicability statement', which documents how a=20
> > > >QNE/RMF interprets and treats the various elements in a QSPEC.
> > > >
> > > >An updated QOSM definition is as follows:
> > > >
> > > >QoS Model (QOSM): A method to achieve QoS for a traffic=20
> flow, e.g.,=20
> > > >IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC=20
> > > >QoS constraints and traffic handling directives a QNE=20
> implementing=20
> > > >that QOSM is capable of supporting and how resources will be=20
> > > >managed by the RMF in such a QNE.
> > > >
> > > >2. The QSPEC1/QSPEC2 semantics will be replaced with
> > Dave's a) --> d)
> > > >semantics:
> > > >A) The source traffic description is mandatory to include by the=20
> > > >QNI and mandatory to interpret by downstream QNEs.  The traffic=20
> > > >description is specified by a traffic model (TMOD) parameter=20
> > > >consisting of 4
> > > >sub-parameters:
> > > >- rate (r)
> > > >- bucket size (b)
> > > >- peak rate (p)
> > > >- minimum policed unit (m)
> > > >The TMOD parameter is a mathematically complete way to=20
> describe a=20
> > > >traffic source.  Examples will be given for setting TMOD to=20
> > > >correspond to specific cases, e.g. where only bandwidth is=20
> > > >available (r=3Dp; b and m set to default values) and other cases=20
> > > >(e.g., TCP).  As such, the bandwidth parameter is eliminated.
> > > >B) All other QSPEC parameters are optional to include by=20
> the QNI,=20
> > > >and may be either mandatory or optional to interpret by=20
> downstream=20
> > > >QNEs.
> > > >That is, the QNI can explicitly specify for each constraint=20
> > > >parameter whether it wishes admission control to succeed=20
> or fail if=20
> > > >that constraint cannot be met.  A downstream QNE distinguishes=20
> > > >whether it is failure to understand the parameter or failure to=20
> > > >meet the constraint that causes the admission control=20
> failure (more=20
> > > >detail on this below).
> > > >C) Additional QSPEC parameters can be defined in updates to the=20
> > > >QSPEC document (or alternatively specified in QOSM applicability=20
> > > >statement documents).  There is already an IANA registry=20
> > > >established for additional QSPEC parameters.
> > > >
> > > >3. The concept of remapping QSPEC parameters is eliminated. =20
> > > >This concept is ill defined and has caused much confusion (there=20
> > > >are few if any sensible remappings).  Rather we define=20
> 'interpret'=20
> > > >a QSPEC parameter to only mean what we now call 'strictly=20
> > > >interpret' (i.e., must conform to the RFCs defining the=20
> parameter=20
> > > >and procedures) and eliminate the remap option part of=20
> 'interpret'.
> > > >
> > > >4. Since remapping is eliminated, the concept of=20
> stacking QSPECs is=20
> > > >also eliminated.  Discussion of tunneling of QNSLP messages is=20
> > > >eliminated since there are no impacts on tunneling required by=20
> > > >QSPEC.
> > > >
> > > >5. Current QSPEC flags are modified as follows:
> > > >A) The QNI sets the M flag for each QSPEC parameter it populates=20
> > > >that it intends must be interpreted or the reservation fails. =20
> > > >Currently the M flag is statically set for every QSPEC parameter.
> > > >B) If the M flag is set, a downstream QNE must interpret the=20
> > > >parameter or the reservation fails.  If a QNE does not=20
> support the=20
> > > >parameter with the M flag set, it sets the N flag and=20
> rejects the=20
> > > >reservation.  If the QNE supports the parameter but=20
> cannot meet the=20
> > > >parameter, it sets the E flag and rejects the reservation.
> > > >C) If the M flag is not set, a downstream QNE should=20
> interpret the=20
> > > >parameter.  If a QNE does not support the parameter, it=20
> sets the N=20
> > > >flag and can optionally accept or reject the=20
> reservation.  If a QNE=20
> > > >supports the parameter but cannot meet the parameter, it=20
> sets the E=20
> > > >flag and optional accepts or rejects the reservation.
> > > >D) The R (remapped parameter) flag and Q (non QOSM) flag are=20
> > > >eliminated.
> > > >
> > > >6. Summary of proposed QSPEC parameters:
> > > >
> > > >Source traffic Description:
> > > > TMOD-1 (mandatory to include)
> > > > TMOD-2 (optional to include)
> > > >
> > > >Constraints (optional to include):
> > > > Path Latency
> > > > Path Jitter
> > > > Path PLR
> > > > Path PER
> > > > Y.1541 class (shorthand notation for above 4 parameters) Slack=20
> > > >term  Priority (Preemption, Defending, Admission, RPH
> > > >Priority)  DSTE class
> > > >
> > > >Handling directives (optional to include):
> > > > excess treatment
> > > >
> > > >Traffic Classifiers (optional to include):
> > > > PHB Class (note that PHB class is set by downstream QNE to tell=20
> > > >QNI how to mark traffic to ensure it will be treated as=20
> committed=20
> > > >by admission
> > > >control)
> > > >
> > > >Eliminated:
> > > > bandwidth
> > > > Ctot, Dtot, Csum, Dsum
> > > >
> > > >7. Editorial changes:
> > > >A) rewrite description of read-only and read-write parameters to=20
> > > >make the rules more clear:
> > > >- which parameters can be ro/rw (function of object in
> > which parameter
> > > >included)
> > > >- can only increase rw values and not delete, etc.
> > > >B) eliminate the distinction of 'control' and 'qos description'
> > > >parameters and replace with suggested taxonomy, as=20
> described above
> > > >C) clarify discussion of packet outcomes from Y.1540 and=20
> place in=20
> > > >appropriate sections
> > > >D) clarify which parameters can be included in <QoS Desired> <QoS
> > > >Available>, <QoS Reserved>, or <QoS Minimum>
> > > >
> > > >Please review and comment on the proposed changes.
> > > >
> > > >Thanks,
> > > >Jerry, Dave, Cornelia
> > > >
> > > >
> > > >-----Original Message-----
> > > >From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> > > >Sent: Monday, October 23, 2006 2:59 AM
> > > >To: nsis@ietf.org
> > > >Subject: [NSIS] Where we stand with the=20
> draft-ietf-nsis-qspec draft
> > > >
> > > >Hi all,
> > > >
> > > >We have had a very large thread on the current=20
> > > >draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear=20
> > > >consensus, so I'd like to try to summarize the discussions.
> > > >
> > > >Several people feel that the current draft contains too many=20
> > > >mandatory parameters and mapping some QoS Models to others is=20
> > > >mixing apples-to-oranges.  The feeling is that=20
> interoperability is=20
> > > >hard to achieve and therefore we should not try to gain=20
> > > >interoperability.  Jukka Manner's message on 17th of October=20
> > > >proposed a way forward.
> > > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
> > > >
> > > >However, several (other) people feel that interoperability is=20
> > > >important and is needed. Giving up on interoperability=20
> would not be=20
> > > >a good thing, and that we are very close to a solution.
> > > > Dave Oran proposed a mechanism to reach a solution on the 19th=20
> > > >October:
> > > >http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
> > > >
> > > >I have asked Jerry and Dave to make a proposal along the=20
> lines of=20
> > > >Dave's message above, and the working group can review=20
> this, to see=20
> > > >if we can gain consensus on this.  I think it would be=20
> sufficient=20
> > > >to send some rough text to the mailing list, and if this is OK,=20
> > > >then we can revise the draft.
> > > >
> > > >One open question to Dave/Jerry about the proposal is,=20
> in the event=20
> > > >that the source cannot be described a mathematically=20
> complete way,=20
> > > >is there any value (or does it make any sense at
> > > >all) in sending an incomplete qspec, or potentially=20
> "wildcarding"=20
> > > >sending some of the mandatory parameters.  I'm wondering if some=20
> > > >sources / deployments may have certain parameters that are=20
> > > >relatively well defined, but not all are so well defined.
> > > >
> > > >thanks,
> > > >John
> > > >
> > > >_______________________________________________
> > > >nsis mailing list
> > > >nsis@ietf.org
> > > >https://www1.ietf.org/mailman/listinfo/nsis
> > > >
> > > >_______________________________________________
> > > >nsis mailing list
> > > >nsis@ietf.org
> > > >https://www1.ietf.org/mailman/listinfo/nsis
> > > >
> > >=20
> > > _______________________________________________
> > > nsis mailing list
> > > nsis@ietf.org
> > > https://www1.ietf.org/mailman/listinfo/nsis
> > >=20
> >=20
>=20
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>=20



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 06:57:17 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfxep-0005vg-GT; Fri, 03 Nov 2006 06:56:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfxe1-0005Pd-4I
	for nsis@ietf.org; Fri, 03 Nov 2006 06:55:33 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfxTC-0001gg-Lo
	for nsis@ietf.org; Fri, 03 Nov 2006 06:44:24 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3Bi9vt005069; Fri, 3 Nov 2006 12:44:14 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <john.loughney@nokia.com>, <gash@att.com>, <jmanner@cs.Helsinki.FI>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 12:44:07 +0100
Message-ID: <002001c6ff3d$61ee4150$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <BAA65A575825454CBB0103267553FCCCED04F5@esebe199.NOE.Nokia.com>
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1AAAElaWAAAXRn4AAAPniw
X-Spam-Score: 0.265 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 12:44:14 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 92df29fa99cf13e554b84c8374345c17
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi John

The text that you have used talks about remapping and for=20
that case QSPEC stacking can be applied for this purpose,
but the main need for QSPEC stacking is given by the
 requirement 5.4.2 in RFC 3762.
Furthermore, in the same section 3.2.8, you could also=20
find the text that does not talk about remapping:

  "By using QSpec staking, an increase in flexibility, modularity and
   signaling performance an be achieved. The flexibility and modularity
   increase can be achieved by providing the possibility to any existing
   or future QoS model applied in a local QoS NSLP aware domain to use
   and stack any required local QoS model information on QoS NSLP
   signaling messages passing through the local domain (see requirement
   5.4.2 in RFC3762).

   The signaling performance increase can be achieved by reducing the
   number of required local QoS NSLP signaling messages, when local QoS
   model information is stacked on the QoS NSLP signaling messages that
   are passing through the local domain."

Best regards,
Georgios

=20

> -----Original Message-----
> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]=20
> Sent: vrijdag 3 november 2006 12:36
> To: karagian@cs.utwente.nl; gash@att.com; jmanner@cs.Helsinki.FI
> Cc: nsis@ietf.org
> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
>=20
> Hi Georgios
>=20
> >QSPEC Stacking is not needed  for remapping.
> >For more details on QSPEC stacking please check Section 3.2.8 in:
> >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt
>=20
> The text in that section says the following:
>=20
>    When a domain wants to apply a certain QoS model to an=20
> incoming per-
>    flow reservation request, each edge of the domain is configured to
>    map the incoming QSPEC object to a local QSPEC object and push that
>    object onto the stack of QSPEC objects. QNEs inside the domain look
>    at the top of the QSPEC object stack to determine which=20
> QoS model to
>=20
> Manner et al.              Expires March 2007                =20
>  [Page 15]
> =0C>=20
> Internet-Draft                  QoS NSLP                   =20
> October 2006
>=20
>    apply for the reservation.
>=20
> =20
> This is clearly remapping from the e2e QSPEC to the local=20
> QSPEC, right?
> since this is the case, could not the local QoS domain just=20
> re-use the existing (e2e) QSPEC - or does the local QoS=20
> domain need to add additional parameters?  I feel like I'm=20
> missing something from you description.
>=20
> thanks,
> John
>=20



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 07:03:11 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfxko-0005Vs-5A; Fri, 03 Nov 2006 07:02:34 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfxkn-0005VR-GW
	for nsis@ietf.org; Fri, 03 Nov 2006 07:02:33 -0500
Received: from mgw-ext13.nokia.com ([131.228.20.172])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfxkj-00051K-0i
	for nsis@ietf.org; Fri, 03 Nov 2006 07:02:33 -0500
Received: from esebh106.NOE.Nokia.com (esebh106.ntc.nokia.com [172.21.138.213])
	by mgw-ext13.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA3C2FVU004818; Fri, 3 Nov 2006 14:02:20 +0200
Received: from esebh102.NOE.Nokia.com ([172.21.138.183]) by
	esebh106.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 14:02:10 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh102.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 14:02:10 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 14:01:54 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED04FA@esebe199.NOE.Nokia.com>
In-Reply-To: <001d01c6ff37$6b50cf20$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1AAAElaWAAATDhIA==
From: <john.loughney@nokia.com>
To: <karagian@cs.utwente.nl>, <gash@att.com>, <jmanner@cs.Helsinki.FI>
X-OriginalArrivalTime: 03 Nov 2006 12:02:10.0206 (UTC)
	FILETIME=[E4217BE0:01C6FF3F]
X-eXpurgate-Category: 1/0
X-eXpurgate-ID: 149371::061103140220-127C5BB0-03AB4B48/0-0/0-0
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 7d33c50f3756db14428398e2bdedd581
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios,

>QSPEC Stacking is not needed  for remapping.
>For more details on QSPEC stacking please check Section 3.2.8 in:
>http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt
>
>Please note that QSEPC stacking is needed for stacking local=20
>objects needed by a local (intra-domain) QoS model that has=20
>different characteristics than an end-to-end QoS model=20
>together with the end-to-end objects needed by the end-to-end=20
>QoS model (see also requirement 5.4.2 in RFC3762).
>Examples of such intra-domain QOS models are models such as=20
>Diffserv based, MEF (Metro Ethernet Forum) based and 3GPP based.=20
>An example of an end-to-end QoS model is the Intserv model.

For stacking, is there any relation between the two QSPECs?  If
so, how are they related?  I am having trouble seeing the relation
between the two QSPECs.

thanks,
JOhn

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 07:17:07 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfxyE-0008Si-6a; Fri, 03 Nov 2006 07:16:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfxyD-0008PF-Ct
	for nsis@ietf.org; Fri, 03 Nov 2006 07:16:25 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfxy9-00072s-UF
	for nsis@ietf.org; Fri, 03 Nov 2006 07:16:25 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3CG0DR006383; Fri, 3 Nov 2006 13:16:14 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <john.loughney@nokia.com>, <gash@att.com>, <jmanner@cs.Helsinki.FI>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 13:15:57 +0100
Message-ID: <002101c6ff41$d786ca50$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <BAA65A575825454CBB0103267553FCCCED04FA@esebe199.NOE.Nokia.com>
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1AAAElaWAAATDhIAABeDxA
X-Spam-Score: 0.266 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 13:16:14 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b431ad66d60be2d47c7bfeb879db82c
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi John 

Of course the two QoS models are related to each other. So the end-to-end
QSPEC 
will activate some events and the local QoS model can include local QOS
parameters, in a
local QSPEC, please see requirement 5.4.2 in RFC3762.

Please note that a QSPEC can include besides the traffic description 
(i.e., Token Bucket, Bandwidth) also other QoS parameters that are needed 
for the local QoS model and are not needed for the end-to-end QoS model. 
In this case stacking is very useful, because the local QSPEC can be
interpreted 
only by the local QoS model and removed at the edges of the local QOSM aware
domain.

Best Regards,
Georgios


> -----Original Message-----
> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com] 
> Sent: vrijdag 3 november 2006 13:02
> To: karagian@cs.utwente.nl; gash@att.com; jmanner@cs.Helsinki.FI
> Cc: nsis@ietf.org
> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> 
> Georgios,
> 
> >QSPEC Stacking is not needed  for remapping.
> >For more details on QSPEC stacking please check Section 3.2.8 in:
> >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt
> >
> >Please note that QSEPC stacking is needed for stacking local objects 
> >needed by a local (intra-domain) QoS model that has different 
> >characteristics than an end-to-end QoS model together with the 
> >end-to-end objects needed by the end-to-end QoS model (see also 
> >requirement 5.4.2 in RFC3762).
> >Examples of such intra-domain QOS models are models such as Diffserv 
> >based, MEF (Metro Ethernet Forum) based and 3GPP based.
> >An example of an end-to-end QoS model is the Intserv model.
> 
> For stacking, is there any relation between the two QSPECs?  
> If so, how are they related?  I am having trouble seeing the 
> relation between the two QSPECs.
> 
> thanks,
> JOhn
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 07:29:45 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfyAR-0003mc-Ib; Fri, 03 Nov 2006 07:29:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfyAQ-0003mU-Ni
	for nsis@ietf.org; Fri, 03 Nov 2006 07:29:02 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfyAJ-0000kg-JL
	for nsis@ietf.org; Fri, 03 Nov 2006 07:29:02 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GfyAG-0003Ne-Se; Fri, 03 Nov 2006 13:28:54 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GfyAF-0003yC-2E; Fri, 03 Nov 2006 13:28:51 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 49E07B597;
	Fri,  3 Nov 2006 13:28:27 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gfy9q-0003nf-Rq; Fri, 03 Nov 2006 13:28:26 +0100
Message-ID: <454B35EA.3000608@tm.uka.de>
Date: Fri, 03 Nov 2006 13:28:26 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Shivanajay Marwaha <Shivanajay.Marwaha@sg.panasonic.com>
Subject: Re: [NSIS] [Fwd: I-D ACTION:draft-bless-nsis-resv-aggr-00.txt]
References: <5F09D220B62F79418461A978CA0921BD0148AB0E@pslexc01.psl.local>
In-Reply-To: <5F09D220B62F79418461A978CA0921BD0148AB0E@pslexc01.psl.local>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-15
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 52e1467c2184c31006318542db5614d5
Cc: Takako Sanda <sanda.takako@jp.panasonic.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Shivanajay,

Shivanajay Marwaha schrieb:
> Hi Roland,
> 
> I read your I-D and I have two questions:-

Thanks for reading the draft.

> 1. Are you also considering route change detection (in section 3.3)
> caused by end-host's mobility. 

Basically we consider any route change that affects the inter-domain
routes of existing flows no matter what was its cause.
However, end-host mobility will be mostly
transparent (i.e. not visible) because the host will move within a
domain (using HMIP, NetLMM, etc.). If it is moving between domains it
will usually get a new IP address, and, thus we would consider it to be
a new flow rather than a rerouted flow.

> 2. Based on the solution in section
> 4.4, in a multi-homing scenario (mentioned in section 3.4), will
> multiple Reserve messages be sent to multiple De-aggregators on
> individual paths?

No, you will always send only one RESERVE per flow.
There are different multi-homing solutions, so there are
different effects in case of failures depending on the
specific deployed solution.
If the MRI stays the same, a failure may cause a re-routing
to the alternative provider. The rerouting may or may not
affect the AS path, so sec. 3.3/4.3 applies.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 08:01:17 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfyeW-0006dI-Mq; Fri, 03 Nov 2006 08:00:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfyeU-0006by-DX
	for nsis@ietf.org; Fri, 03 Nov 2006 08:00:06 -0500
Received: from mail121.messagelabs.com ([216.82.241.195])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfyeO-0006t2-Nt
	for nsis@ietf.org; Fri, 03 Nov 2006 08:00:06 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-7.tower-121.messagelabs.com!1162558799!14045310!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 18176 invoked from network); 3 Nov 2006 12:59:59 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-7.tower-121.messagelabs.com with SMTP;
	3 Nov 2006 12:59:59 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA3CxxIb027481; 
	Fri, 3 Nov 2006 07:59:59 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA3CxsIc027458; Fri, 3 Nov 2006 07:59:54 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 06:59:53 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F26@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <001e01c6ff38$995c43d0$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQAAIJ7gAABDyLAA==
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "Georgios Karagiannis" <karagian@cs.utwente.nl>, <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a0534e6179a1e260079328e8b03c7901
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios,=20

> -----Original Message-----
> From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]=20
> Sent: Friday, November 03, 2006 6:10 AM
> To: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> Subject: RE: [NSIS] jerry, dave,cornelia's proposal very=20
> confusing and not solving problem
>=20
> Hi Jerry
>=20
> Thank you for the clarification!
> Regarding the interpretation of the token bucket parameter=20
> I have another problem on understanding your proposal.
>=20
> In your proposal you say:
> >The TMOD parameter is a mathematically complete way to describe a=20
> >traffic source.  Examples will be given for setting TMOD to=20
> correspond=20
> >to specific cases, e.g. where only bandwidth is available=20
> (r=3Dp; b and m=20
> >set to default values) and other cases (e.g., TCP).  As such, the=20
> >bandwidth parameter is eliminated.
>=20
> Does the above mean that in specific cases a QNE can=20
> interpret only the
> Token Bucket=20
> peak bandwidth parameter and ignore all the other Token=20
> Bucket parameters?
>=20
> 1) If the above is true, then how a QNI could know that a QNE=20
> used only the
> Token Bucket=20
> Peak bandwidth parameter and another QNE used all the Token Bucket
> parameters?
>=20
> 2) If the above is not true then it means that a QoS model=20
> that needs to use
> only=20
> the Token Bucket peak bandwidth parameter it will not be able to use=20
> the token bucket specification, thus it will not be able to use NSIS.
>=20
>=20
> Which of the two alternatives is the right one?

If the QNI intends that only bandwidth be specified to describe the
traffic source then it sets TMOD with r=3Dp and b/m to default values.  =
If
the QNI intends that a token bucket be specified to describe the traffic
source then it sets the 4 parameters accordingly. =20

The QNE is obliged to (strictly) interpret TMOD as specified by the QNI.
There is no remapping to override or replace the intent of the QNI.

Jerry

>=20
>=20
> Best Regards,
> Georgios
>=20
>=20
>=20
> =20
>=20
> > -----Original Message-----
> > From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com]=20
> > Sent: vrijdag 3 november 2006 11:39
> > To: Georgios Karagiannis; nsis@ietf.org
> > Cc: Ash, Gerald R (Jerry), NPE
> > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very=20
> > confusing and not solving problem
> >=20
> > Georgios,=20
> >=20
> > > -----Original Message-----
> > > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > > Sent: Friday, November 03, 2006 1:27 AM
> > > To: nsis@ietf.org
> > > Subject: [NSIS] jerry, dave,cornelia's proposal very=20
> > confusing and not=20
> > > solving problem
> > >=20
> > > Hi all
> > >=20
> > > After rereading the proposal of Jerry, Dave, Cornelia I=20
> > have seen that=20
> > > the main problem that we have had until now in the mailing=20
> > discussions=20
> > > is not solved. Mainly the number of mandatory parameters to=20
> > interpret=20
> > > by all the QNEs in the universe is not solved.
> >=20
> > It is solved.  There is only one mandatory parameter that=20
> > must be interpreted by all QNE/RMFs (traffic description =3D=20
> > TMOD).  All other parameters are optional to interpret and=20
> > can be ignored by a QNE/RMF.
> > I.e., there is only one 'qspec1' parameter (although=20
> > qspec1/qspec2 terminology is no longer used).
> > =20
> > > The proposal mentions in point 2) point B):
> > >=20
> > > "All other QSPEC parameters are optional to include by=20
> the QNI, and=20
> > > may be either mandatory or optional to interpret by=20
> > downstream QNEs."
> > >=20
> > > Meaning that if a QNI includes the Y.1541 class and wants=20
> that all=20
> > > downstream QNEs interpret this parameter, then all=20
> downstream QNEs=20
> > > must be able to strictly interpret the Y.1541 class.
> > > Jerry, Dave, Cornelia is this true?
> >=20
> > No.  If the Y.1541 parameter is populated by the QNI and the=20
> > QNE does not support the Y.1541 parameter then:
> > A) if the M flag is set by the QNI for the Y.1541 parameter,=20
> > the QNE sets the N flag for the Y.1541 parameter and rejects=20
> > the reservation
> > B) if the M flag is not set by the QNI for the Y.1541=20
> > parameter, the QNE sets the N flag for the Y.1541 parameter=20
> > and optionally either accepts or rejects the reservation=20
> > (i.e., the QNE can simply ignore the Y.1541 parameter).
> > =20
> > > Meaning that all QNEs that want to use NSIS must be=20
> implemented in=20
> > > such a way that if a QNI mandates the use of a parameter, e.g.,=20
> > > Y.1541, then the QNE must strictly interpret it, otherwise=20
> > it has to=20
> > > reject the parameter or the admission request.
> > >=20
> > > But this is exactly the problem that we wanted to solve by=20
> > having this=20
> > > discussions until now. Thus the problem is not solved.
> >=20
> > I disagree.  All parameters except TMOD can be=20
> > *not-supported* by a QNE.
> > If a QNI truly wants according to its QoS procedure being=20
> > implemented that a parameter must be interpreted or rejected,=20
> > then no other outcome makes sense other than to reject the=20
> > reservation if a parameter is not supported.  The QNI governs=20
> > the QoS outcome of the reservation.
> >=20
> > The proposal is also similar to an earlier suggestion (by=20
> > Roland?) that a QNI can decide which parameters are mandatory=20
> > to interpret.
> > =20
> > > Additionally, Jerry, Dave and Cornelia change the main=20
> > concepts of QoS=20
> > > based NSIS, please see my previous e-mails that I have sent=20
> > yesterday=20
> > > on this.
> > >=20
> > > I have already proposed a solution to solve the problems=20
> > that we have=20
> > > had in the nsis mailing list, see a copy of that proposal below:
> > > -------------------------
> > >=20
> > > Another proposal for solving the interoperability issue and=20
> > the issue=20
> > > of having too many parameters was to:
> > > * use only one mandatory parameter, i.e., Token Bucket that=20
> > could be=20
> > > remapped into the peak Bandwidth parameter.
> > > Thus except the Token Bucket parameter,
> > > * change the QSPEC-1 parameters into QSPEC-2 parameters
> >=20
> > It is equivalent.  If you change Y.1541 to qspec2 and a QNE=20
> > doesn't support Y.1541, it sets the N flag and ignores the=20
> > parameter.  This is the same as the proposal as long as the=20
> > QNI doesn't care whether Y.1541 is supported.  However, if=20
> > the QNI does care that Y.1541 is supported, it will=20
> > reject/tear the reservation if the N flag is set.  The result=20
> > is the same as the proposal!
> >=20
> > Jerry
> >=20
> > > * describe the remapping procedure of the Token Bucket into=20
> > the peak=20
> > > Bandwidth parameter in the QSPEC template draft.
> > >=20
> > > After that we can submit the QSPEC template draft for last=20
> > WG call. If=20
> > > after the deployment phase we conclude that more mandatory=20
> > parameters=20
> > > are needed for interoperability then we could work on an=20
> > interworking=20
> > > draft.
> > >=20
> > > --------------
> > >=20
> > > Best Regards,
> > > Georgios
> > >=20
> > > _______________________________________________
> > > nsis mailing list
> > > nsis@ietf.org
> > > https://www1.ietf.org/mailman/listinfo/nsis
> > >=20
> >=20
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> >=20
>=20
>=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 08:16:55 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfyu8-0008FV-77; Fri, 03 Nov 2006 08:16:16 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfyu3-0008AF-FR
	for nsis@ietf.org; Fri, 03 Nov 2006 08:16:11 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfynd-0008Lu-Bq
	for nsis@ietf.org; Fri, 03 Nov 2006 08:09:39 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3D9MNK008664; Fri, 3 Nov 2006 14:09:27 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, <nsis@ietf.org>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 14:09:20 +0100
Message-ID: <002b01c6ff49$494b87a0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F26@KCCLUST06EVS1.ugd.att.com>
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQAAIJ7gAABDyLAAAAyJ+g
X-Spam-Score: 0.421 () AWL,J_CHICKENPOX_48
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 14:09:28 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8068004c042dabd7f1301bcc80e039df
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

So the alternative 2) holds, meaning:
 2) If the above is not true then it means that a QoS model 
that needs to use only the Token Bucket peak bandwidth parameter it will not
be 
able to use the token bucket specification, thus it will not be able 
to use NSIS.

Thanks for the clarification!
This means that Diffserv based QoS models that need to use only bandwidth 
will not be able to use NSIS.

I really do not understand what kind of proposal is this??????


Best regards,
Georgios



> -----Original Message-----
> From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com] 
> Sent: vrijdag 3 november 2006 14:00
> To: Georgios Karagiannis; nsis@ietf.org
> Cc: Ash, Gerald R (Jerry), NPE
> Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing and not solving problem
> 
> Georgios, 
> 
> > -----Original Message-----
> > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > Sent: Friday, November 03, 2006 6:10 AM
> > To: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing and 
> > not solving problem
> > 
> > Hi Jerry
> > 
> > Thank you for the clarification!
> > Regarding the interpretation of the token bucket parameter I have 
> > another problem on understanding your proposal.
> > 
> > In your proposal you say:
> > >The TMOD parameter is a mathematically complete way to describe a 
> > >traffic source.  Examples will be given for setting TMOD to
> > correspond
> > >to specific cases, e.g. where only bandwidth is available
> > (r=p; b and m
> > >set to default values) and other cases (e.g., TCP).  As such, the 
> > >bandwidth parameter is eliminated.
> > 
> > Does the above mean that in specific cases a QNE can interpret only 
> > the Token Bucket peak bandwidth parameter and ignore all the other 
> > Token Bucket parameters?
> > 
> > 1) If the above is true, then how a QNI could know that a QNE used 
> > only the Token Bucket Peak bandwidth parameter and another QNE used 
> > all the Token Bucket parameters?
> > 
> > 2) If the above is not true then it means that a QoS model 
> that needs 
> > to use only the Token Bucket peak bandwidth parameter it 
> will not be 
> > able to use the token bucket specification, thus it will 
> not be able 
> > to use NSIS.
> > 
> > 
> > Which of the two alternatives is the right one?
> 
> If the QNI intends that only bandwidth be specified to 
> describe the traffic source then it sets TMOD with r=p and 
> b/m to default values.  If the QNI intends that a token 
> bucket be specified to describe the traffic source then it 
> sets the 4 parameters accordingly.  
> 
> The QNE is obliged to (strictly) interpret TMOD as specified 
> by the QNI.
> There is no remapping to override or replace the intent of the QNI.
> 
> Jerry
> 
> > 
> > 
> > Best Regards,
> > Georgios
> > 
> > 
> > 
> >  
> > 
> > > -----Original Message-----
> > > From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com]
> > > Sent: vrijdag 3 november 2006 11:39
> > > To: Georgios Karagiannis; nsis@ietf.org
> > > Cc: Ash, Gerald R (Jerry), NPE
> > > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing 
> > > and not solving problem
> > > 
> > > Georgios,
> > > 
> > > > -----Original Message-----
> > > > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > > > Sent: Friday, November 03, 2006 1:27 AM
> > > > To: nsis@ietf.org
> > > > Subject: [NSIS] jerry, dave,cornelia's proposal very
> > > confusing and not
> > > > solving problem
> > > > 
> > > > Hi all
> > > > 
> > > > After rereading the proposal of Jerry, Dave, Cornelia I
> > > have seen that
> > > > the main problem that we have had until now in the mailing
> > > discussions
> > > > is not solved. Mainly the number of mandatory parameters to
> > > interpret
> > > > by all the QNEs in the universe is not solved.
> > > 
> > > It is solved.  There is only one mandatory parameter that must be 
> > > interpreted by all QNE/RMFs (traffic description = TMOD). 
>  All other 
> > > parameters are optional to interpret and can be ignored by a 
> > > QNE/RMF.
> > > I.e., there is only one 'qspec1' parameter (although
> > > qspec1/qspec2 terminology is no longer used).
> > >  
> > > > The proposal mentions in point 2) point B):
> > > > 
> > > > "All other QSPEC parameters are optional to include by
> > the QNI, and
> > > > may be either mandatory or optional to interpret by
> > > downstream QNEs."
> > > > 
> > > > Meaning that if a QNI includes the Y.1541 class and wants
> > that all
> > > > downstream QNEs interpret this parameter, then all
> > downstream QNEs
> > > > must be able to strictly interpret the Y.1541 class.
> > > > Jerry, Dave, Cornelia is this true?
> > > 
> > > No.  If the Y.1541 parameter is populated by the QNI and the QNE 
> > > does not support the Y.1541 parameter then:
> > > A) if the M flag is set by the QNI for the Y.1541 
> parameter, the QNE 
> > > sets the N flag for the Y.1541 parameter and rejects the 
> reservation
> > > B) if the M flag is not set by the QNI for the Y.1541 
> parameter, the 
> > > QNE sets the N flag for the Y.1541 parameter and 
> optionally either 
> > > accepts or rejects the reservation (i.e., the QNE can 
> simply ignore 
> > > the Y.1541 parameter).
> > >  
> > > > Meaning that all QNEs that want to use NSIS must be
> > implemented in
> > > > such a way that if a QNI mandates the use of a parameter, e.g., 
> > > > Y.1541, then the QNE must strictly interpret it, otherwise
> > > it has to
> > > > reject the parameter or the admission request.
> > > > 
> > > > But this is exactly the problem that we wanted to solve by
> > > having this
> > > > discussions until now. Thus the problem is not solved.
> > > 
> > > I disagree.  All parameters except TMOD can be
> > > *not-supported* by a QNE.
> > > If a QNI truly wants according to its QoS procedure being 
> > > implemented that a parameter must be interpreted or 
> rejected, then 
> > > no other outcome makes sense other than to reject the 
> reservation if 
> > > a parameter is not supported.  The QNI governs the QoS outcome of 
> > > the reservation.
> > > 
> > > The proposal is also similar to an earlier suggestion (by
> > > Roland?) that a QNI can decide which parameters are mandatory to 
> > > interpret.
> > >  
> > > > Additionally, Jerry, Dave and Cornelia change the main
> > > concepts of QoS
> > > > based NSIS, please see my previous e-mails that I have sent
> > > yesterday
> > > > on this.
> > > > 
> > > > I have already proposed a solution to solve the problems
> > > that we have
> > > > had in the nsis mailing list, see a copy of that proposal below:
> > > > -------------------------
> > > > 
> > > > Another proposal for solving the interoperability issue and
> > > the issue
> > > > of having too many parameters was to:
> > > > * use only one mandatory parameter, i.e., Token Bucket that
> > > could be
> > > > remapped into the peak Bandwidth parameter.
> > > > Thus except the Token Bucket parameter,
> > > > * change the QSPEC-1 parameters into QSPEC-2 parameters
> > > 
> > > It is equivalent.  If you change Y.1541 to qspec2 and a 
> QNE doesn't 
> > > support Y.1541, it sets the N flag and ignores the 
> parameter.  This 
> > > is the same as the proposal as long as the QNI doesn't 
> care whether 
> > > Y.1541 is supported.  However, if the QNI does care that 
> Y.1541 is 
> > > supported, it will reject/tear the reservation if the N 
> flag is set.  
> > > The result is the same as the proposal!
> > > 
> > > Jerry
> > > 
> > > > * describe the remapping procedure of the Token Bucket into
> > > the peak
> > > > Bandwidth parameter in the QSPEC template draft.
> > > > 
> > > > After that we can submit the QSPEC template draft for last
> > > WG call. If
> > > > after the deployment phase we conclude that more mandatory
> > > parameters
> > > > are needed for interoperability then we could work on an
> > > interworking
> > > > draft.
> > > > 
> > > > --------------
> > > > 
> > > > Best Regards,
> > > > Georgios
> > > > 
> > > > _______________________________________________
> > > > nsis mailing list
> > > > nsis@ietf.org
> > > > https://www1.ietf.org/mailman/listinfo/nsis
> > > > 
> > > 
> > > _______________________________________________
> > > nsis mailing list
> > > nsis@ietf.org
> > > https://www1.ietf.org/mailman/listinfo/nsis
> > > 
> > 
> > 
> > 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 08:17:00 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfytx-00080M-C6; Fri, 03 Nov 2006 08:16:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfytu-0007sD-Kd
	for nsis@ietf.org; Fri, 03 Nov 2006 08:16:02 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gfytt-0000wq-0x
	for nsis@ietf.org; Fri, 03 Nov 2006 08:16:02 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3DFnHl008891; Fri, 3 Nov 2006 14:15:54 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, <john.loughney@nokia.com>,
	<jmanner@cs.Helsinki.FI>
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 14:15:47 +0100
Message-ID: <002c01c6ff4a$301cbb90$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F27@KCCLUST06EVS1.ugd.att.com>
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1AAAElaWAAATDhIAABeDxAAAGAsnAAAKPrcA==
X-Spam-Score: 0.273 () AWL,FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 14:15:55 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: c83ccb5cc10e751496398f1233ca9c3a
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

Another model that will need stacking is for example the ITU-T RACF model,
see:
http://www.ietf.org/internet-drafts/draft-zhang-nsis-interdomain-qosm-03.txt

I think however that your question is wrong!

NSIS is designed to allow modularity, such new QoS models can be 
developed in the future.
Today only three QoS models are specified and it does not mean 
that because two of them do not neeed to use stacking (Y.1541 and 
Controlled load) it does not mean that 
none of the future QoS models will not need to use stacking.

Best Regards,
Georgios

> -----Original Message-----
> From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com] 
> Sent: vrijdag 3 november 2006 14:01
> To: Georgios Karagiannis; john.loughney@nokia.com; 
> jmanner@cs.Helsinki.FI
> Cc: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> 
> Georgios, 
> 
> > -----Original Message-----
> > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > Sent: Friday, November 03, 2006 7:16 AM
> > To: john.loughney@nokia.com; Ash, Gerald R (Jerry), NPE; 
> > jmanner@cs.Helsinki.FI
> > Cc: nsis@ietf.org
> > Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> > 
> > Hi John
> > 
> > Of course the two QoS models are related to each other. So the 
> > end-to-end QSPEC will activate some events and the local 
> QoS model can 
> > include local QOS parameters, in a local QSPEC, please see 
> requirement 
> > 5.4.2 in RFC3762.
> > 
> > Please note that a QSPEC can include besides the traffic 
> description 
> > (i.e., Token Bucket, Bandwidth) also other QoS parameters that are 
> > needed for the local QoS model and are not needed for the 
> end-to-end 
> > QoS model.
> > In this case stacking is very useful, because the local 
> QSPEC can be 
> > interpreted only by the local QoS model and removed at the edges of 
> > the local QOSM aware domain.
> 
> You haven't given any concrete example of how the second 
> QSPEC should be used and why.
> 
> If you are referring to RMD using the second QSPEC to 
> piggyback certain control parameters, applicable only to RMD, 
> inside a local domain, then that isn't a justification for 
> stacking a QSPEC.
> 
> Please provide a concrete example of QSPEC stacking that 
> justifies the need, where no remapping of qspec parameters is allowed.
> 
> Jerry
> 
> > 
> > Best Regards,
> > Georgios
> > 
> > 
> > > -----Original Message-----
> > > From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> > > Sent: vrijdag 3 november 2006 13:02
> > > To: karagian@cs.utwente.nl; gash@att.com; jmanner@cs.Helsinki.FI
> > > Cc: nsis@ietf.org
> > > Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> > > 
> > > Georgios,
> > > 
> > > >QSPEC Stacking is not needed  for remapping.
> > > >For more details on QSPEC stacking please check Section 3.2.8 in:
> > > 
> >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt
> > > >
> > > >Please note that QSEPC stacking is needed for stacking
> > local objects
> > > >needed by a local (intra-domain) QoS model that has different 
> > > >characteristics than an end-to-end QoS model together with the 
> > > >end-to-end objects needed by the end-to-end QoS model (see also 
> > > >requirement 5.4.2 in RFC3762).
> > > >Examples of such intra-domain QOS models are models such
> > as Diffserv
> > > >based, MEF (Metro Ethernet Forum) based and 3GPP based.
> > > >An example of an end-to-end QoS model is the Intserv model.
> > > 
> > > For stacking, is there any relation between the two QSPECs?  
> > > If so, how are they related?  I am having trouble seeing the 
> > > relation between the two QSPECs.
> > > 
> > > thanks,
> > > JOhn
> > > 
> > 
> > 
> > 
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 08:18:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfyvT-00035z-Bo; Fri, 03 Nov 2006 08:17:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfyuL-0000IS-Jy
	for nsis@ietf.org; Fri, 03 Nov 2006 08:16:29 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gfyfp-00075N-3j
	for nsis@ietf.org; Fri, 03 Nov 2006 08:01:34 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-4.tower-146.messagelabs.com!1162558888!7959401!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 23270 invoked from network); 3 Nov 2006 13:01:28 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-4.tower-146.messagelabs.com with SMTP;
	3 Nov 2006 13:01:28 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA3D1SXv028299; 
	Fri, 3 Nov 2006 08:01:28 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA3D0jNx027988; Fri, 3 Nov 2006 08:01:20 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 07:01:08 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F27@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <002101c6ff41$d786ca50$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb/LTWIxDMr/BXfQ3SO01jlKziwkQABEk1AAAElaWAAATDhIAABeDxAAAGAsnA=
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "Georgios Karagiannis" <karagian@cs.utwente.nl>, <john.loughney@nokia.com>,
	<jmanner@cs.Helsinki.FI>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d185fa790257f526fedfd5d01ed9c976
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios,=20

> -----Original Message-----
> From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]=20
> Sent: Friday, November 03, 2006 7:16 AM
> To: john.loughney@nokia.com; Ash, Gerald R (Jerry), NPE;=20
> jmanner@cs.Helsinki.FI
> Cc: nsis@ietf.org
> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
>=20
> Hi John=20
>=20
> Of course the two QoS models are related to each other. So=20
> the end-to-end
> QSPEC=20
> will activate some events and the local QoS model can include=20
> local QOS
> parameters, in a
> local QSPEC, please see requirement 5.4.2 in RFC3762.
>=20
> Please note that a QSPEC can include besides the traffic description=20
> (i.e., Token Bucket, Bandwidth) also other QoS parameters=20
> that are needed=20
> for the local QoS model and are not needed for the end-to-end=20
> QoS model.=20
> In this case stacking is very useful, because the local QSPEC can be
> interpreted=20
> only by the local QoS model and removed at the edges of the=20
> local QOSM aware domain.

You haven't given any concrete example of how the second QSPEC should be
used and why.

If you are referring to RMD using the second QSPEC to piggyback certain
control parameters, applicable only to RMD, inside a local domain, then
that isn't a justification for stacking a QSPEC.

Please provide a concrete example of QSPEC stacking that justifies the
need, where no remapping of qspec parameters is allowed.

Jerry

>=20
> Best Regards,
> Georgios
>=20
>=20
> > -----Original Message-----
> > From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]=20
> > Sent: vrijdag 3 november 2006 13:02
> > To: karagian@cs.utwente.nl; gash@att.com; jmanner@cs.Helsinki.FI
> > Cc: nsis@ietf.org
> > Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
> >=20
> > Georgios,
> >=20
> > >QSPEC Stacking is not needed  for remapping.
> > >For more details on QSPEC stacking please check Section 3.2.8 in:
> > >http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt
> > >
> > >Please note that QSEPC stacking is needed for stacking=20
> local objects=20
> > >needed by a local (intra-domain) QoS model that has different=20
> > >characteristics than an end-to-end QoS model together with the=20
> > >end-to-end objects needed by the end-to-end QoS model (see also=20
> > >requirement 5.4.2 in RFC3762).
> > >Examples of such intra-domain QOS models are models such=20
> as Diffserv=20
> > >based, MEF (Metro Ethernet Forum) based and 3GPP based.
> > >An example of an end-to-end QoS model is the Intserv model.
> >=20
> > For stacking, is there any relation between the two QSPECs? =20
> > If so, how are they related?  I am having trouble seeing the=20
> > relation between the two QSPECs.
> >=20
> > thanks,
> > JOhn
> >=20
>=20
>=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 08:31:51 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gfz8T-0001J8-PL; Fri, 03 Nov 2006 08:31:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gfz8S-0001J2-Cf
	for nsis@ietf.org; Fri, 03 Nov 2006 08:31:04 -0500
Received: from mail120.messagelabs.com ([216.82.250.83])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gfz8P-0004C7-HP
	for nsis@ietf.org; Fri, 03 Nov 2006 08:31:04 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-12.tower-120.messagelabs.com!1162560659!16013626!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 11022 invoked from network); 3 Nov 2006 13:31:00 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-12.tower-120.messagelabs.com with SMTP;
	3 Nov 2006 13:31:00 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA3DUxT3011257; 
	Fri, 3 Nov 2006 08:30:59 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA3DUqA0011203; Fri, 3 Nov 2006 08:30:53 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 07:30:52 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F28@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <002b01c6ff49$494b87a0$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQAAIJ7gAABDyLAAAAyJ+gAABmOhA=
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: "Georgios Karagiannis" <karagian@cs.utwente.nl>, <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8a4bcf8f67063cac573319207fe3db35
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios,=20

> -----Original Message-----
> From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]=20
> Sent: Friday, November 03, 2006 8:09 AM
> To: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> Subject: RE: [NSIS] jerry, dave,cornelia's proposal very=20
> confusing and not solving problem
>=20
> Hi Jerry
>=20
> So the alternative 2) holds, meaning:
>  2) If the above is not true then it means that a QoS model=20
> that needs to use only the Token Bucket peak bandwidth=20
> parameter it will not
> be=20
> able to use the token bucket specification, thus it will not be able=20
> to use NSIS.
>=20
> Thanks for the clarification!
> This means that Diffserv based QoS models that need to use=20
> only bandwidth=20
> will not be able to use NSIS.

If a QNI wants to specify TMOD as bandwidth only, there is a way to do
that as I said (see below in my previous email).  This has nothing to do
with DiffServ, I think you mean it doesn't work with the way you've
specified RMD and therefore you don't like it.
=20
> I really do not understand what kind of proposal is this??????

Token bucket has always been a qspec1/mandatory parameter, and must be
interpreted as a token bucket.  That has not changed!

It seems you want RMD to ignore token bucket also and only have
bandwidth for RMD.

Please note, NSIS isn't being designed around RMD.  Please try to think
how RMD might work within the proposal rather than rejecting things that
don't immediately fit.  It might just be possible if you gave it a
little more thought.

Jerry
=20
>=20
> Best regards,
> Georgios
>=20
>=20
>=20
> > -----Original Message-----
> > From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com]=20
> > Sent: vrijdag 3 november 2006 14:00
> > To: Georgios Karagiannis; nsis@ietf.org
> > Cc: Ash, Gerald R (Jerry), NPE
> > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very=20
> > confusing and not solving problem
> >=20
> > Georgios,=20
> >=20
> > > -----Original Message-----
> > > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > > Sent: Friday, November 03, 2006 6:10 AM
> > > To: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> > > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very=20
> > confusing and=20
> > > not solving problem
> > >=20
> > > Hi Jerry
> > >=20
> > > Thank you for the clarification!
> > > Regarding the interpretation of the token bucket parameter I have=20
> > > another problem on understanding your proposal.
> > >=20
> > > In your proposal you say:
> > > >The TMOD parameter is a mathematically complete way to=20
> describe a=20
> > > >traffic source.  Examples will be given for setting TMOD to
> > > correspond
> > > >to specific cases, e.g. where only bandwidth is available
> > > (r=3Dp; b and m
> > > >set to default values) and other cases (e.g., TCP).  As=20
> such, the=20
> > > >bandwidth parameter is eliminated.
> > >=20
> > > Does the above mean that in specific cases a QNE can=20
> interpret only=20
> > > the Token Bucket peak bandwidth parameter and ignore all=20
> the other=20
> > > Token Bucket parameters?
> > >=20
> > > 1) If the above is true, then how a QNI could know that a=20
> QNE used=20
> > > only the Token Bucket Peak bandwidth parameter and=20
> another QNE used=20
> > > all the Token Bucket parameters?
> > >=20
> > > 2) If the above is not true then it means that a QoS model=20
> > that needs=20
> > > to use only the Token Bucket peak bandwidth parameter it=20
> > will not be=20
> > > able to use the token bucket specification, thus it will=20
> > not be able=20
> > > to use NSIS.
> > >=20
> > >=20
> > > Which of the two alternatives is the right one?
> >=20
> > If the QNI intends that only bandwidth be specified to=20
> > describe the traffic source then it sets TMOD with r=3Dp and=20
> > b/m to default values.  If the QNI intends that a token=20
> > bucket be specified to describe the traffic source then it=20
> > sets the 4 parameters accordingly. =20
> >=20
> > The QNE is obliged to (strictly) interpret TMOD as specified=20
> > by the QNI.
> > There is no remapping to override or replace the intent of the QNI.
> >=20
> > Jerry
> >=20
> > >=20
> > >=20
> > > Best Regards,
> > > Georgios
> > >=20
> > >=20
> > >=20
> > > =20
> > >=20
> > > > -----Original Message-----
> > > > From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com]
> > > > Sent: vrijdag 3 november 2006 11:39
> > > > To: Georgios Karagiannis; nsis@ietf.org
> > > > Cc: Ash, Gerald R (Jerry), NPE
> > > > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very=20
> > confusing=20
> > > > and not solving problem
> > > >=20
> > > > Georgios,
> > > >=20
> > > > > -----Original Message-----
> > > > > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > > > > Sent: Friday, November 03, 2006 1:27 AM
> > > > > To: nsis@ietf.org
> > > > > Subject: [NSIS] jerry, dave,cornelia's proposal very
> > > > confusing and not
> > > > > solving problem
> > > > >=20
> > > > > Hi all
> > > > >=20
> > > > > After rereading the proposal of Jerry, Dave, Cornelia I
> > > > have seen that
> > > > > the main problem that we have had until now in the mailing
> > > > discussions
> > > > > is not solved. Mainly the number of mandatory parameters to
> > > > interpret
> > > > > by all the QNEs in the universe is not solved.
> > > >=20
> > > > It is solved.  There is only one mandatory parameter=20
> that must be=20
> > > > interpreted by all QNE/RMFs (traffic description =3D TMOD).=20
> >  All other=20
> > > > parameters are optional to interpret and can be ignored by a=20
> > > > QNE/RMF.
> > > > I.e., there is only one 'qspec1' parameter (although
> > > > qspec1/qspec2 terminology is no longer used).
> > > > =20
> > > > > The proposal mentions in point 2) point B):
> > > > >=20
> > > > > "All other QSPEC parameters are optional to include by
> > > the QNI, and
> > > > > may be either mandatory or optional to interpret by
> > > > downstream QNEs."
> > > > >=20
> > > > > Meaning that if a QNI includes the Y.1541 class and wants
> > > that all
> > > > > downstream QNEs interpret this parameter, then all
> > > downstream QNEs
> > > > > must be able to strictly interpret the Y.1541 class.
> > > > > Jerry, Dave, Cornelia is this true?
> > > >=20
> > > > No.  If the Y.1541 parameter is populated by the QNI=20
> and the QNE=20
> > > > does not support the Y.1541 parameter then:
> > > > A) if the M flag is set by the QNI for the Y.1541=20
> > parameter, the QNE=20
> > > > sets the N flag for the Y.1541 parameter and rejects the=20
> > reservation
> > > > B) if the M flag is not set by the QNI for the Y.1541=20
> > parameter, the=20
> > > > QNE sets the N flag for the Y.1541 parameter and=20
> > optionally either=20
> > > > accepts or rejects the reservation (i.e., the QNE can=20
> > simply ignore=20
> > > > the Y.1541 parameter).
> > > > =20
> > > > > Meaning that all QNEs that want to use NSIS must be
> > > implemented in
> > > > > such a way that if a QNI mandates the use of a=20
> parameter, e.g.,=20
> > > > > Y.1541, then the QNE must strictly interpret it, otherwise
> > > > it has to
> > > > > reject the parameter or the admission request.
> > > > >=20
> > > > > But this is exactly the problem that we wanted to solve by
> > > > having this
> > > > > discussions until now. Thus the problem is not solved.
> > > >=20
> > > > I disagree.  All parameters except TMOD can be
> > > > *not-supported* by a QNE.
> > > > If a QNI truly wants according to its QoS procedure being=20
> > > > implemented that a parameter must be interpreted or=20
> > rejected, then=20
> > > > no other outcome makes sense other than to reject the=20
> > reservation if=20
> > > > a parameter is not supported.  The QNI governs the QoS=20
> outcome of=20
> > > > the reservation.
> > > >=20
> > > > The proposal is also similar to an earlier suggestion (by
> > > > Roland?) that a QNI can decide which parameters are=20
> mandatory to=20
> > > > interpret.
> > > > =20
> > > > > Additionally, Jerry, Dave and Cornelia change the main
> > > > concepts of QoS
> > > > > based NSIS, please see my previous e-mails that I have sent
> > > > yesterday
> > > > > on this.
> > > > >=20
> > > > > I have already proposed a solution to solve the problems
> > > > that we have
> > > > > had in the nsis mailing list, see a copy of that=20
> proposal below:
> > > > > -------------------------
> > > > >=20
> > > > > Another proposal for solving the interoperability issue and
> > > > the issue
> > > > > of having too many parameters was to:
> > > > > * use only one mandatory parameter, i.e., Token Bucket that
> > > > could be
> > > > > remapped into the peak Bandwidth parameter.
> > > > > Thus except the Token Bucket parameter,
> > > > > * change the QSPEC-1 parameters into QSPEC-2 parameters
> > > >=20
> > > > It is equivalent.  If you change Y.1541 to qspec2 and a=20
> > QNE doesn't=20
> > > > support Y.1541, it sets the N flag and ignores the=20
> > parameter.  This=20
> > > > is the same as the proposal as long as the QNI doesn't=20
> > care whether=20
> > > > Y.1541 is supported.  However, if the QNI does care that=20
> > Y.1541 is=20
> > > > supported, it will reject/tear the reservation if the N=20
> > flag is set. =20
> > > > The result is the same as the proposal!
> > > >=20
> > > > Jerry
> > > >=20
> > > > > * describe the remapping procedure of the Token Bucket into
> > > > the peak
> > > > > Bandwidth parameter in the QSPEC template draft.
> > > > >=20
> > > > > After that we can submit the QSPEC template draft for last
> > > > WG call. If
> > > > > after the deployment phase we conclude that more mandatory
> > > > parameters
> > > > > are needed for interoperability then we could work on an
> > > > interworking
> > > > > draft.
> > > > >=20
> > > > > --------------
> > > > >=20
> > > > > Best Regards,
> > > > > Georgios
> > > > >=20
> > > > > _______________________________________________
> > > > > nsis mailing list
> > > > > nsis@ietf.org
> > > > > https://www1.ietf.org/mailman/listinfo/nsis
> > > > >=20
> > > >=20
> > > > _______________________________________________
> > > > nsis mailing list
> > > > nsis@ietf.org
> > > > https://www1.ietf.org/mailman/listinfo/nsis
> > > >=20
> > >=20
> > >=20
> > >=20
> >=20
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> >=20
>=20
>=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 08:39:15 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfzFd-00082B-Nr; Fri, 03 Nov 2006 08:38:29 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfzFc-000823-78
	for nsis@ietf.org; Fri, 03 Nov 2006 08:38:28 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GfzFZ-0005LG-IJ
	for nsis@ietf.org; Fri, 03 Nov 2006 08:38:28 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3Dc3UH009682; Fri, 3 Nov 2006 14:38:06 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, <nsis@ietf.org>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 14:38:00 +0100
Message-ID: <002f01c6ff4d$4af40ab0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F28@KCCLUST06EVS1.ugd.att.com>
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQAAIJ7gAABDyLAAAAyJ+gAABmOhAAAJkeQA==
X-Spam-Score: 0.424 () AWL,J_CHICKENPOX_48
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 14:38:08 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 32a65c0bf5eb4ec26489239c7cdd0636
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

RMD-QOSM used the NSIS features that have been approved by NSIS WG and IESG,
see 
RFC3762.

Your proposal is not right, since it does not comply to RFC 3762.

So maybe you should think on redrawing your proposal!


Best Regards,
Georgios


> -----Original Message-----
> From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com] 
> Sent: vrijdag 3 november 2006 14:31
> To: Georgios Karagiannis; nsis@ietf.org
> Cc: Ash, Gerald R (Jerry), NPE
> Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing and not solving problem
> 
> Georgios, 
> 
> > -----Original Message-----
> > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > Sent: Friday, November 03, 2006 8:09 AM
> > To: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing and 
> > not solving problem
> > 
> > Hi Jerry
> > 
> > So the alternative 2) holds, meaning:
> >  2) If the above is not true then it means that a QoS model 
> that needs 
> > to use only the Token Bucket peak bandwidth parameter it 
> will not be 
> > able to use the token bucket specification, thus it will 
> not be able 
> > to use NSIS.
> > 
> > Thanks for the clarification!
> > This means that Diffserv based QoS models that need to use only 
> > bandwidth will not be able to use NSIS.
> 
> If a QNI wants to specify TMOD as bandwidth only, there is a 
> way to do that as I said (see below in my previous email).  
> This has nothing to do with DiffServ, I think you mean it 
> doesn't work with the way you've specified RMD and therefore 
> you don't like it.
>  
> > I really do not understand what kind of proposal is this??????
> 
> Token bucket has always been a qspec1/mandatory parameter, 
> and must be interpreted as a token bucket.  That has not changed!
> 
> It seems you want RMD to ignore token bucket also and only 
> have bandwidth for RMD.
> 
> Please note, NSIS isn't being designed around RMD.  Please 
> try to think how RMD might work within the proposal rather 
> than rejecting things that don't immediately fit.  It might 
> just be possible if you gave it a little more thought.
> 
> Jerry
>  
> > 
> > Best regards,
> > Georgios
> > 
> > 
> > 
> > > -----Original Message-----
> > > From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com]
> > > Sent: vrijdag 3 november 2006 14:00
> > > To: Georgios Karagiannis; nsis@ietf.org
> > > Cc: Ash, Gerald R (Jerry), NPE
> > > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing 
> > > and not solving problem
> > > 
> > > Georgios,
> > > 
> > > > -----Original Message-----
> > > > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > > > Sent: Friday, November 03, 2006 6:10 AM
> > > > To: Ash, Gerald R (Jerry), NPE; nsis@ietf.org
> > > > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very
> > > confusing and
> > > > not solving problem
> > > > 
> > > > Hi Jerry
> > > > 
> > > > Thank you for the clarification!
> > > > Regarding the interpretation of the token bucket 
> parameter I have 
> > > > another problem on understanding your proposal.
> > > > 
> > > > In your proposal you say:
> > > > >The TMOD parameter is a mathematically complete way to
> > describe a
> > > > >traffic source.  Examples will be given for setting TMOD to
> > > > correspond
> > > > >to specific cases, e.g. where only bandwidth is available
> > > > (r=p; b and m
> > > > >set to default values) and other cases (e.g., TCP).  As
> > such, the
> > > > >bandwidth parameter is eliminated.
> > > > 
> > > > Does the above mean that in specific cases a QNE can
> > interpret only
> > > > the Token Bucket peak bandwidth parameter and ignore all
> > the other
> > > > Token Bucket parameters?
> > > > 
> > > > 1) If the above is true, then how a QNI could know that a
> > QNE used
> > > > only the Token Bucket Peak bandwidth parameter and
> > another QNE used
> > > > all the Token Bucket parameters?
> > > > 
> > > > 2) If the above is not true then it means that a QoS model
> > > that needs
> > > > to use only the Token Bucket peak bandwidth parameter it
> > > will not be
> > > > able to use the token bucket specification, thus it will
> > > not be able
> > > > to use NSIS.
> > > > 
> > > > 
> > > > Which of the two alternatives is the right one?
> > > 
> > > If the QNI intends that only bandwidth be specified to 
> describe the 
> > > traffic source then it sets TMOD with r=p and b/m to 
> default values.  
> > > If the QNI intends that a token bucket be specified to 
> describe the 
> > > traffic source then it sets the 4 parameters accordingly.
> > > 
> > > The QNE is obliged to (strictly) interpret TMOD as 
> specified by the 
> > > QNI.
> > > There is no remapping to override or replace the intent 
> of the QNI.
> > > 
> > > Jerry
> > > 
> > > > 
> > > > 
> > > > Best Regards,
> > > > Georgios
> > > > 
> > > > 
> > > > 
> > > >  
> > > > 
> > > > > -----Original Message-----
> > > > > From: Ash, Gerald R (Jerry), NPE [mailto:gash@att.com]
> > > > > Sent: vrijdag 3 november 2006 11:39
> > > > > To: Georgios Karagiannis; nsis@ietf.org
> > > > > Cc: Ash, Gerald R (Jerry), NPE
> > > > > Subject: RE: [NSIS] jerry, dave,cornelia's proposal very
> > > confusing
> > > > > and not solving problem
> > > > > 
> > > > > Georgios,
> > > > > 
> > > > > > -----Original Message-----
> > > > > > From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]
> > > > > > Sent: Friday, November 03, 2006 1:27 AM
> > > > > > To: nsis@ietf.org
> > > > > > Subject: [NSIS] jerry, dave,cornelia's proposal very
> > > > > confusing and not
> > > > > > solving problem
> > > > > > 
> > > > > > Hi all
> > > > > > 
> > > > > > After rereading the proposal of Jerry, Dave, Cornelia I
> > > > > have seen that
> > > > > > the main problem that we have had until now in the mailing
> > > > > discussions
> > > > > > is not solved. Mainly the number of mandatory parameters to
> > > > > interpret
> > > > > > by all the QNEs in the universe is not solved.
> > > > > 
> > > > > It is solved.  There is only one mandatory parameter
> > that must be
> > > > > interpreted by all QNE/RMFs (traffic description = TMOD). 
> > >  All other
> > > > > parameters are optional to interpret and can be ignored by a 
> > > > > QNE/RMF.
> > > > > I.e., there is only one 'qspec1' parameter (although
> > > > > qspec1/qspec2 terminology is no longer used).
> > > > >  
> > > > > > The proposal mentions in point 2) point B):
> > > > > > 
> > > > > > "All other QSPEC parameters are optional to include by
> > > > the QNI, and
> > > > > > may be either mandatory or optional to interpret by
> > > > > downstream QNEs."
> > > > > > 
> > > > > > Meaning that if a QNI includes the Y.1541 class and wants
> > > > that all
> > > > > > downstream QNEs interpret this parameter, then all
> > > > downstream QNEs
> > > > > > must be able to strictly interpret the Y.1541 class.
> > > > > > Jerry, Dave, Cornelia is this true?
> > > > > 
> > > > > No.  If the Y.1541 parameter is populated by the QNI
> > and the QNE
> > > > > does not support the Y.1541 parameter then:
> > > > > A) if the M flag is set by the QNI for the Y.1541
> > > parameter, the QNE
> > > > > sets the N flag for the Y.1541 parameter and rejects the
> > > reservation
> > > > > B) if the M flag is not set by the QNI for the Y.1541
> > > parameter, the
> > > > > QNE sets the N flag for the Y.1541 parameter and
> > > optionally either
> > > > > accepts or rejects the reservation (i.e., the QNE can
> > > simply ignore
> > > > > the Y.1541 parameter).
> > > > >  
> > > > > > Meaning that all QNEs that want to use NSIS must be
> > > > implemented in
> > > > > > such a way that if a QNI mandates the use of a
> > parameter, e.g.,
> > > > > > Y.1541, then the QNE must strictly interpret it, otherwise
> > > > > it has to
> > > > > > reject the parameter or the admission request.
> > > > > > 
> > > > > > But this is exactly the problem that we wanted to solve by
> > > > > having this
> > > > > > discussions until now. Thus the problem is not solved.
> > > > > 
> > > > > I disagree.  All parameters except TMOD can be
> > > > > *not-supported* by a QNE.
> > > > > If a QNI truly wants according to its QoS procedure being 
> > > > > implemented that a parameter must be interpreted or
> > > rejected, then
> > > > > no other outcome makes sense other than to reject the
> > > reservation if
> > > > > a parameter is not supported.  The QNI governs the QoS
> > outcome of
> > > > > the reservation.
> > > > > 
> > > > > The proposal is also similar to an earlier suggestion (by
> > > > > Roland?) that a QNI can decide which parameters are
> > mandatory to
> > > > > interpret.
> > > > >  
> > > > > > Additionally, Jerry, Dave and Cornelia change the main
> > > > > concepts of QoS
> > > > > > based NSIS, please see my previous e-mails that I have sent
> > > > > yesterday
> > > > > > on this.
> > > > > > 
> > > > > > I have already proposed a solution to solve the problems
> > > > > that we have
> > > > > > had in the nsis mailing list, see a copy of that
> > proposal below:
> > > > > > -------------------------
> > > > > > 
> > > > > > Another proposal for solving the interoperability issue and
> > > > > the issue
> > > > > > of having too many parameters was to:
> > > > > > * use only one mandatory parameter, i.e., Token Bucket that
> > > > > could be
> > > > > > remapped into the peak Bandwidth parameter.
> > > > > > Thus except the Token Bucket parameter,
> > > > > > * change the QSPEC-1 parameters into QSPEC-2 parameters
> > > > > 
> > > > > It is equivalent.  If you change Y.1541 to qspec2 and a
> > > QNE doesn't
> > > > > support Y.1541, it sets the N flag and ignores the
> > > parameter.  This
> > > > > is the same as the proposal as long as the QNI doesn't
> > > care whether
> > > > > Y.1541 is supported.  However, if the QNI does care that
> > > Y.1541 is
> > > > > supported, it will reject/tear the reservation if the N
> > > flag is set.  
> > > > > The result is the same as the proposal!
> > > > > 
> > > > > Jerry
> > > > > 
> > > > > > * describe the remapping procedure of the Token Bucket into
> > > > > the peak
> > > > > > Bandwidth parameter in the QSPEC template draft.
> > > > > > 
> > > > > > After that we can submit the QSPEC template draft for last
> > > > > WG call. If
> > > > > > after the deployment phase we conclude that more mandatory
> > > > > parameters
> > > > > > are needed for interoperability then we could work on an
> > > > > interworking
> > > > > > draft.
> > > > > > 
> > > > > > --------------
> > > > > > 
> > > > > > Best Regards,
> > > > > > Georgios
> > > > > > 
> > > > > > _______________________________________________
> > > > > > nsis mailing list
> > > > > > nsis@ietf.org
> > > > > > https://www1.ietf.org/mailman/listinfo/nsis
> > > > > > 
> > > > > 
> > > > > _______________________________________________
> > > > > nsis mailing list
> > > > > nsis@ietf.org
> > > > > https://www1.ietf.org/mailman/listinfo/nsis
> > > > > 
> > > > 
> > > > 
> > > > 
> > > 
> > > _______________________________________________
> > > nsis mailing list
> > > nsis@ietf.org
> > > https://www1.ietf.org/mailman/listinfo/nsis
> > > 
> > 
> > 
> > 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 08:41:35 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GfzI0-0000Sd-Gj; Fri, 03 Nov 2006 08:40:56 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GfzHz-0000SX-0i
	for nsis@ietf.org; Fri, 03 Nov 2006 08:40:55 -0500
Received: from mail.gmx.net ([213.165.64.20])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GfzHp-0005h0-H8
	for nsis@ietf.org; Fri, 03 Nov 2006 08:40:54 -0500
Received: (qmail invoked by alias); 03 Nov 2006 13:40:43 -0000
Received: from unknown (EHLO [172.16.68.106]) [83.210.50.6]
	by mail.gmx.net (mp039) with SMTP; 03 Nov 2006 14:40:43 +0100
X-Authenticated: #29516787
Message-ID: <454B46D3.1000900@gmx.net>
Date: Fri, 03 Nov 2006 14:40:35 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.7 (Windows/20060909)
MIME-Version: 1.0
To: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F20@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F20@KCCLUST06EVS1.ugd.att.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 79899194edc4f33a41f49410777972f8
Cc: NSIS <nsis@ietf.org>
Subject: [NSIS] QSPEC Face-to-Face Chat: Which Reception?
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry,

it's good that you know the place very well.

Let's pick the main hotel reception where the conference rooms are.

Ciao
Hannes

Ash, Gerald R (Jerry), NPE wrote:
>> What about Sunday, 3pm at the Sheraton San Diego Hotel (= 
>> meeting venue) in front of the hotel reception.
> 
> OK for me.  But please clarify which reception we should meet at (there
> are 2), there is the Annex reception where most people are staying and
> the main hotel reception (where the conference rooms are), which is it?
> 
> Thanks,
> Jerry


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 09:59:19 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg0V1-0003uj-KV; Fri, 03 Nov 2006 09:58:27 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg0V0-0003ub-Hv
	for nsis@ietf.org; Fri, 03 Nov 2006 09:58:26 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg0Uu-00079U-Lq
	for nsis@ietf.org; Fri, 03 Nov 2006 09:58:26 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gg0Uc-0005mO-PM; Fri, 03 Nov 2006 15:58:03 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Gg0Ub-0002zv-Ab; Fri, 03 Nov 2006 15:58:01 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 2E551B4ED;
	Fri,  3 Nov 2006 15:57:56 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gg0UV-00040b-Oh; Fri, 03 Nov 2006 15:57:55 +0100
Message-ID: <454B58F3.4070709@tm.uka.de>
Date: Fri, 03 Nov 2006 15:57:55 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] jerry, dave,	cornelia's proposal very confusing and not
	solving problem
References: <002b01c6ff49$494b87a0$4c0d5982@dynamic.ewi.utwente.nl>
In-Reply-To: <002b01c6ff49$494b87a0$4c0d5982@dynamic.ewi.utwente.nl>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 39bd8f8cbb76cae18b7e23f7cf6b2b9f
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

Georgios Karagiannis wrote:
> So the alternative 2) holds, meaning:
>  2) If the above is not true then it means that a QoS model 
> that needs to use only the Token Bucket peak bandwidth parameter it will not
> be 
> able to use the token bucket specification, thus it will not be able 
> to use NSIS.
> 
> Thanks for the clarification!
> This means that Diffserv based QoS models that need to use only bandwidth 
> will not be able to use NSIS.

I guess that there is a lack of understanding DiffServ here.
DiffServ is not restricted to EF-based services only. If you
perform admission control based on the peak rate only, it's
your engineering decision. Some other people may take a look
at the whole token bucket parameters and achieve higher utilization.
It's up to the signaled constraints and your local QoS model.
For other services you might be more interested in the burstiness of the
traffic, too. Depending on the constraints that are signaled,
you are free to map the request into any appropriate DiffServ
class (PDB/PHB set) that is able to meet the signaled constraints.
So I guess the new proposal is very well suited to support also
DiffServ-based QoS support in domains.

> I really do not understand what kind of proposal is this??????

If you care about stacking:
Why not defining your own QSPEC that allows to carry your
domain-specific parameters as well as the original encapsulated
external QSPEC. This would be a form of encapsulation, but it
could probably be used instead of stacking.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 10:03:14 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg0Z3-0001Es-Aw; Fri, 03 Nov 2006 10:02:37 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg0Z1-0001En-NS
	for nsis@ietf.org; Fri, 03 Nov 2006 10:02:35 -0500
Received: from mgw-ext11.nokia.com ([131.228.20.170])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg0Yx-0008H9-8N
	for nsis@ietf.org; Fri, 03 Nov 2006 10:02:35 -0500
Received: from esebh107.NOE.Nokia.com (esebh107.ntc.nokia.com [172.21.143.143])
	by mgw-ext11.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA3F1grL000541; Fri, 3 Nov 2006 17:01:46 +0200
Received: from esebh104.NOE.Nokia.com ([172.21.143.34]) by
	esebh107.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 17:01:42 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh104.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 17:01:42 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 17:01:16 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED050C@esebe199.NOE.Nokia.com>
In-Reply-To: <002f01c6ff4d$4af40ab0$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQAAIJ7gAABDyLAAAAyJ+gAABmOhAAAJkeQAAC5aKQ
From: <john.loughney@nokia.com>
To: <karagian@cs.utwente.nl>, <gash@att.com>, <nsis@ietf.org>
X-OriginalArrivalTime: 03 Nov 2006 15:01:42.0087 (UTC)
	FILETIME=[F8AC1570:01C6FF58]
X-eXpurgate-Category: 1/0
X-eXpurgate-ID: 149371::061103170146-01A8ABB0-34EE554B/0-0/0-0
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: cf4fa59384e76e63313391b70cd0dd25
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios,

>RMD-QOSM used the NSIS features that have been approved by=20
>NSIS WG and IESG, see RFC3762.
>
>Your proposal is not right, since it does not comply to RFC 3762.
>
>So maybe you should think on redrawing your proposal!

RFC3762 is "Telephone Number Mapping (ENUM) Service Registration for
H.323"
Which document do you really mean?

http://www.ietf.org/rfc/rfc3762.txt?number=3D3762

John

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 10:18:42 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg0nr-0002ej-MZ; Fri, 03 Nov 2006 10:17:55 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg0nq-0002ca-M2
	for nsis@ietf.org; Fri, 03 Nov 2006 10:17:54 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg0nl-0005lv-Hv
	for nsis@ietf.org; Fri, 03 Nov 2006 10:17:54 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3FHRt3013180; Fri, 3 Nov 2006 16:17:44 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <john.loughney@nokia.com>, <gash@att.com>, <nsis@ietf.org>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 16:17:24 +0100
Message-ID: <005a01c6ff5b$2e2ebe80$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <BAA65A575825454CBB0103267553FCCCED050C@esebe199.NOE.Nokia.com>
Thread-Index: Acb/Ec8imeDt6VL6TceQKERKsP/7yAAGq9gQAAIJ7gAABDyLAAAAyJ+gAABmOhAAAJkeQAAC5aKQAACw8dA=
X-Spam-Score: 0.42 () AWL,J_CHICKENPOX_48
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 16:17:44 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7baded97d9887f7a0c7e8a33c2e3ea1b
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi John

Sorry, I meant RFC 3726, see below:
http://www.ietf.org/rfc/rfc3726.txt
 
Best Regards,
Georgios

> -----Original Message-----
> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com] 
> Sent: vrijdag 3 november 2006 16:01
> To: karagian@cs.utwente.nl; gash@att.com; nsis@ietf.org
> Subject: RE: [NSIS] jerry, dave,cornelia's proposal very 
> confusing and not solving problem
> 
> Georgios,
> 
> >RMD-QOSM used the NSIS features that have been approved by 
> NSIS WG and 
> >IESG, see RFC3762.
> >
> >Your proposal is not right, since it does not comply to RFC 3762.
> >
> >So maybe you should think on redrawing your proposal!
> 
> RFC3762 is "Telephone Number Mapping (ENUM) Service 
> Registration for H.323"
> Which document do you really mean?
> 
> http://www.ietf.org/rfc/rfc3762.txt?number=3762
> 
> John
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 10:27:43 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg0wk-000342-DJ; Fri, 03 Nov 2006 10:27:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg0wj-00033x-J7
	for nsis@ietf.org; Fri, 03 Nov 2006 10:27:05 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg0wi-00011T-3D
	for nsis@ietf.org; Fri, 03 Nov 2006 10:27:05 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gg0wZ-0004Ei-Ty; Fri, 03 Nov 2006 16:26:56 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Gg0wW-0006yK-86; Fri, 03 Nov 2006 16:26:52 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 71955B4EC;
	Fri,  3 Nov 2006 16:26:11 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gg0vr-00042H-6L; Fri, 03 Nov 2006 16:26:11 +0100
Message-ID: <454B5F92.4090505@tm.uka.de>
Date: Fri, 03 Nov 2006 16:26:10 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Lars Westberg <Lars.westberg@ericsson.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>	<F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>
	<454B1268.4030405@ericsson.com>
In-Reply-To: <454B1268.4030405@ericsson.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: c1c65599517f9ac32519d043c37c5336
Cc: john.loughney@nokia.com, nsis@ietf.org, "Attila BÃ¡"@ira.uka.de, David R Oran <oran@cisco.com>, Georgios Karagiannis <karagian@cs.utwente.nl>, "der \(IJ/ETH\)" <attila.bader@ericsson.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Lars,

Lars Westberg wrote:
> the main reason for bringing the discussion of parameters has nothing to
> with RMD. It is an arhitectural limitation. Just compared it to RSVP.
> RSVP do not have any hardcoded QoS-paraneters into the core part of the
> protocol specification, while NSIS has????

No, NSIS signaling carries the QSPEC which is opaque for the protocol.

> Given the present proposal in the QoS-spec, the architecture is much
> less open for new use than RSVP. Should NSIS make the same mistake as RSVP ?

It is not less open.

> comment about RMD:
> 
> the design target for RMD is twofold:
>     a) support for PCN-like bandwidth management.
>     b) scalable per-flow reservation.
> 
> during many years of discussion, we can conclude that additional
> functionality like RSVP-aggregation have ben design to cope with the
> state-explosion for per-flow reservation. RMD is developed to solve it
> inother approach: scalable per-flow reseravtion with the simillar
> resource guarantee as RSVP.

Actually I don't see why RMD is more scalable than RSVP aggregation
(at least if reservation-based mode of RMD is considered).
If compared to RSVP aggregation, RMD generates more signaling messages.
If I understood the RMD-draft correctly, every inter-domain/external
RESERVE message causes the QNE Ingress node to send an intra-Domain
RESERVE. Thus, interior nodes have to process per-(micro)flow signaling
messages which is avoided in the RSVP aggregation if the aggregate
reservation capacity is not adapted with every flow.

> Our target was actually to improve RSVP more improved scalabillity to
> solve a problem within the mobile access network. But, NSIS becomes the
> answer to the standardization discission so we introduced it in this
> context instead.

What is the intended status of RMD? Informational or PS?

Regards,
 Roland



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 10:28:24 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg0xP-0003MS-Pd; Fri, 03 Nov 2006 10:27:47 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg0xO-0003MN-AJ
	for nsis@ietf.org; Fri, 03 Nov 2006 10:27:46 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg0xM-0001IH-Sb
	for nsis@ietf.org; Fri, 03 Nov 2006 10:27:46 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3FRYRf013464; Fri, 3 Nov 2006 16:27:39 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Roland Bless'" <bless@tm.uka.de>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and notsolving problem
Date: Fri, 3 Nov 2006 16:27:32 +0100
Message-ID: <005b01c6ff5c$98242ea0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <454B58F3.4070709@tm.uka.de>
Thread-Index: Acb/WML5k1vqg49sREiDpOJAkjc/EwAAn7bQ
X-Spam-Score: 0.122 () AWL
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 16:27:40 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b7b9551d71acde901886cc48bfc088a6
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland

Please see in line!
 

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: vrijdag 3 november 2006 15:58
> To: Georgios Karagiannis
> Cc: 'Ash, Gerald R (Jerry), NPE'; nsis@ietf.org
> Subject: Re: [NSIS] jerry, dave, cornelia's proposal very 
> confusing and notsolving problem
> 
> Hi Georgios,
> 
> Georgios Karagiannis wrote:
> > So the alternative 2) holds, meaning:
> >  2) If the above is not true then it means that a QoS model 
> that needs 
> > to use only the Token Bucket peak bandwidth parameter it 
> will not be 
> > able to use the token bucket specification, thus it will 
> not be able 
> > to use NSIS.
> > 
> > Thanks for the clarification!
> > This means that Diffserv based QoS models that need to use only 
> > bandwidth will not be able to use NSIS.
> 
> I guess that there is a lack of understanding DiffServ here.
> DiffServ is not restricted to EF-based services only. If you 
> perform admission control based on the peak rate only, it's 
> your engineering decision. Some other people may take a look 
> at the whole token bucket parameters and achieve higher utilization.
> It's up to the signaled constraints and your local QoS model.
> For other services you might be more interested in the 
> burstiness of the traffic, too. Depending on the constraints 
> that are signaled, you are free to map the request into any 
> appropriate DiffServ class (PDB/PHB set) that is able to meet 
> the signaled constraints.
> So I guess the new proposal is very well suited to support 
> also DiffServ-based QoS support in domains.

Georgios: I agree, but it can not be used in all types of Diffserv based
domains.
The proposal restricts the possibility on performing admission control 
based on the peak rate only! That is my problem!
The proposal does not allow this kind of flexibility that 
we wanted to have in NSIS.
I remember some severe discussions on this point some years ago where we
came up
with the possibility on supporting both Token Bucket and Bandwidth
parameters.


> 
> > I really do not understand what kind of proposal is this??????
> 
> If you care about stacking:
> Why not defining your own QSPEC that allows to carry your 
> domain-specific parameters as well as the original 
> encapsulated external QSPEC. This would be a form of 
> encapsulation, but it could probably be used instead of stacking.

Georgios: 
This is what I want to do with stacking. Stack/encapsulate a RMD-QSPEC above
an end-to-end 
QSPEC. This is however not allowed in this proposal.

Best Regards,
Georgios


> 
> Regards,
>  Roland
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 11:12:49 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg1du-0004NR-I5; Fri, 03 Nov 2006 11:11:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg1dt-0004NM-7R
	for nsis@ietf.org; Fri, 03 Nov 2006 11:11:41 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg1do-0000k6-Pb
	for nsis@ietf.org; Fri, 03 Nov 2006 11:11:41 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gg1dl-0006Hb-EO; Fri, 03 Nov 2006 17:11:33 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Gg1dl-0005D5-3W; Fri, 03 Nov 2006 17:11:33 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id A873FB5AF;
	Fri,  3 Nov 2006 17:11:32 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gg1dk-00044w-98; Fri, 03 Nov 2006 17:11:32 +0100
Message-ID: <454B6A33.8030505@tm.uka.de>
Date: Fri, 03 Nov 2006 17:11:31 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] jerry, dave,
	cornelia's proposal very confusing and notsolving problem
References: <005b01c6ff5c$98242ea0$4c0d5982@dynamic.ewi.utwente.nl>
In-Reply-To: <005b01c6ff5c$98242ea0$4c0d5982@dynamic.ewi.utwente.nl>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 97adf591118a232206bdb5a27b217034
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

Georgios Karagiannis schrieb:
> Georgios: I agree, but it can not be used in all types of Diffserv based
> domains.
> The proposal restricts the possibility on performing admission control 
> based on the peak rate only! That is my problem!
> The proposal does not allow this kind of flexibility that 
> we wanted to have in NSIS.
> I remember some severe discussions on this point some years ago where we
> came up
> with the possibility on supporting both Token Bucket and Bandwidth
> parameters.

I can only agree with Dave here, that bandwidth only
without specifying any time period is not really usable.

>>> I really do not understand what kind of proposal is this??????
>> If you care about stacking:
>> Why not defining your own QSPEC that allows to carry your 
>> domain-specific parameters as well as the original 
>> encapsulated external QSPEC. This would be a form of 
>> encapsulation, but it could probably be used instead of stacking.
> 
> Georgios: 
> This is what I want to do with stacking. Stack/encapsulate a RMD-QSPEC above
> an end-to-end 
> QSPEC. This is however not allowed in this proposal.

It is allowed if you define your own RMD-QSPEC
with an encapsulation method for an e2e QSPEC.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 11:13:36 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg1f9-00055i-FM; Fri, 03 Nov 2006 11:12:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg1f8-00054E-8e
	for nsis@ietf.org; Fri, 03 Nov 2006 11:12:58 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg1f5-0000pL-Iw
	for nsis@ietf.org; Fri, 03 Nov 2006 11:12:58 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gg1ev-0006Vk-Ut; Fri, 03 Nov 2006 17:12:49 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Gg1ev-0005LR-30; Fri, 03 Nov 2006 17:12:45 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 061AEB5AF;
	Fri,  3 Nov 2006 17:12:44 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gg1et-000450-O8; Fri, 03 Nov 2006 17:12:43 +0100
Message-ID: <454B6A7A.6000407@tm.uka.de>
Date: Fri, 03 Nov 2006 17:12:42 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Lars Westberg <Lars.westberg@ericsson.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>	<F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>
	<454B1268.4030405@ericsson.com>
In-Reply-To: <454B1268.4030405@ericsson.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 769a46790fb42fbb0b0cc700c82f7081
Cc: john.loughney@nokia.com, Georgios Karagiannis <karagian@cs.utwente.nl>,
	David R Oran <oran@cisco.com>, nsis@ietf.org,
	Attila Bader <attila.bader@ericsson.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Lars,

Lars Westberg wrote:
> the main reason for bringing the discussion of parameters has nothing to
> with RMD. It is an arhitectural limitation. Just compared it to RSVP.
> RSVP do not have any hardcoded QoS-paraneters into the core part of the
> protocol specification, while NSIS has????

No, NSIS signaling carries the QSPEC which is opaque for the protocol.

> Given the present proposal in the QoS-spec, the architecture is much
> less open for new use than RSVP. Should NSIS make the same mistake as RSVP ?

It is not less open.

> comment about RMD:
> 
> the design target for RMD is twofold:
>     a) support for PCN-like bandwidth management.
>     b) scalable per-flow reservation.
> 
> during many years of discussion, we can conclude that additional
> functionality like RSVP-aggregation have ben design to cope with the
> state-explosion for per-flow reservation. RMD is developed to solve it
> inother approach: scalable per-flow reseravtion with the simillar
> resource guarantee as RSVP.

Actually I don't see why RMD is more scalable than RSVP aggregation
(at least if reservation-based mode of RMD is considered).
If compared to RSVP aggregation, RMD generates more signaling messages.
If I understood the RMD-draft correctly, every inter-domain/external
RESERVE message causes the QNE Ingress node to send an intra-Domain
RESERVE. Thus, interior nodes have to process per-(micro)flow signaling
messages which is avoided in the RSVP aggregation if the aggregate
reservation capacity is not adapted with every flow.

> Our target was actually to improve RSVP more improved scalabillity to
> solve a problem within the mobile access network. But, NSIS becomes the
> answer to the standardization discission so we introduced it in this
> context instead.

What is the intended status of RMD? Informational or PS?

Regards,
 Roland




_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 11:18:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg1jf-0006vI-8B; Fri, 03 Nov 2006 11:17:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg1jd-0006ud-QN
	for nsis@ietf.org; Fri, 03 Nov 2006 11:17:37 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg1ja-0001OZ-3c
	for nsis@ietf.org; Fri, 03 Nov 2006 11:17:37 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3GHGcb015418; Fri, 3 Nov 2006 17:17:27 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Roland Bless'" <bless@tm.uka.de>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and notsolving problem
Date: Fri, 3 Nov 2006 17:17:13 +0100
Message-ID: <006001c6ff63$8907c1f0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <454B6A33.8030505@tm.uka.de>
Thread-Index: Acb/Yrs/LmMCCP2kTIy79JGmmOkBSgAABo6w
X-Spam-Score: 0.12 () AWL
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 17:17:28 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 41c17b4b16d1eedaa8395c26e9a251c4
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland

> I can only agree with Dave here, that bandwidth only without 
> specifying any time period is not really usable.

Georgios: In many environments it is which will benefit
from such features.


> It is allowed if you define your own RMD-QSPEC with an 
> encapsulation method for an e2e QSPEC.

Georgios: Well, probably yes, but then all QNEs in the local domain will
have to inspect all 
QSPEC parameters including the e2e QSPEC parameters.

Best regards,
Georgios


> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: vrijdag 3 november 2006 17:12
> To: Georgios Karagiannis
> Cc: 'Ash, Gerald R (Jerry), NPE'; nsis@ietf.org
> Subject: Re: [NSIS] jerry, dave, cornelia's proposal very 
> confusing and notsolving problem
> 
> Hi Georgios,
> 
> Georgios Karagiannis schrieb:
> > Georgios: I agree, but it can not be used in all types of Diffserv 
> > based domains.
> > The proposal restricts the possibility on performing 
> admission control 
> > based on the peak rate only! That is my problem!
> > The proposal does not allow this kind of flexibility that 
> we wanted to 
> > have in NSIS.
> > I remember some severe discussions on this point some years 
> ago where 
> > we came up with the possibility on supporting both Token Bucket and 
> > Bandwidth parameters.
> 
> I can only agree with Dave here, that bandwidth only without 
> specifying any time period is not really usable.
> 
> >>> I really do not understand what kind of proposal is this??????
> >> If you care about stacking:
> >> Why not defining your own QSPEC that allows to carry your 
> >> domain-specific parameters as well as the original encapsulated 
> >> external QSPEC. This would be a form of encapsulation, but 
> it could 
> >> probably be used instead of stacking.
> > 
> > Georgios: 
> > This is what I want to do with stacking. Stack/encapsulate 
> a RMD-QSPEC 
> > above an end-to-end QSPEC. This is however not allowed in this 
> > proposal.
> 
> It is allowed if you define your own RMD-QSPEC with an 
> encapsulation method for an e2e QSPEC.
> 
> Regards,
>  Roland
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 11:28:11 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg1su-0005eB-42; Fri, 03 Nov 2006 11:27:12 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg1st-0005e4-6y
	for nsis@ietf.org; Fri, 03 Nov 2006 11:27:11 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg1sm-0002Lo-Ba
	for nsis@ietf.org; Fri, 03 Nov 2006 11:27:11 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gg1sj-0001N9-1P; Fri, 03 Nov 2006 17:27:01 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Gg1si-0007Ad-KK; Fri, 03 Nov 2006 17:27:00 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id D737DB592;
	Fri,  3 Nov 2006 17:26:59 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gg1sh-00046R-I0; Fri, 03 Nov 2006 17:26:59 +0100
Message-ID: <454B6DD2.3050704@tm.uka.de>
Date: Fri, 03 Nov 2006 17:26:58 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] jerry, dave,
	cornelia's proposal very confusing and notsolving problem
References: <006001c6ff63$8907c1f0$4c0d5982@dynamic.ewi.utwente.nl>
In-Reply-To: <006001c6ff63$8907c1f0$4c0d5982@dynamic.ewi.utwente.nl>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cf4fa59384e76e63313391b70cd0dd25
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

Georgios Karagiannis wrote:
>> It is allowed if you define your own RMD-QSPEC with an 
>> encapsulation method for an e2e QSPEC.
> 
> Georgios: Well, probably yes, but then all QNEs in the local domain will
> have to inspect all 
> QSPEC parameters including the e2e QSPEC parameters.

I don't see any problem here, you can simply skip/ignore the
encapsulated e2e QSPEC object just as you ignore the other
stacked QSPEC in the earlier proposal.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 11:30:46 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg1vk-0006NT-Vg; Fri, 03 Nov 2006 11:30:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg1vj-0006NN-Pq
	for nsis@ietf.org; Fri, 03 Nov 2006 11:30:07 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg1vg-0002dc-EQ
	for nsis@ietf.org; Fri, 03 Nov 2006 11:30:07 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3GTrsm015823; Fri, 3 Nov 2006 17:29:59 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Roland Bless'" <bless@tm.uka.de>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and notsolving problem
Date: Fri, 3 Nov 2006 17:29:51 +0100
Message-ID: <006801c6ff65$4c7602e0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <454B6DD2.3050704@tm.uka.de>
Thread-Index: Acb/ZOMvaz8uV5y5RmqPz6gLdB51lAAADtsg
X-Spam-Score: 0.119 () AWL
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 17:29:59 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 69a74e02bbee44ab4f8eafdbcedd94a1
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland

But what you propose is just stacking, but with another name, i.e., 
encapsulation.
I do not think that the new proposal allows this.

Best Regards,
Georgios
 

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: vrijdag 3 november 2006 17:27
> To: Georgios Karagiannis
> Cc: 'Ash, Gerald R (Jerry), NPE'; nsis@ietf.org
> Subject: Re: [NSIS] jerry, dave, cornelia's proposal very 
> confusing and notsolving problem
> 
> Hi Georgios,
> 
> Georgios Karagiannis wrote:
> >> It is allowed if you define your own RMD-QSPEC with an 
> encapsulation 
> >> method for an e2e QSPEC.
> > 
> > Georgios: Well, probably yes, but then all QNEs in the local domain 
> > will have to inspect all QSPEC parameters including the e2e QSPEC 
> > parameters.
> 
> I don't see any problem here, you can simply skip/ignore the 
> encapsulated e2e QSPEC object just as you ignore the other 
> stacked QSPEC in the earlier proposal.
> 
> Regards,
>  Roland
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 11:35:53 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg20i-0008Q9-4f; Fri, 03 Nov 2006 11:35:16 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg20h-0008NS-12
	for nsis@ietf.org; Fri, 03 Nov 2006 11:35:15 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg20f-00035x-Nz
	for nsis@ietf.org; Fri, 03 Nov 2006 11:35:15 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gg20d-0003FY-F3; Fri, 03 Nov 2006 17:35:12 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Gg20c-0000S1-UF; Fri, 03 Nov 2006 17:35:10 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 7B65AB592;
	Fri,  3 Nov 2006 17:35:09 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gg20b-00047T-3u; Fri, 03 Nov 2006 17:35:09 +0100
Message-ID: <454B6FBC.9070609@tm.uka.de>
Date: Fri, 03 Nov 2006 17:35:08 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] jerry, dave,
	cornelia's proposal very confusing and notsolving problem
References: <006801c6ff65$4c7602e0$4c0d5982@dynamic.ewi.utwente.nl>
In-Reply-To: <006801c6ff65$4c7602e0$4c0d5982@dynamic.ewi.utwente.nl>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 1ac7cc0a4cd376402b85bc1961a86ac2
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

Georgios Karagiannis wrote:
> But what you propose is just stacking, but with another name, i.e., 
> encapsulation.
> I do not think that the new proposal allows this.

It's independent of the new proposal. The new QSPEC proposal
is not required to support this.
You can specify this simply and independently of other QSPECs
within your own RMD-QSPEC. No need to put it in the new e2e QSPEC.

Regards
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 11:39:56 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg24a-00047v-36; Fri, 03 Nov 2006 11:39:16 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg24Z-00047q-1V
	for nsis@ietf.org; Fri, 03 Nov 2006 11:39:15 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg24X-0003MD-MC
	for nsis@ietf.org; Fri, 03 Nov 2006 11:39:15 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kA3Gd1ED016085; Fri, 3 Nov 2006 17:39:08 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Roland Bless'" <bless@tm.uka.de>
Subject: RE: [NSIS] jerry, dave,
	cornelia's proposal very confusing and notsolving problem
Date: Fri, 3 Nov 2006 17:38:59 +0100
Message-ID: <006901c6ff66$93430d70$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <454B6FBC.9070609@tm.uka.de>
Thread-Index: Acb/ZganwusFi2SeT+6fZTPMPMKViwAAEfNQ
X-Spam-Score: 0.118 () AWL
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 03 Nov 2006 17:39:08 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b30eb7682a596edff707698f4a80f7d
Cc: "'Ash, Gerald R \(Jerry\), NPE'" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland

I think it is beter to discuss this on Sonday!

What you actually say is that stacking is allowed but 
it is not specified in the QSPEC template draft.
Okay, but it then should be specified in the QoS-NSLP 
draft and in the QOSM model drafts.

Best Regards,
Georgios

 

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: vrijdag 3 november 2006 17:35
> To: Georgios Karagiannis
> Cc: 'Ash, Gerald R (Jerry), NPE'; nsis@ietf.org
> Subject: Re: [NSIS] jerry, dave, cornelia's proposal very 
> confusing and notsolving problem
> 
> Hi Georgios,
> 
> Georgios Karagiannis wrote:
> > But what you propose is just stacking, but with another name, i.e., 
> > encapsulation.
> > I do not think that the new proposal allows this.
> 
> It's independent of the new proposal. The new QSPEC proposal 
> is not required to support this.
> You can specify this simply and independently of other QSPECs 
> within your own RMD-QSPEC. No need to put it in the new e2e QSPEC.
> 
> Regards
>  Roland
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 14:16:13 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg4Vs-0005WM-UL; Fri, 03 Nov 2006 14:15:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg4Vo-0005VE-RA
	for nsis@ietf.org; Fri, 03 Nov 2006 14:15:34 -0500
Received: from sj-iport-5.cisco.com ([171.68.10.87])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg4Vl-0002wC-GN
	for nsis@ietf.org; Fri, 03 Nov 2006 14:15:32 -0500
Received: from sj-dkim-2.cisco.com ([171.71.179.186])
	by sj-iport-5.cisco.com with ESMTP; 03 Nov 2006 11:15:29 -0800
X-IronPort-AV: i="4.09,386,1157353200"; 
	d="scan'208"; a="339232337:sNHT51128752"
Received: from sj-core-5.cisco.com (sj-core-5.cisco.com [171.71.177.238])
	by sj-dkim-2.cisco.com (8.12.11.20060308/8.12.11) with ESMTP id
	kA3JFSlf016666; Fri, 3 Nov 2006 11:15:28 -0800
Received: from xbh-sjc-211.amer.cisco.com (xbh-sjc-211.cisco.com
	[171.70.151.144])
	by sj-core-5.cisco.com (8.12.10/8.12.6) with ESMTP id kA3JFSW4002997;
	Fri, 3 Nov 2006 11:15:28 -0800 (PST)
Received: from xfe-sjc-211.amer.cisco.com ([171.70.151.174]) by
	xbh-sjc-211.amer.cisco.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 11:15:28 -0800
Received: from [10.32.244.219] ([10.32.244.219]) by xfe-sjc-211.amer.cisco.com
	with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 11:15:28 -0800
In-Reply-To: <hvOvd46k.1162535221.3187200.karagian@ewi.utwente.nl>
References: <hvOvd46k.1162535221.3187200.karagian@ewi.utwente.nl>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <092E8588-6CA7-4ECA-BB3D-077A95E16F0F@cisco.com>
Content-Transfer-Encoding: 7bit
From: Fred Baker <fred@cisco.com>
Subject: Re: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 11:15:26 -0800
To: Georgios Karagiannis <karagian@cs.utwente.nl>
X-Mailer: Apple Mail (2.752.2)
X-OriginalArrivalTime: 03 Nov 2006 19:15:28.0205 (UTC)
	FILETIME=[6C254FD0:01C6FF7C]
DKIM-Signature: a=rsa-sha1; q=dns; l=1475; t=1162581328; x=1163445328;
	c=relaxed/simple; s=sjdkim2002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=fred@cisco.com; z=From:Fred=20Baker=20<fred@cisco.com>
	|Subject:Re=3A=20[NSIS]=20jerry, =20dave,
	=20cornelia's=20proposal=20very=20confusi
	ng=20and=20not=20solving=20problem;
	X=v=3Dcisco.com=3B=20h=3DsgZmBVt5zjFFlFCm8CtHxB4l9vI=3D;
	b=v7DcUUs/tbY5nz0QcBbpxD9uVXNpZswTFjTrDjUZVD3+w80Rdd0fVgjYuhPzEXyEwR94R12M
	Y7ViHCJMuxvD0opFK3wS/CDwOTgM5liSvLKgumTO+IqrozTWL0KmyqgM;
Authentication-Results: sj-dkim-2.cisco.com; header.From=fred@cisco.com;
	dkim=pass ( sig from cisco.com verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9466e0365fc95844abaf7c3f15a05c7d
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

On Nov 2, 2006, at 10:27 PM, Georgios Karagiannis wrote:
> The proposal mentions in point 2) point B):
>
> "All other QSPEC parameters are optional to include by the QNI, and  
> may
> be either mandatory or optional to interpret by downstream QNEs."

If I may point out a problem...

Generally speaking, optional capabilities create a diode. Some of us  
will recall the antics os the ISO protocol suite in the 1980's. It  
was in many ways a good piece of work, and there are some bits (IS-IS  
and ASN.1 come to mind, and the conceptual carry-over of ES-IS into  
IPv6 ND) that are still in use. The big thing that made it hard to  
use and ultimately triggered its collapse was the optional sets of  
parameters - every product built was compliant under some profile of  
selected options or another, but few were really interoperable. As a  
result, at the end of the day the solution didn't work.

With something where one has multiple products from multiple vendors  
that wind up constituting a set of needles that a transaction has to  
simultaneously thread, optional capabilities are a real liability. In  
the end, it is the mandatory-to-implement-and-use subset that will be  
interoperable end to end.

I would really suggest that we use optional capabilities as an  
upgrade path - if someone doesn't understand an option it will be  
passed along unchanged - but in the general case avoid optional  
capabilities like the plague.

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 16:48:46 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg6tU-0002gl-KJ; Fri, 03 Nov 2006 16:48:08 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg6tT-0002gW-51
	for nsis@ietf.org; Fri, 03 Nov 2006 16:48:07 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg6tS-0005Ui-0z
	for nsis@ietf.org; Fri, 03 Nov 2006 16:48:07 -0500
Received: from sj-dkim-2.cisco.com ([171.71.179.186])
	by sj-iport-6.cisco.com with ESMTP; 03 Nov 2006 13:48:05 -0800
X-IronPort-AV: i="4.09,386,1157353200"; 
	d="scan'208"; a="85242184:sNHT105970086"
Received: from sj-core-5.cisco.com (sj-core-5.cisco.com [171.71.177.238])
	by sj-dkim-2.cisco.com (8.12.11.20060308/8.12.11) with ESMTP id
	kA3Lm516007681; Fri, 3 Nov 2006 13:48:05 -0800
Received: from imail.cisco.com (sjc12-sbr-sw3-3f5.cisco.com [172.19.96.182])
	by sj-core-5.cisco.com (8.12.10/8.12.6) with ESMTP id kA3Lm5W4024100;
	Fri, 3 Nov 2006 13:48:05 -0800 (PST)
Received: from [192.168.100.106] (sjc-vpn-hwcore-391.cisco.com [10.21.153.135])
	by imail.cisco.com (8.12.11/8.12.10) with ESMTP id kA3LZ50h020711;
	Fri, 3 Nov 2006 13:35:05 -0800
In-Reply-To: <454B1268.4030405@ericsson.com>
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>
	<F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>
	<454B1268.4030405@ericsson.com>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <8203F813-69C3-4BBE-ADF2-8DC009E0F87C@cisco.com>
Content-Transfer-Encoding: 7bit
From: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 07:32:21 -0500
To: Lars Westberg <Lars.westberg@ericsson.com>
X-Mailer: Apple Mail (2.752.2)
Authentication-Results: sj-dkim-2.cisco.com; header.From=oran@cisco.com;
	dkim=pass ( sig from cisco.com verified; ); 
DKIM-Signature: a=rsa-sha1; q=dns; l=22259; t=1162590485; x=1163454485;
	c=relaxed/simple; s=sjdkim2002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=oran@cisco.com;
	z=From:David=20R=20Oran=20<oran@cisco.com>
	|Subject:Re=3A=20[NSIS]=20Proposed=20Updates=20to=20the=20QSPEC=20Draft;
	X=v=3Dcisco.com=3B=20h=3Da2zo2nRtey2Ue6oN5FICmok8I60=3D;
	b=S/05UsBiIzfPFRSbdtDFaSIklB3KE5YNA09aURfxF+SmyqkAKLRJPtAhFYjuWMR69UQZdxhx
	F613Oq4PnksBSvfhRxVR/yHoLKdMjIlED+rtiTFUyYPIM9zRYRTPZZ0T;
X-Spam-Score: 0.2 (/)
X-Scan-Signature: c07eeb7900970a16fe4056cc74ae9ce2
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


On Nov 3, 2006, at 4:56 AM, Lars Westberg wrote:

>
> Let's start with the historic aspect of QoS. Since ATM was defined  
> during beginning of 90, the discussion about traffic managment  
> (QOS) started to explode. Since then,  all standardization  
> organization has tried to defined QOS-parameters based on their own  
> scenario. The same happends in IETF with RSVP and Int.serv. We have  
> today parameters and signalling sequence diagram defined  
> differently in each standards-body especially for different access  
> networks.
>
Are you saying this is good,or bad, or what? The IETF is about  
interoperability, not defining shell standards into which the widest  
possible variety of non-interoperable things can be accommodated. In  
the world of transports, we certainly could have defined stackable  
transports, with remapping of the transport semantics at various  
domain points, and had something that worked sometimes. But, the end- 
to-end nature of the architecture would have been compromised and I  
think taking the same complex view in NSIS is well along that road.

I'm quite cognizant that some of this stuff goes counter to  
agreements that have been close to consensus in the working group for  
a long time, and had we not all come to logger-heads over the issue  
of mandatory QoS parameters this might have sailed through (at least  
until it hit the IESG). My assessment (again - let me emphasize that  
it's a personal view) was that the lack of closure over the  
parameters in fact revealed some deeper architectural difficulties,  
especially around QOS models and the role they play in the NSIS NSLP  
and QSPEC design.

So, if the WG wants to go back to the prior direction, I'll just  
crawl back in my hole and apologize for rocking the boat.

I happen to fundamentally disagree with the notion that different QoS  
models require different traffic descriptions by the traffic source.  
The traffic generation process is what it is, and defining different  
QoS models that preserve the exact source traffic generation pattern  
at the receiver to differing degrees of accuracy is perfectly  
appropriate. However, if you alter the source traffic description in  
the middle through stacking or remapping this property is lost.

I think that the the situation where "We have today parameters and  
signalling sequence diagram defined differently in each standards- 
body especially for different access networks" is a problem, not a  
good thing, and the way to deal with it is not to import every one of  
those into NSIS, but rather to map "down" to the particular domain's  
model in individual QNEs. If they in turn need to do some private  
signaling among themselves (e.g. DOCSIS, 802.11e) that's their business.

I also disagree with the notion that by choosing a 4-parameter  
traffic model for describing traffic sources somehow means that you  
are constrained to an INTSERV QOS model in terms of the QNEs and  
their RMFs that actually admit and handle the traffic. The QOSM (to  
my way of thinking) defined by what constraints and handling  
directives can be accommodated for the stated traffic description,  
not that it imposes a traffic model or description on the source.

>
> Another problem that we experiencing is that the application  
> programmer do not understand QoS-API. They actually do not know how  
> to set the parameters.
This is an independent problem but I'll observe that by having a  
large number of parameters, all optional, with no understanding of  
what they are and how they relate to how the application generates  
its traffic cannot possibly make it easier for the programmer to  
understand QoS. This argues for a single, mathematically complete and  
consistent method of describing a traffic source that all application  
scan use.

> The basic problem is that the QNI will not generate appropriate  
> parameter set given the functionality in the sub-path the  
> signalling is traversing.
I agree that the QNI will not, but further that it should not, and  
further that we should not have parameters in NSIS that   don't  
relate to providing end-to-end QoS.

> so, the reason for QoS-model stacking is to allow a network entity  
> to introduce more advanced QoS-functionality that is required for  
> the specific sub-path.
>
This is what aggregation and tunneling are for, no?

> However, the specificly added parameters may also be needed between  
> two sub-paths i.e. different access networks with a backbones  
> (Internet?) in between.
>
Can you give an example? In the multi-domain models I'm familiar  
with, the domain-specific parameters are computed on domain entry and  
are not tunneled because the domain boundaries always or nearly  
always also correspond to trust boundaries (i.e. if you have a domain  
path A-X-B with A and be using the same QoSM, B will not believe QoSM- 
specific parameters passed from A through X anyway).

> so I recommend that the stacking feature should be a part of the  
> QoS-spec template so we can allow for new usage scenario of the  
> NSIS. If it is taken away, I do not se any difference between RSVP  
> and NSIS.
>
There are a lot of differences, some in form and some in function,  
but one thing I agree with is that the general RSVP model of  
separating traffic description from QoS admission and data plane  
functions and treating the traffic description as end-to-middle-to- 
end is not broken and that there's no reason for NSIS to deviate from  
it.

> I do not understand how the process works. Should we now start a  
> blank-sheet argue functionality from beginning ?????
>
Of course not.

> the main reason for bringing the discussion of parameters has  
> nothing to with RMD. It is an arhitectural limitation. Just  
> compared it to RSVP. RSVP do not have any hardcoded QoS-paraneters  
> into the core part of the protocol specification, while NSIS has????
Maybe to do a better job?

> Given the present proposal in the QoS-spec, the architecture is  
> much less open for new use than RSVP. Should NSIS make the same  
> mistake as RSVP ?
>
RSVP's problems had little to do with its QoS description  
capabilities, but arguing the fate of RSVP is unlikely to help us  
resolve the general architectural questions that have come up as a  
result of not being able to reach closure on the QSPEC draft.

> comment about RMD:
>
> the design target for RMD is twofold:
>     a) support for PCN-like bandwidth management.
>     b) scalable per-flow reservation.
>
> during many years of discussion, we can conclude that additional  
> functionality like RSVP-aggregation have ben design to cope with  
> the state-explosion for per-flow reservation. RMD is developed to  
> solve it inother approach: scalable per-flow reseravtion with the  
> simillar resource guarantee as RSVP.
>
Understood. However, what I read into this is that RMD is a QoS  
signaling architecture of its own and the desire is to co-mingle it  
with NSIS. That runs counter to the goal of universal  
interoperability, IMO.

> Our target was actually to improve RSVP more improved scalabillity  
> to solve a problem within the mobile access network. But, NSIS  
> becomes the answer to the standardization discission so we  
> introduced it in this context instead.
>
Also understood, and other things being equal, a worthy endeavor.

Cheers, Dave.

> regards Lasse
>
> David R Oran wrote:
>> I think some detailed discussion of this is going to be needed,
>> because after considerable pondering the three of us could not at all
>> figure out what either remapping or QSPEC stacking was good for. It
>> certainly introduces a vast amount of complexity and confusion as
>> currently formulated.
>>
>> It appears at least to me that one of three situation obtains:
>> (a) we (well, at least I) don't understand RMD
>> (b) there is some general QoS concept that we are missing - we went
>> through some complex scenarios like reservation aggregation any
>> multiple QOSMs chained together and still could not figure out how
>> one needed either stacking or remapping of QSPEC parameters
>> (c) RMD is really a completely different QoS signaling architecture
>> from the rest of NSIS and is being shoehorned in by complicating
>> QNSLP and QSPEC to allow multiple QoS signaling schemes to masquerade
>> as a single architecture.
>>
>> I really hope it's (a) or (b) since then we can probably resolve this
>> by either putting back the missing functionality or figuring out
>> another simple way to accommodate RMD. What I'm most afraid of is
>> that this is really (c).
>>
>>  From the little I understand RMD, it's really like mapping onto a
>> custom L2 QoS scheme like ATM or DOCSIS, and that rather than doing
>> the mapping the way any conventional L3 RMF would do it, you co-
>> mingle RMD's model with NSIS's model.
>>
>> But, I could of course be wildly wrong here.
>>
>> Dave.
>>
>>
>> On Nov 2, 2006, at 6:00 AM, Georgios Karagiannis wrote:
>>
>> > Hi John
>> >
>> > I have just come back from a trip and I have just look into  
>> point 4):
>> >>> 4. Since remapping is eliminated, the concept of stacking  
>> QSPECs is
>> >>> also eliminated.  Discussion of tunneling of QNSLP messages is
>> >>> eliminated since there are no impacts on tunneling required by
>> >>> QSPEC.
>> >
>> > QSPEC stacking is one of the significant concepts in the QoS based
>> > NSIS
>> > solution.
>> > This concept provides the additional flexibility that is missing
>> > from RSVP.
>> > Furthermore, all the work that we have done until now is based on
>> > the concept of QSPEC stacking. Thus what you actually say is to  
>> just
>> > forget the work that we have done
>> > until now, in e.g.the RMD-QOSM draft, and start from the beginning.
>> >
>> >
>> > The main problem that we have had until now was not the QSPEC  
>> stacking
>> > issue,
>> > but it was the huge number of mandatory paremeters. Actualy I
>> > consider the
>> > issue of removing the
>> > QSPEC stacking concept a much more severe problem on achieving the
>> > flexibility
>> > that we wanted to achieve with NSIS than the huge number of  
>> mandatory
>> > parameters.
>> >
>> > Therefore, I am severely against removing the QSPEC stacking  
>> concept!
>> >
>> > Best Regards,
>> > Georgios
>> >
>> >
>> >
>> >
>> >
>> >> -----Original Message-----
>> >> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
>> >> Sent: donderdag 2 november 2006 8:35
>> >> To: gash@att.com; nsis@ietf.org
>> >> Cc: cornelia.kappler@siemens.com; oran@cisco.com;
>> >> cornelia.kappler@gmx.net
>> >> Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
>> >>
>> >> Jerry, Dave and Cornelia;
>> >>
>> >> This looks like a very good simplification of the QSPEC, and
>> >> I think that is a good idea.  I think that there will be some
>> >> implications to the QoSM drafts - I think QoS Model drafts
>> >> will need to contain some more applicability statement type
>> >> of text, along with deployment considerations.
>> >>
>> >> A plus side of this is that I think that the QoS template
>> >> will become more broadly applicable to other uses than just
>> >> NSIS QoS Signaling, so I think that is a gain as well.
>> >>
>> >> I'd really like people to review this and send comments on it
>> >> so that we could really focus the discussion next week and
>> >> get this work finally resolved.
>> >>
>> >> Thanks a lot for working through this, I think this is a good
>> >> basis for a proposal.
>> >>
>> >> thanks,
>> >> John
>> >>
>> >>> -----Original Message-----
>> >>> From: ext Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
>> >>> Sent: 02 November, 2006 00:13
>> >>> To: nsis@ietf.org
>> >>> Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran;
>> >>> Kappler,Cornelia
>> >>> Subject: [NSIS] Proposed Updates to the QSPEC Draft
>> >>>
>> >>> Hi All,
>> >>>
>> >>> Per John's suggestion below we've discussed Dave Oran's
>> >> suggestions at
>> >>> length and have the following proposal.
>> >>>
>> >>> We think the proposed changes address many concerns raised
>> >> on the list
>> >>> over the last few months and will require some rather basic
>> >> changes in
>> >>> the QSPEC document
>> >>> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
>> >>>
>> >>> We appreciate Dave's suggestions and careful review of the  
>> document.
>> >>>
>> >>> Here is a summary of the proposed changes:
>> >>>
>> >>> 1. The notion of QOSMs will be played down in the updated draft.
>> >>> Language to the effect that 'QNSLP/QSPEC can signal for
>> >> different QOSMs
>> >>> across multiple domains' will be replaced by the notion that
>> >>> QNSLP/QSPEC allows different elements on the path to implement
>> >>> different data plane QoS mechanisms that meets the
>> >> constraints included
>> >>> in the QSPEC.  A QOSM describes common capabilities among
>> >> QNEs to act
>> >>> consistently when requested to admit traffic and in treating
>> >> admitted
>> >>> traffic.
>> >>> A QOSM ID need not be defined or signaled and a QNE need not  
>> support
>> >>> any particular QOSM.  A 'QOSM specification' would become a
>> >>> 'QNSLP/QSPEC applicability statement', which documents how a  
>> QNE/RMF
>> >>> interprets and treats the various elements in a QSPEC.
>> >>>
>> >>> An updated QOSM definition is as follows:
>> >>>
>> >>> QoS Model (QOSM): A method to achieve QoS for a traffic flow,  
>> e.g.,
>> >>> IntServ Controlled Load.  A QOSM specifies what sub-set of  
>> QSPEC QoS
>> >>> constraints and traffic handling directives a QNE implementing  
>> that
>> >>> QOSM is capable of supporting and how resources will be
>> >> managed by the
>> >>> RMF in such a QNE.
>> >>>
>> >>> 2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --
>> >>> > d)
>> >>> semantics:
>> >>> A) The source traffic description is mandatory to include by  
>> the QNI
>> >>> and mandatory to interpret by downstream QNEs.  The traffic
>> >> description
>> >>> is specified by a traffic model (TMOD) parameter consisting of 4
>> >>> sub-parameters:
>> >>> - rate (r)
>> >>> - bucket size (b)
>> >>> - peak rate (p)
>> >>> - minimum policed unit (m)
>> >>> The TMOD parameter is a mathematically complete way to describe a
>> >>> traffic source.  Examples will be given for setting TMOD to
>> >> correspond
>> >>> to specific cases, e.g. where only bandwidth is available
>> >> (r=p; b and m
>> >>> set to default values) and other cases (e.g., TCP).  As such, the
>> >>> bandwidth parameter is eliminated.
>> >>> B) All other QSPEC parameters are optional to include by the
>> >> QNI, and
>> >>> may be either mandatory or optional to interpret by downstream  
>> QNEs.
>> >>> That is, the QNI can explicitly specify for each constraint
>> >> parameter
>> >>> whether it wishes admission control to succeed or fail if that
>> >>> constraint cannot be met.  A downstream QNE distinguishes
>> >> whether it is
>> >>> failure to understand the parameter or failure to meet the
>> >> constraint
>> >>> that causes the admission control failure (more detail on
>> >> this below).
>> >>> C) Additional QSPEC parameters can be defined in updates to
>> >> the QSPEC
>> >>> document (or alternatively specified in QOSM applicability  
>> statement
>> >>> documents).  There is already an IANA registry established for
>> >>> additional QSPEC parameters.
>> >>>
>> >>> 3. The concept of remapping QSPEC parameters is eliminated.
>> >>> This concept is ill defined and has caused much confusion  
>> (there are
>> >>> few if any sensible remappings).  Rather we define
>> >> 'interpret' a QSPEC
>> >>> parameter to only mean what we now call 'strictly  
>> interpret' (i.e.,
>> >>> must conform to the RFCs defining the parameter and  
>> procedures) and
>> >>> eliminate the remap option part of 'interpret'.
>> >>>
>> >>> 4. Since remapping is eliminated, the concept of stacking  
>> QSPECs is
>> >>> also eliminated.  Discussion of tunneling of QNSLP messages is
>> >>> eliminated since there are no impacts on tunneling required by
>> >>> QSPEC.
>> >>>
>> >>> 5. Current QSPEC flags are modified as follows:
>> >>> A) The QNI sets the M flag for each QSPEC parameter it
>> >> populates that
>> >>> it intends must be interpreted or the reservation fails.
>> >> Currently the
>> >>> M flag is statically set for every QSPEC parameter.
>> >>> B) If the M flag is set, a downstream QNE must interpret the
>> >> parameter
>> >>> or the reservation fails.  If a QNE does not support the
>> >> parameter with
>> >>> the M flag set, it sets the N flag and rejects the
>> >> reservation.  If the
>> >>> QNE supports the parameter but cannot meet the parameter, it
>> >> sets the E
>> >>> flag and rejects the reservation.
>> >>> C) If the M flag is not set, a downstream QNE should interpret  
>> the
>> >>> parameter.  If a QNE does not support the parameter, it sets
>> >> the N flag
>> >>> and can optionally accept or reject the reservation.  If a
>> >> QNE supports
>> >>> the parameter but cannot meet the parameter, it sets the E  
>> flag and
>> >>> optional accepts or rejects the reservation.
>> >>> D) The R (remapped parameter) flag and Q (non QOSM) flag are
>> >>> eliminated.
>> >>>
>> >>> 6. Summary of proposed QSPEC parameters:
>> >>>
>> >>> Source traffic Description:
>> >>> TMOD-1 (mandatory to include)
>> >>> TMOD-2 (optional to include)
>> >>>
>> >>> Constraints (optional to include):
>> >>> Path Latency
>> >>> Path Jitter
>> >>> Path PLR
>> >>> Path PER
>> >>> Y.1541 class (shorthand notation for above 4 parameters)
>> >> Slack term
>> >>> Priority (Preemption, Defending, Admission, RPH
>> >>> Priority)  DSTE class
>> >>>
>> >>> Handling directives (optional to include):
>> >>> excess treatment
>> >>>
>> >>> Traffic Classifiers (optional to include):
>> >>> PHB Class (note that PHB class is set by downstream QNE to  
>> tell QNI
>> >>> how to mark traffic to ensure it will be treated as committed by
>> >>> admission
>> >>> control)
>> >>>
>> >>> Eliminated:
>> >>> bandwidth
>> >>> Ctot, Dtot, Csum, Dsum
>> >>>
>> >>> 7. Editorial changes:
>> >>> A) rewrite description of read-only and read-write
>> >> parameters to make
>> >>> the rules more clear:
>> >>> - which parameters can be ro/rw (function of object in which
>> >> parameter
>> >>> included)
>> >>> - can only increase rw values and not delete, etc.
>> >>> B) eliminate the distinction of 'control' and 'qos description'
>> >>> parameters and replace with suggested taxonomy, as described  
>> above
>> >>> C) clarify discussion of packet outcomes from Y.1540 and place in
>> >>> appropriate sections
>> >>> D) clarify which parameters can be included in <QoS Desired> <QoS
>> >>> Available>, <QoS Reserved>, or <QoS Minimum>
>> >>>
>> >>> Please review and comment on the proposed changes.
>> >>>
>> >>> Thanks,
>> >>> Jerry, Dave, Cornelia
>> >>>
>> >>>
>> >>> -----Original Message-----
>> >>> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
>> >>> Sent: Monday, October 23, 2006 2:59 AM
>> >>> To: nsis@ietf.org
>> >>> Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec  
>> draft
>> >>>
>> >>> Hi all,
>> >>>
>> >>> We have had a very large thread on the current
>> >>> draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear
>> >> consensus,
>> >>> so I'd like to try to summarize the discussions.
>> >>>
>> >>> Several people feel that the current draft contains too many
>> >> mandatory
>> >>> parameters and mapping some QoS Models to others is mixing
>> >>> apples-to-oranges.  The feeling is that interoperability is  
>> hard to
>> >>> achieve and therefore we should not try to gain interoperability.
>> >>> Jukka Manner's message on 17th of October proposed a way forward.
>> >>> http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html
>> >>>
>> >>> However, several (other) people feel that interoperability
>> >> is important
>> >>> and is needed. Giving up on interoperability would not be a
>> >> good thing,
>> >>> and that we are very close to a solution.
>> >>> Dave Oran proposed a mechanism to reach a solution on the 19th
>> >>> October:
>> >>> http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html
>> >>>
>> >>> I have asked Jerry and Dave to make a proposal along the lines of
>> >>> Dave's message above, and the working group can review this,
>> >> to see if
>> >>> we can gain consensus on this.  I think it would be
>> >> sufficient to send
>> >>> some rough text to the mailing list, and if this is OK, then  
>> we can
>> >>> revise the draft.
>> >>>
>> >>> One open question to Dave/Jerry about the proposal is, in the  
>> event
>> >>> that the source cannot be described a mathematically
>> >> complete way, is
>> >>> there any value (or does it make any sense at
>> >>> all) in sending an incomplete qspec, or potentially "wildcarding"
>> >>> sending some of the mandatory parameters.  I'm wondering if some
>> >>> sources / deployments may have certain parameters that are
>> >> relatively
>> >>> well defined, but not all are so well defined.
>> >>>
>> >>> thanks,
>> >>> John
>> >>>
>> >>> _______________________________________________
>> >>> nsis mailing list
>> >>> nsis@ietf.org
>> >>> https://www1.ietf.org/mailman/listinfo/nsis
>> >>>
>> >>> _______________________________________________
>> >>> nsis mailing list
>> >>> nsis@ietf.org
>> >>> https://www1.ietf.org/mailman/listinfo/nsis
>> >>>
>> >>
>> >> _______________________________________________
>> >> nsis mailing list
>> >> nsis@ietf.org
>> >> https://www1.ietf.org/mailman/listinfo/nsis
>> >>
>> >
>> >
>> >
>> > _______________________________________________
>> > nsis mailing list
>> > nsis@ietf.org
>> > https://www1.ietf.org/mailman/listinfo/nsis
>>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 16:52:53 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg6xQ-00073u-If; Fri, 03 Nov 2006 16:52:12 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg6xP-0006zG-5Q
	for nsis@ietf.org; Fri, 03 Nov 2006 16:52:11 -0500
Received: from sj-iport-5.cisco.com ([171.68.10.87])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg6xO-0007vI-9D
	for nsis@ietf.org; Fri, 03 Nov 2006 16:52:11 -0500
Received: from sj-dkim-6.cisco.com ([171.68.10.81])
	by sj-iport-5.cisco.com with ESMTP; 03 Nov 2006 13:52:10 -0800
X-IronPort-AV: i="4.09,386,1157353200"; 
	d="scan'208"; a="339285168:sNHT84890456"
Received: from sj-core-3.cisco.com (sj-core-3.cisco.com [171.68.223.137])
	by sj-dkim-6.cisco.com (8.12.11.20060308/8.12.11) with ESMTP id
	kA3Lq96b006468; Fri, 3 Nov 2006 13:52:09 -0800
Received: from imail.cisco.com (sjc12-sbr-sw3-3f5.cisco.com [172.19.96.182])
	by sj-core-3.cisco.com (8.12.10/8.12.6) with ESMTP id kA3Lq9bF022958;
	Fri, 3 Nov 2006 13:52:09 -0800 (PST)
Received: from [192.168.100.106] (sjc-vpn-hwcore-391.cisco.com [10.21.153.135])
	by imail.cisco.com (8.12.11/8.12.10) with ESMTP id kA3Ld80t020799;
	Fri, 3 Nov 2006 13:39:08 -0800
In-Reply-To: <004601c6fe91$27c60640$4c0d5982@dynamic.ewi.utwente.nl>
References: <004601c6fe91$27c60640$4c0d5982@dynamic.ewi.utwente.nl>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=ISO-8859-1; delsp=yes; format=flowed
Message-Id: <3A01242E-A7F7-4BCB-8C49-A452382F36D9@cisco.com>
Content-Transfer-Encoding: quoted-printable
From: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 07:36:34 -0500
To: Georgios Karagiannis <karagian@cs.utwente.nl>
X-Mailer: Apple Mail (2.752.2)
Authentication-Results: sj-dkim-6.cisco.com; header.From=oran@cisco.com;
	dkim=pass ( sig from cisco.com verified; ); 
DKIM-Signature: a=rsa-sha1; q=dns; l=15844; t=1162590729; x=1163454729;
	c=relaxed/simple; s=sjdkim6002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=oran@cisco.com;
	z=From:David=20R=20Oran=20<oran@cisco.com>
	|Subject:Re=3A=20[NSIS]=20Proposed=20Updates=20to=20the=20QSPEC=20Draft;
	X=v=3Dcisco.com=3B=20h=3D+fbDsKMN+x04uJSpnC5JZWLxxUg=3D;
	b=uNnEOE901SAgZHIPlPrMAS457W9uYXGiMZrGvGpf+Zw7f0HHa2v8f1FfMaKHHe7DMeewthtl
	hyFy+B8t319bL3In1S1v+AfepxixOuCQ754lZgQVd6Xmsq3uLG0zF4eW;
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 958aa603499a3de6b2b87d68741ed60e
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


On Nov 2, 2006, at 10:11 AM, Georgios Karagiannis wrote:

> Hi Dave
>
> You mention that remapping of the Token bucket to bandwidth is not =20
> allowed
> but it is allowed
> to interpret the peak bandwidth parameter value.
>
Correct.

> Do you mean that a QNE can use the peak bandwidth parameter and ignore
> all the other Token Bucket parameters?
>
If the QNE deterministically has enough buffering to absorb any =20
burst, and its scheduling accuracy is sufficient to not worry about =20
the distortions of having a non-zero minimum-policed unit, and that =20
it prefers to be pessimistic in admission control by considering only =20=

peak and not average rate, then yes.

> Regarding the signifficance of the QSPEC stacking please refer to =20
> Section
> 3.2.8 in:
> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qos-nslp-12.txt
>
I'll go back and read that again. Thanks.

> Best Regards,
> Georgios
>
>
>> -----Original Message-----
>> From: David R Oran [mailto:oran@cisco.com]
>> Sent: donderdag 2 november 2006 15:21
>> To: Georgios Karagiannis
>> Cc: 'Ash, Gerald R (Jerry), ALABS'; nsis@ietf.org; 'Cornelia
>> Kappler'; 'Kappler,Cornelia'; 'Attila B=E1der (IJ/ETH)' ; 'Lars
>> Westberg (KI/EAB)'; 'Magnus Westerlund'; 'Lars Eggert';
>> john.loughney@nokia.com; stiemerling@netlab.nec.de
>> Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
>>
>>
>> On Nov 2, 2006, at 8:38 AM, Georgios Karagiannis wrote:
>>
>>> Hi Jerry, Dave and Cornelia
>>>
>>> Here are some more detail comments on your proposal!
>>>
>>> This proposal actually changes completely the QOS based concept of
>>> NSIS and not only the QSPEC draft. Therefore, I do not
>> agree with the
>>> proposal.
>>>
>>> The problems that we have had until now, in the nsis mailing
>>> discussions, were related to the huge number of mandatory
>> parameters
>>> that all nodes in the universe had to support.
>>> This proposal proposes only one parameter Token Bucket,
>> which is good,
>>> but it removes the possibility from NSIS to use QoS signaling for
>>> Diffserv domains, which is against the NSIS charter specification.
>>>
>> How do you reach this conclusion? Could you give an example
>> of what doesn't work for diffserv?
>>
>>> Moreover, the remapping possibility from the peak bandwidth (Token
>>> Bucket)
>>> value to
>>> the Bandwidth value that can be used in RMD-QOSM aware like
>> nodes is
>>> eliminated.
>> Yes, this was intentional. There is no need to remap in the
>> NSIS signaling. A given QOSM's RMF can simply interpret the
>> peak bandwidth parameter as their
>> admitted/shaped/policed/queued bandwidth. This behavior would
>> be documented in the QOSM application document for RMD.
>>
>>> I do not understand why, since we have had a consensus on
>> this point
>>> in the nsis mailing list.
>> What seems to have happened (at least from my vantage point
>> as a long- time lurker and occasional contributor) is that as
>> we got down to reach closure on issues around the mandatory
>> nature of QSPEC parameters and why certain QOSMs wanted
>> certain parameters for certain purposes, we got to a
>> deadlock. Removing the deadlock simply by blowing away any
>> common interoperable behavior for NSIS (by declaring
>> everything optional) would likely have run straight into a
>> brick wall in the IESG. This prompted my, Jerry and Cornelia
>> to take a step back, put our heads together, and try to
>> figure out what the underlying problems were. Our assessment,
>> captured in the proposal we just made, was that in order to
>> maintain flexibility but still have global interoperability
>> we needed to step back and rethink some of the more complex
>> and confusing parts of the QSPEC design. We did that and you
>> see the result. It's a matter of point of view whether one
>> considers this "big change" or "valuable simplification".
>>
>>> This proposal also removes the QSPEC stacking concept,
>> which is one of
>>> the most significant concepts in the QoS based NSIS solution.
>> So, we clearly missed the significance, so perhaps you need
>> to go back and re-educate us. No other QOSM I know of other
>> than RMD claims to need this functionality.
>>
>>> This concept provides the additional flexibility that is
>> missing from
>>> RSVP.
>>> Furthermore, all the work that we have done until now is
>> based on the
>>> concept of QSPEC stacking. Thus what you actually say is to just
>>> forget the work that we have done until now, in e.g.the RMD-QOSM
>>> draft, and start from the beginning.
>>>
>> Not necessarily. What we may need to do is rethink a bit how
>> RMD integrates with NSIS as a QOSM. the other possibility is
>> for the WG to reject our approach and stay with what's in the
>> QPSEC draft today.
>>
>>> Furthermore, I wonder why Attila Bader has been left out of the
>>> discussions that you have had to come up to this proposal.
>>> =46rom what I know Jerry, Cornelia and Attila were the editors of =
the
>>> QSPEC template draft.
>>>
>>> More comments in line!
>>>
>>>
>>>> -----Original Message-----
>>>> From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]
>>>> Sent: woensdag 1 november 2006 23:13
>>>> To: nsis@ietf.org
>>>> Cc: Ash, Gerald R (Jerry), ALABS; Cornelia Kappler; David R Oran;
>>>> Kappler,Cornelia
>>>> Subject: [NSIS] Proposed Updates to the QSPEC Draft
>>>>
>>>> Hi All,
>>>>
>>>> Per John's suggestion below we've discussed Dave Oran's
>> suggestions
>>>> at length and have the following proposal.
>>>>
>>>> We think the proposed changes address many concerns raised on the
>>>> list over the last few months and will require some rather basic
>>>> changes in the QSPEC document
>>>> http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
>>>>
>>>> We appreciate Dave's suggestions and careful review of the
>> document.
>>>>
>>>> Here is a summary of the proposed changes:
>>>>
>>>> 1. The notion of QOSMs will be played down in the updated draft.
>>>> Language to the effect that 'QNSLP/QSPEC can signal for different
>>>> QOSMs across multiple domains' will be replaced by the notion that
>>>> QNSLP/QSPEC allows different elements on the path to implement
>>>> different data plane QoS mechanisms that meets the constraints
>>>> included in the QSPEC.  A QOSM describes common capabilities among
>>>> QNEs to act consistently when requested to admit traffic and in
>>>> treating admitted traffic.  A QOSM ID need not be defined
>> or signaled
>>>> and a QNE need not support any particular QOSM.  A 'QOSM
>>>> specification'
>>>> would become a 'QNSLP/QSPEC applicability statement',
>> which documents
>>>> how a QNE/RMF interprets and treats the various elements
>> in a QSPEC.
>>>
>>>
>>> Georgios: This is a severe change in the NSIS QoS based
>> concept, which
>>> we have agreed
>>> 4/5 years ago to use in our work.
>>> The current agreed concept compared to RSVP provides a flexibility
>>> advantage that is missing from RSVP.
>>>
>> It is a change/simplification. What I don't get (I won't
>> speak for Jerry or Cornelia, but neither of them could come
>> up with a succinct explanation when we discussed this) is
>> exactly what this flexibility is and why it matters.
>>
>>>
>>>
>>>> An updated QOSM definition is as follows:
>>>>
>>>> QoS Model (QOSM): A method to achieve QoS for a traffic
>> flow, e.g.,
>>>> IntServ Controlled Load.  A QOSM specifies what sub-set of
>> QSPEC QoS
>>>> constraints and traffic handling directives a QNE
>> implementing that
>>>> QOSM is capable of supporting and how resources will be managed by
>>>> the RMF in such a QNE.
>>>>
>>>> 2. The QSPEC1/QSPEC2 semantics will be replaced with
>> Dave's a) --> d)
>>>> semantics:
>>>> A) The source traffic description is mandatory to include
>> by the QNI
>>>> and mandatory to interpret by downstream QNEs.  The traffic
>>>> description is specified by a traffic model (TMOD) parameter
>>>> consisting of 4
>>>> sub-parameters:
>>>> - rate (r)
>>>> - bucket size (b)
>>>> - peak rate (p)
>>>> - minimum policed unit (m)
>>>> The TMOD parameter is a mathematically complete way to describe a
>>>> traffic source.  Examples will be given for setting TMOD to
>>>> correspond to specific cases, e.g. where only bandwidth is
>> available
>>>> (r=3Dp; b and m set to default values) and other cases
>> (e.g., TCP).  As
>>>> such, the bandwidth parameter is eliminated.
>>>
>>> Georgios: What you actually now say is that a Diffserv
>> domain cannot
>>> use NSIS for QoS signaling. As you probably know the NSIS charter
>>> specifies that NSIS should provide QoS signaling for
>> Diffserv domains.
>>> This is actually against the NSIS charter specification.
>>>
>> I'm lost. what leads you to this conclusion?
>>
>>>
>>>> B) All other QSPEC parameters are optional to include by
>> the QNI, and
>>>> may be either mandatory or optional to interpret by
>> downstream QNEs.
>>>> That is, the QNI can explicitly specify for each
>> constraint parameter
>>>> whether it wishes admission control to succeed or fail if that
>>>> constraint cannot be met.  A downstream QNE distinguishes
>> whether it
>>>> is failure to understand the parameter or failure to meet the
>>>> constraint that causes the admission control failure (more
>> detail on
>>>> this below).
>>>
>>> Georgios: In other words, RMD-QOSM cannot any more use NSIS for QoS
>>> signaling ???
>>>
>>>
>> Possibly, since I'm not sure the goal of NSIS was ever to be
>> just a schematic shell for inserting individual QOSM-specific
>> signaling protocols. If it was, then clearly my understanding
>> of the NSIS architecture and what I read in the NSIS charter
>> does not correspond to the WG consensus our changes to the
>> QSPEC draft would be ill-advised.
>>
>>>
>>>> C) Additional QSPEC parameters can be defined in updates
>> to the QSPEC
>>>> document (or alternatively specified in QOSM applicability
>> statement
>>>> documents).  There is already an IANA registry established for
>>>> additional QSPEC parameters.
>>>>
>>>> 3. The concept of remapping QSPEC parameters is eliminated.
>>>> This concept is ill defined and has caused much confusion
>> (there are
>>>> few if any sensible remappings).  Rather we define 'interpret' a
>>>> QSPEC parameter to only mean what we now call 'strictly interpret'
>>>> (i.e., must conform to the RFCs defining the parameter and
>>>> procedures) and eliminate the remap option part of 'interpret'.
>>>
>>>
>>> Georgios: Again the peak bandwidth in the Token Bucket
>> parameter can
>>> be remapped to Bandwidth. Many people already agreed with this.
>> If you mean "interpreted by the RMD QOSM as bandwidth" then
>> we are in agreement. If you mean "modified in the NSLP
>> signaling and QSPEC by replacing the <peak bandwidth
>> parameter> with a <bandwidth> parameter, possibly stacking
>> the new QPEC with <bandwidth> instead of <peak bandwidth>"
>> then that's where we disagreed with the current approach in
>> fact departed from the current spec.
>>
>> The problem with both remapping and stacking is that they do
>> not constrain the semantics of the end-to-end signaling, and
>> introduce considerable complexity into the protocol.
>>
>>>  Thus I do not see why you do not want to support this feature.
>>> Probably, because in this way you will actually exclude the use and
>>> the application of the RMD-QOSM work ???
>>>
>>>
>>>
>>>> 4. Since remapping is eliminated, the concept of stacking
>> QSPECs is
>>>> also eliminated.  Discussion of tunneling of QNSLP messages is
>>>> eliminated since there are no impacts on tunneling
>> required by QSPEC.
>>>
>>> Georgios:
>>>
>>> QSPEC stacking is one of the significant concepts in the QoS based
>>> NSIS solution.
>>> This concept provides the additional flexibility that is
>> missing from
>>> RSVP.
>>> Furthermore, all the work that we have done until now is
>> based on the
>>> concept of QSPEC stacking. Thus what you actually say is to just
>>> forget the work that we have done until now, in e.g.the RMD-QOSM
>>> draft, and start from the beginning.
>>> The main problem that we have had until now was not the
>> QSPEC stacking
>>> issue, but it was the huge number of mandatory paremeters.
>> Actualy I
>>> consider the issue of removing the QSPEC stacking concept a
>> much more
>>> severe problem on achieving the flexibility that we wanted
>> to achieve
>>> with NSIS than the huge number of mandatory parameters.
>>>>
>>>> 5. Current QSPEC flags are modified as follows:
>>>> A) The QNI sets the M flag for each QSPEC parameter it
>> populates that
>>>> it intends must be interpreted or the reservation fails.
>> Currently
>>>> the M flag is statically set for every QSPEC parameter.
>>>> B) If the M flag is set, a downstream QNE must interpret the
>>>> parameter or the reservation fails.  If a QNE does not support the
>>>> parameter with the M flag set, it sets the N flag and rejects the
>>>> reservation.  If the QNE supports the parameter but cannot
>> meet the
>>>> parameter, it sets the E flag and rejects the reservation.
>>>> C) If the M flag is not set, a downstream QNE should interpret the
>>>> parameter.  If a QNE does not support the parameter, it sets the N
>>>> flag and can optionally accept or reject the reservation.
>> If a QNE
>>>> supports the parameter but cannot meet the parameter, it
>> sets the E
>>>> flag and optional accepts or rejects the reservation.
>>>> D) The R (remapped parameter) flag and Q (non QOSM) flag are
>>>> eliminated.
>>>
>>> Georgios: I do not agree to completely remove the
>> remapping, see the
>>> reason given on the previous points  and I also disagree to
>> remove the
>>> R and Q flags listed above!
>>>
>>>
>>>
>>>> 6. Summary of proposed QSPEC parameters:
>>>>
>>>> Source traffic Description:
>>>>  TMOD-1 (mandatory to include)
>>>>  TMOD-2 (optional to include)
>>>>
>>>> Constraints (optional to include):
>>>>  Path Latency
>>>>  Path Jitter
>>>>  Path PLR
>>>>  Path PER
>>>>  Y.1541 class (shorthand notation for above 4 parameters)
>>>> Slack term  Priority (Preemption, Defending, Admission, RPH
>>>> Priority)  DSTE class
>>>>
>>>> Handling directives (optional to include):
>>>>  excess treatment
>>>>
>>>> Traffic Classifiers (optional to include):
>>>>  PHB Class (note that PHB class is set by downstream QNE to
>>>> tell QNI how to mark traffic to ensure it will be treated as
>>>> committed by admission
>>>> control)
>>>>
>>>> Eliminated:
>>>>  bandwidth
>>>>  Ctot, Dtot, Csum, Dsum
>>>
>>> Georgios: I do not agree, See above!
>>>
>>>> 7. Editorial changes:
>>>> A) rewrite description of read-only and read-write parameters
>>>> to make the rules more clear:
>>>> - which parameters can be ro/rw (function of object in which
>>>> parameter
>>>> included)
>>>> - can only increase rw values and not delete, etc.
>>>> B) eliminate the distinction of 'control' and 'qos description'
>>>> parameters and replace with suggested taxonomy, as described above
>>>> C) clarify discussion of packet outcomes from Y.1540 and
>>>> place in appropriate sections
>>>> D) clarify which parameters can be included in <QoS Desired> <QoS
>>>> Available>, <QoS Reserved>, or <QoS Minimum>
>>>
>>> Georgios: I do not agree, See above!
>>>
>>>
>>>> Please review and comment on the proposed changes.
>>>> Thanks,
>>>> Jerry, Dave, Cornelia
>>>
>>>
>>> Best Regards,
>>> Georgios
>>>
>>
>
>
>
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 16:56:50 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gg71J-0005Rd-L5; Fri, 03 Nov 2006 16:56:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gg71I-0005RY-Au
	for nsis@ietf.org; Fri, 03 Nov 2006 16:56:12 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gg71F-0002Qr-TT
	for nsis@ietf.org; Fri, 03 Nov 2006 16:56:12 -0500
Received: from sj-dkim-2.cisco.com ([171.71.179.186])
	by sj-iport-6.cisco.com with ESMTP; 03 Nov 2006 13:56:09 -0800
X-IronPort-AV: i="4.09,386,1157353200"; 
	d="scan'208"; a="85246677:sNHT57816297"
Received: from sj-core-5.cisco.com (sj-core-5.cisco.com [171.71.177.238])
	by sj-dkim-2.cisco.com (8.12.11.20060308/8.12.11) with ESMTP id
	kA3Lu9Ru018227; Fri, 3 Nov 2006 13:56:09 -0800
Received: from imail.cisco.com (sjc12-sbr-sw3-3f5.cisco.com [172.19.96.182])
	by sj-core-5.cisco.com (8.12.10/8.12.6) with ESMTP id kA3Lu8W4001140;
	Fri, 3 Nov 2006 13:56:08 -0800 (PST)
Received: from [192.168.100.106] (sjc-vpn-hwcore-391.cisco.com [10.21.153.135])
	by imail.cisco.com (8.12.11/8.12.10) with ESMTP id kA3Lh8Jj020871;
	Fri, 3 Nov 2006 13:43:08 -0800
In-Reply-To: <Pine.LNX.4.64.0611031140410.7180@sbz-31.cs.Helsinki.FI>
References: <454A0764.3070202@gmx.net> <454B0BFE.3000708@gmx.net>
	<Pine.LNX.4.64.0611031140410.7180@sbz-31.cs.Helsinki.FI>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <8277BEB3-F2C2-43FD-B931-A9ABF7B18866@cisco.com>
Content-Transfer-Encoding: 7bit
From: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] QSPEC Face-to-Face Chat: Another Try
Date: Fri, 3 Nov 2006 07:37:20 -0500
To: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
X-Mailer: Apple Mail (2.752.2)
Authentication-Results: sj-dkim-2.cisco.com; header.From=oran@cisco.com;
	dkim=pass ( sig from cisco.com verified; ); 
DKIM-Signature: a=rsa-sha1; q=dns; l=1283; t=1162590969; x=1163454969;
	c=relaxed/simple; s=sjdkim2002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=oran@cisco.com;
	z=From:David=20R=20Oran=20<oran@cisco.com>
	|Subject:Re=3A=20[NSIS]=20QSPEC=20Face-to-Face=20Chat=3A=20Another=20Try;
	X=v=3Dcisco.com=3B=20h=3DgwbH9sdHc95I5M8poltyif0coJ0=3D;
	b=PyZ4P5wy8xsRAaFmiuyo0ycyPAHS2j0u68JYkFvNfOfdUQn2xKGRIgQK6oUnnIfgLideIUss
	S0lCPEBFr/S1TR9Znz3FaaNMoOiywPKRLElcWwagl/85gDFj8rJUTeU5;
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 5a9a1bd6c2d06a21d748b7d0070ddcb8
Cc: NSIS <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

and me.

On Nov 3, 2006, at 4:40 AM, Jukka MJ Manner wrote:

>
> Works for me.
>
> Jukka
>
> On Fri, 3 Nov 2006, Hannes Tschofenig wrote:
>
>> Hi all,
>>
>> let me try it again.
>>
>> What about Sunday, 3pm at the Sheraton San Diego Hotel (= meeting  
>> venue) in
>> front of the hotel reception.
>>
>> Would this work for someone?
>>
>> Ciao
>> Hannes
>>
>> Hannes Tschofenig wrote:
>>> Hi all,
>>>
>>> I wonder how many NSIS folks are available for a QSPEC face-to-face
>>> discussion Sunday morning?
>>>
>>> Since many folks have to be at company internal meetings during  
>>> lunch time
>>> and want to see some of the tutorials I would suggest 9am at the
>>> Sheraton San Diego Hotel (= meeting venue) in front of the hotel  
>>> reception.
>>>
>>> Is this a useful idea?
>>>
>>> Ciao
>>> Hannes
>>>
>>> _______________________________________________
>>> nsis mailing list
>>> nsis@ietf.org
>>> https://www1.ietf.org/mailman/listinfo/nsis
>>
>>
>> _______________________________________________
>> nsis mailing list
>> nsis@ietf.org
>> https://www1.ietf.org/mailman/listinfo/nsis
>>
>
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 20:08:40 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgA0c-0005Q8-NI; Fri, 03 Nov 2006 20:07:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgA0b-0005Nk-4c
	for nsis@ietf.org; Fri, 03 Nov 2006 20:07:41 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgA0Y-0008UG-RG
	for nsis@ietf.org; Fri, 03 Nov 2006 20:07:41 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Fri, 3 Nov 2006 20:07:24 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006110320071905364 ; Fri, 03 Nov 2006 20:07:20 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <VX80R5QT>; Fri, 3 Nov 2006 20:07:18 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF00@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: David R Oran <oran@cisco.com>,
	Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: RE: [NSIS] jerry, dave,cornelia's proposal very confusing and not
	solving problem
Date: Fri, 3 Nov 2006 20:07:15 -0500 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 04 Nov 2006 01:07:16.0070 (UTC)
	FILETIME=[9169E860:01C6FFAD]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 769a46790fb42fbb0b0cc700c82f7081
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Inline [RRR]

  _____  

From: nsis-bounces@ietf.org on behalf of David R Oran
Sent: Fri 11/3/2006 8:43 AM
To: Georgios Karagiannis
Cc: nsis@ietf.org
Subject: Re: [NSIS] jerry, dave,cornelia's proposal very confusing and not
solving problem



> After that we can submit the QSPEC template draft for last WG call. If
> after the deployment phase we conclude that more mandatory parameters
> are needed for interoperability then we could work on an interworking
> draft.
>
The fact we might need interworking drafts is a big red flag for me 
anyway.

[RRR] Is not the case the whole QSPEC draft shows how all these RSVP,
DiffServ, IntServ, Y.1541, 3GPP/Y.1541, and DSTE are interworking? Does the
draft not show that QSPEC template has a magic for interworking among all of
them? 

[RRR] The question is this: Can we we not find which ones interwork and
which ones are not? If so, why should the NSIS QSPEC draft deals with those
QOS models that do NOT interwork among themselves?

[RRR] If there are some "bad"apples of QOS models have been injected using
some "brute" forces knowing they will NEVER interwork, why the whole NSIS WG
members, in fact the whole IETF community, will pay the heavy price of
"NON-Interowkring?"

[RRR] Why is the NSIS QSPEC standard putting the seeds of
"Non-Interoperability"  mixing between "Water" and "Oil" in the same draft
deliberately in a "Standard Track RFC?"

Cheers, DaveO.

> --------------
>
> Best Regards,
> Georgios
>



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 20:32:53 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgAOB-0000XS-Fh; Fri, 03 Nov 2006 20:32:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgAOA-0000WL-8K
	for nsis@ietf.org; Fri, 03 Nov 2006 20:32:02 -0500
Received: from mclmx2.mail.saic.com ([149.8.64.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgAO8-0003Ye-TL
	for nsis@ietf.org; Fri, 03 Nov 2006 20:32:02 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx2.mail.saic.com; Fri, 3 Nov 2006 20:31:44 -0500
Received: from 0015-ITS-EXBH01.us.saic.com ([10.43.229.18])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006110320314405271 ; Fri, 03 Nov 2006 20:31:44 -0500
Received: from 0591-ITS-EXMP02.us.saic.com ([10.42.81.248]) by
	0015-ITS-EXBH01.us.saic.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 3 Nov 2006 20:31:44 -0500
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Fri, 3 Nov 2006 20:31:43 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF01@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: Acb/khubB9CXtfvFQXu1Ug8UkeR+UAAG74R+
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl><F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com><454B1268.4030405@ericsson.com>
	<8203F813-69C3-4BBE-ADF2-8DC009E0F87C@cisco.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: "David R Oran" <oran@cisco.com>,
	"Lars Westberg" <Lars.westberg@ericsson.com>
X-OriginalArrivalTime: 04 Nov 2006 01:31:44.0806 (UTC)
	FILETIME=[FCD93C60:01C6FFB0]
X-Spam-Score: 0.5 (/)
X-Scan-Signature: 7da5a831c477fb6ef97f379a05fb683c
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1889664083=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1889664083==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C6FFB0.FC50EBD2"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C6FFB0.FC50EBD2
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Inline [RRR]

________________________________

From: nsis-bounces@ietf.org on behalf of David R Oran
Sent: Fri 11/3/2006 7:32 AM
To: Lars Westberg
Cc: NSIS Group
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft




On Nov 3, 2006, at 4:56 AM, Lars Westberg wrote:

My assessment (again - let me emphasize that=20
it's a personal view) was that the lack of closure over the=20
parameters in fact revealed some deeper architectural difficulties,=20
especially around QOS models and the role they play in the NSIS NSLP=20
and QSPEC design.

[RRR] Here we are loosing our sights. We have generalized all QOS Models =
in the "same" level. It is the root cause of all problems. We must NOT =
treat "apple" and "oragnge" in the same level.

[RRR] Let me ask one fundamental question: What are the properties of a =
given QOS model that characterize it as QOS model? If we do not analyze =
this fundamental problem, then we will NEVER be able be to solve any QOS =
problems.

[RRR] Did you ever has the chance to comapre the properties of so-called =
QOS models: Y.1541, 3GPP/Y.1541, and DSTE (group I)? Do they the same =
properties like those of RSVP, DiffServ, and IntServ (Group II)? Are the =
properties of this group I not fundamentally different?

[RRR] Again, all QOS models of Group II (RSVP, DiffServ, and IntServ) =
interwork among themselves. So, their QSPEC spec also interwork among =
theselves. So, it is the perfect goal of the IETF and NSIS WG and we are =
done.

[RRR] All QOS models of Group I (Y.1541, 3GPP/Y.1541, and DSTE) do not =
interwork among themselves. So, their QSPEC spec will also NEVER work =
among themselves. The interworking anomg Groups I and II QOS models is a =
DREAM to internity.

[RRR] What is the conclusion after 2 yeasrs of hard works of all =
members? The answer is this: Drop all bad apples of QOS models of Group =
I (Y.1541, 3GPP/Y.1541, and DSTE ) from the QSPEC draft.

So, if the WG wants to go back to the prior direction, I'll just=20
crawl back in my hole and apologize for rocking the boat.

[RRR] You did a very hard work. What you need to suggest is this: Drop =
all bad apples of QOS models of Group I (Y.1541, 3GPP/Y.1541, and DSTE ) =
from the QSPEC draft. Work with only Group II QoS models (RSVP, =
DiffServ, and IntServ).

[RRR] May also suggest: Do individual QSPEC drafts for Y.1451 and DSTE.


Cheers, Dave.




------_=_NextPart_001_01C6FFB0.FC50EBD2
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<HTML dir=3Dltr><HEAD><TITLE>Re: [NSIS] Proposed Updates to the QSPEC =
Draft</TITLE>=0A=
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dunicode">=0A=
<META content=3D"MSHTML 6.00.2900.2963" name=3DGENERATOR><BASE =
href=3Dhttps://mclmail.saic.com/exchange/royrr/Drafts/RE:%20%5BNSIS%5D%20=
Proposed%20Updates%20to%20the%20QSPEC%20Draft.EML/1_text.htm></HEAD>=0A=
<BODY>=0A=
<DIV id=3DidOWAReplyText74121 dir=3Dltr>=0A=
<DIV dir=3Dltr><FONT face=3DArial color=3D#000000 size=3D2>Inline =
[RRR]</FONT></DIV></DIV>=0A=
<DIV dir=3Dltr><BR>=0A=
<HR tabIndex=3D-1>=0A=
<FONT face=3DTahoma size=3D2><B>From:</B> nsis-bounces@ietf.org on =
behalf of David R Oran<BR><B>Sent:</B> Fri 11/3/2006 7:32 =
AM<BR><B>To:</B> Lars Westberg<BR><B>Cc:</B> NSIS =
Group<BR><B>Subject:</B> Re: [NSIS] Proposed Updates to the QSPEC =
Draft<BR></FONT><BR></DIV>=0A=
<DIV><BR>=0A=
<P><FONT size=3D2>On Nov 3, 2006, at 4:56 AM, Lars Westberg =
wrote:<BR><BR>My assessment (again - let me emphasize that&nbsp;<BR>it's =
a personal view) was that the lack of closure over =
the&nbsp;<BR>parameters in fact revealed some deeper architectural =
difficulties,&nbsp;<BR>especially around QOS models and the role they =
play in the NSIS NSLP&nbsp;<BR>and QSPEC design.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] Here we are loosing our sights. We have =
generalized all QOS Models in the "same" level. It is the root cause of =
all problems. We must NOT treat "apple" and "oragnge" in the same =
level.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] Let me ask one fundamental question: What are =
the properties of a given QOS model that characterize it as QOS model? =
If we do not analyze this fundamental problem, then we will NEVER be =
able be to solve any QOS problems.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] Did you ever has the chance to comapre the =
properties of so-called QOS models: Y.1541, 3GPP/Y.1541, and DSTE (group =
I)? Do they the same properties like those of RSVP, DiffServ, and =
IntServ (Group II)? Are the properties of this group I&nbsp;not =
fundamentally different?</FONT></P>=0A=
<P><FONT size=3D2>[RRR] Again, all QOS models of Group II (RSVP, =
DiffServ, and IntServ) interwork among themselves. So, their =
QSPEC&nbsp;spec also interwork among theselves. So, it is the perfect =
goal of the IETF and NSIS WG and we are done.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] All QOS models of Group I (Y.1541, 3GPP/Y.1541, =
and DSTE) do not interwork among themselves. So, their QSPEC spec will =
also NEVER work among themselves. The interworking anomg Groups I and II =
QOS models&nbsp;is&nbsp;a DREAM to internity.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] What is the conclusion after 2 yeasrs of hard =
works of all members? The answer is this: Drop all bad apples of QOS =
models of Group I (Y.1541, 3GPP/Y.1541, and DSTE ) from the QSPEC =
draft.<BR><BR>So, if the WG wants to go back to the prior direction, =
I'll just&nbsp;<BR>crawl back in my hole and apologize for rocking the =
boat.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] You did a very hard work. What you need to =
suggest is this: Drop all bad apples of QOS models of Group I (Y.1541, =
3GPP/Y.1541, and DSTE ) from the QSPEC draft. Work with only Group II =
QoS models (RSVP, DiffServ, and IntServ).</FONT></P>=0A=
<P><FONT size=3D2>[RRR] May also suggest: Do individual QSPEC drafts for =
Y.1451 and DSTE.<BR><BR><BR>Cheers, =
Dave.<BR><BR></P></FONT></DIV></BODY></HTML>
------_=_NextPart_001_01C6FFB0.FC50EBD2--


--===============1889664083==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1889664083==--




From nsis-bounces@ietf.org Fri Nov 03 21:09:25 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgAxJ-0003nj-On; Fri, 03 Nov 2006 21:08:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgAxI-0003ne-9z
	for nsis@ietf.org; Fri, 03 Nov 2006 21:08:20 -0500
Received: from mclmx2.mail.saic.com ([149.8.64.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgAxF-0000SY-0V
	for nsis@ietf.org; Fri, 03 Nov 2006 21:08:20 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx2.mail.saic.com; Fri, 3 Nov 2006 21:08:03 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006110321080313874 ; Fri, 03 Nov 2006 21:08:03 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <VX80R7VW>; Fri, 3 Nov 2006 21:08:03 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF03@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: Fred Baker <fred@cisco.com>, Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: RE: [NSIS] jerry, dave,cornelia's proposal very confusing and not
	solving problem
Date: Fri, 3 Nov 2006 21:07:56 -0500 
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 04 Nov 2006 02:07:57.0031 (UTC)
	FILETIME=[0B987770:01C6FFB6]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e8a67952aa972b528dd04570d58ad8fe
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

I understand the old story ....
 
Let us also see that this QSPEC draft is dealing with "Interoperability."
Please see that the draft itself has used the "interoperability" word
probably more 100 times. Please the all the big figures of
"Interoperability."
 
That is, the draft itself is nothing but "interoperability."
 
If it is so, why are we forcing "Non-Interoperability" in the first place?
If there is a way to solve this problem, what is the problem? One of the
suggestions may be look into the thread as follows:
 

http://www1.ietf.org/mail-archive/web/nsis/current/msg07149.html
<http://www1.ietf.org/mail-archive/web/nsis/current/msg07149.html> 

 


  _____  

From: nsis-bounces@ietf.org on behalf of Fred Baker
Sent: Fri 11/3/2006 2:15 PM
To: Georgios Karagiannis
Cc: nsis@ietf.org
Subject: Re: [NSIS] jerry, dave,cornelia's proposal very confusing and not
solving problem



On Nov 2, 2006, at 10:27 PM, Georgios Karagiannis wrote:
> The proposal mentions in point 2) point B):
>
> "All other QSPEC parameters are optional to include by the QNI, and 
> may
> be either mandatory or optional to interpret by downstream QNEs."

If I may point out a problem...

Generally speaking, optional capabilities create a diode. Some of us 
will recall the antics os the ISO protocol suite in the 1980's. It 
was in many ways a good piece of work, and there are some bits (IS-IS 
and ASN.1 come to mind, and the conceptual carry-over of ES-IS into 
IPv6 ND) that are still in use. The big thing that made it hard to 
use and ultimately triggered its collapse was the optional sets of 
parameters - every product built was compliant under some profile of 
selected options or another, but few were really interoperable. As a 
result, at the end of the day the solution didn't work.

With something where one has multiple products from multiple vendors 
that wind up constituting a set of needles that a transaction has to 
simultaneously thread, optional capabilities are a real liability. In 
the end, it is the mandatory-to-implement-and-use subset that will be 
interoperable end to end.

I would really suggest that we use optional capabilities as an 
upgrade path - if someone doesn't understand an option it will be 
passed along unchanged - but in the general case avoid optional 
capabilities like the plague.




_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 03 22:05:43 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgBpr-0000P4-9g; Fri, 03 Nov 2006 22:04:43 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgBpp-0000ME-Nf
	for nsis@ietf.org; Fri, 03 Nov 2006 22:04:41 -0500
Received: from sj-iport-6.cisco.com ([171.71.176.117])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgBpo-00026M-DL
	for nsis@ietf.org; Fri, 03 Nov 2006 22:04:41 -0500
Received: from sj-dkim-3.cisco.com ([171.71.179.195])
	by sj-iport-6.cisco.com with ESMTP; 03 Nov 2006 19:04:39 -0800
X-IronPort-AV: i="4.09,387,1157353200"; 
	d="scan'208"; a="85368839:sNHT49729671"
Received: from sj-core-5.cisco.com (sj-core-5.cisco.com [171.71.177.238])
	by sj-dkim-3.cisco.com (8.12.11.20060308/8.12.11) with ESMTP id
	kA434dQd020351; Fri, 3 Nov 2006 19:04:39 -0800
Received: from imail.cisco.com (sjc12-sbr-sw3-3f5.cisco.com [172.19.96.182])
	by sj-core-5.cisco.com (8.12.10/8.12.6) with ESMTP id kA434dW4007351;
	Fri, 3 Nov 2006 19:04:39 -0800 (PST)
Received: from [10.71.2.136] (sjc-vpn6-20.cisco.com [10.21.120.20])
	by imail.cisco.com (8.12.11/8.12.10) with ESMTP id kA42pc77027596;
	Fri, 3 Nov 2006 18:51:38 -0800
In-Reply-To: <454B58F3.4070709@tm.uka.de>
References: <002b01c6ff49$494b87a0$4c0d5982@dynamic.ewi.utwente.nl>
	<454B58F3.4070709@tm.uka.de>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <E9A03781-490C-45CA-9B36-ABA8B4FE8033@cisco.com>
Content-Transfer-Encoding: 7bit
From: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] jerry, dave,
	cornelia's proposal very confusing and not solving problem
Date: Fri, 3 Nov 2006 22:04:31 -0500
To: Roland Bless <bless@tm.uka.de>
X-Mailer: Apple Mail (2.752.2)
Authentication-Results: sj-dkim-3.cisco.com; header.From=oran@cisco.com;
	dkim=pass ( sig from cisco.com verified; ); 
DKIM-Signature: a=rsa-sha1; q=dns; l=1882; t=1162609479; x=1163473479;
	c=relaxed/simple; s=sjdkim3002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=oran@cisco.com;
	z=From:David=20R=20Oran=20<oran@cisco.com>
	|Subject:Re=3A=20[NSIS]=20jerry, =20dave,
	=09cornelia's=20proposal=20very=20confusi
	ng=20and=20not=20solving=20problem;
	X=v=3Dcisco.com=3B=20h=3DL/H77xCILBzp9ZIoXcXixiYTGqM=3D;
	b=pnBVLjkQ9bB7RiWuDuI8hcQ3WY0Z14H60DWYp3zEJWlHPpDiv9HI22/AV9Fu8GbyODM+GYje
	g4vx/Hxo14AJpTbYUmxUNfySXoZ6K9aYu/SV+PwmU77QPuMH7zTMeUv6;
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0ddefe323dd869ab027dbfff7eff0465
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


On Nov 3, 2006, at 9:57 AM, Roland Bless wrote:

> Hi Georgios,
>
> Georgios Karagiannis wrote:
>> So the alternative 2) holds, meaning:
>>  2) If the above is not true then it means that a QoS model
>> that needs to use only the Token Bucket peak bandwidth parameter  
>> it will not
>> be
>> able to use the token bucket specification, thus it will not be able
>> to use NSIS.
>>
>> Thanks for the clarification!
>> This means that Diffserv based QoS models that need to use only  
>> bandwidth
>> will not be able to use NSIS.
>
> I guess that there is a lack of understanding DiffServ here.
> DiffServ is not restricted to EF-based services only. If you
> perform admission control based on the peak rate only, it's
> your engineering decision. Some other people may take a look
> at the whole token bucket parameters and achieve higher utilization.
> It's up to the signaled constraints and your local QoS model.
> For other services you might be more interested in the burstiness  
> of the
> traffic, too. Depending on the constraints that are signaled,
> you are free to map the request into any appropriate DiffServ
> class (PDB/PHB set) that is able to meet the signaled constraints.
> So I guess the new proposal is very well suited to support also
> DiffServ-based QoS support in domains.
>
YES! Extremely well put. Thank you.

>> I really do not understand what kind of proposal is this??????
>
> If you care about stacking:
> Why not defining your own QSPEC that allows to carry your
> domain-specific parameters as well as the original encapsulated
> external QSPEC. This would be a form of encapsulation, but it
> could probably be used instead of stacking.
>
> Regards,
>  Roland
>
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sat Nov 04 16:21:31 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgSw6-00010E-Od; Sat, 04 Nov 2006 16:20:18 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgSw4-0000ya-NP
	for nsis@ietf.org; Sat, 04 Nov 2006 16:20:16 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgSsH-0001vh-MP
	for nsis@ietf.org; Sat, 04 Nov 2006 16:16:23 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GgSs9-0001KG-G2; Sat, 04 Nov 2006 22:16:14 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GgSs9-00062i-3C; Sat, 04 Nov 2006 22:16:13 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 7F5FCB594;
	Sat,  4 Nov 2006 22:16:12 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GgSs5-0005mK-TE; Sat, 04 Nov 2006 22:16:12 +0100
Message-ID: <454D02E6.2090505@tm.uka.de>
Date: Sat, 04 Nov 2006 22:15:18 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.7) Gecko/20060909 Thunderbird/1.5.0.7 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>	<F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>	<454B1268.4030405@ericsson.com>
	<8203F813-69C3-4BBE-ADF2-8DC009E0F87C@cisco.com>
In-Reply-To: <8203F813-69C3-4BBE-ADF2-8DC009E0F87C@cisco.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 82c9bddb247d9ba4471160a9a865a5f3
Cc: Lars Westberg <Lars.westberg@ericsson.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi,

David R Oran wrote:
> IESG). My assessment (again - let me emphasize that it's a personal
> view) was that the lack of closure over the parameters in fact revealed
> some deeper architectural difficulties, especially around QOS models and
> the role they play in the NSIS NSLP and QSPEC design.

Exactly.

> I happen to fundamentally disagree with the notion that different QoS
> models require different traffic descriptions by the traffic source. The
> traffic generation process is what it is, and defining different QoS
> models that preserve the exact source traffic generation pattern at the
> receiver to differing degrees of accuracy is perfectly appropriate.
> However, if you alter the source traffic description in the middle
> through stacking or remapping this property is lost.

Yep.

> I think that the the situation where "We have today parameters and
> signalling sequence diagram defined differently in each standards-body
> especially for different access networks" is a problem, not a good
> thing, and the way to deal with it is not to import every one of those
> into NSIS, but rather to map "down" to the particular domain's model in
> individual QNEs. If they in turn need to do some private signaling among
> themselves (e.g. DOCSIS, 802.11e) that's their business.

Agree.

> I also disagree with the notion that by choosing a 4-parameter traffic
> model for describing traffic sources somehow means that you are
> constrained to an INTSERV QOS model in terms of the QNEs and their RMFs
> that actually admit and handle the traffic. The QOSM (to my way of
> thinking) defined by what constraints and handling directives can be
> accommodated for the stated traffic description, not that it imposes a
> traffic model or description on the source.

Yes.


> Understood. However, what I read into this is that RMD is a QoS
> signaling architecture of its own and the desire is to co-mingle it with
> NSIS. That runs counter to the goal of universal interoperability, IMO.

That matches nearly exactly my opionion which I expressed once at
the first interim meeting in Helsinki back in 2002. I said that this
doesn't need to be standardized in NSIS in the first place, because
it's intra-domain edge-to-edge signaling where you can use any
proprietary protocol. I felt that NSIS would tackle also more the
inter-domain issues in QoS signaling (at least when reading the old
charter).

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sun Nov 05 10:58:35 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgkNn-0004IK-LU; Sun, 05 Nov 2006 10:58:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgkNm-0004IF-K4
	for nsis@ietf.org; Sun, 05 Nov 2006 10:58:02 -0500
Received: from mclmx2.mail.saic.com ([149.8.64.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgkNk-0002Ha-VQ
	for nsis@ietf.org; Sun, 05 Nov 2006 10:58:02 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx2.mail.saic.com; Sun, 5 Nov 2006 10:57:28 -0500
Received: from 0015-ITS-EXBH01.us.saic.com ([10.43.229.18])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006110510572026034 ; Sun, 05 Nov 2006 10:57:22 -0500
Received: from 0591-ITS-EXMP02.us.saic.com ([10.42.81.248]) by
	0015-ITS-EXBH01.us.saic.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Sun, 5 Nov 2006 10:57:20 -0500
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] Proposed Updates to the QSPEC Draft
Date: Sun, 5 Nov 2006 10:57:19 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF04@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Proposed Updates to the QSPEC Draft
Thread-Index: AccAV6cr2iSHf1EaS/C45cJ1c7CHBAAk+OfJ
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>	<F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>	<454B1268.4030405@ericsson.com>
	<8203F813-69C3-4BBE-ADF2-8DC009E0F87C@cisco.com>
	<454D02E6.2090505@tm.uka.de>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: "Roland Bless" <bless@tm.uka.de>,
    "David R Oran" <oran@cisco.com>
X-OriginalArrivalTime: 05 Nov 2006 15:57:20.0185 (UTC)
	FILETIME=[1328EA90:01C700F3]
X-Spam-Score: 0.3 (/)
X-Scan-Signature: dd887a8966a4c4c217a52303814d0b5f
Cc: Lars Westberg <Lars.westberg@ericsson.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0161499500=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============0161499500==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C700F3.12BDB7F8"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C700F3.12BDB7F8
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Inline [RRR]

________________________________

From: nsis-bounces@ietf.org on behalf of Roland Bless
Sent: Sat 11/4/2006 4:15 PM
To: David R Oran
Cc: Lars Westberg; NSIS Group
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft



Hi,

David R Oran wrote:
> IESG). My assessment (again - let me emphasize that it's a personal
> view) was that the lack of closure over the parameters in fact =
revealed
> some deeper architectural difficulties, especially around QOS models =
and
> the role they play in the NSIS NSLP and QSPEC design.

Exactly.

[RRR] Now it reveals the fact that there is no socpe for =
interoperability in using different QOS models. If it is so, did Dave =
and Roland analyze the root cause of this interoperability problem?

[RRR] If you folks do not analyze for each QOS model and want to =
generalize for all QOS models, you are NOT doing your technical analysis =
correctly and are sending the WRONG message to the entire NSIS technical =
community mixing apples and oragnes as if all QOS models have the same =
properties.

[RRR} Did you folks know that RSVP, DiffServ, and IntServ QOS models =
interowrk among themselves using QSPEC template with NSIS QOS signaling =
messages. If so, why are you not ponting out this?

[RRR} Is it not the case the so-called QOS models like Y.1451, =
3GPP/Y.1541, and DSTE that do NOT qualify to be QOS models in the first =
place are the root cause of all problems? Why do you not speak out for =
these cases specifically? What is your problem?

> I happen to fundamentally disagree with the notion that different QoS
> models require different traffic descriptions by the traffic source. =
The
> traffic generation process is what it is, and defining different QoS
> models that preserve the exact source traffic generation pattern at =
the
> receiver to differing degrees of accuracy is perfectly appropriate.
> However, if you alter the source traffic description in the middle
> through stacking or remapping this property is lost.

Yep.

[RRR] Again, it is a too general statement. Let us taken an example and =
analyze for each QOS model and examine where and how each QOS model =
fails. If we do not do this, there is no clear conlusion in support of =
this to "generic" statement.

[RRR] If you analyze each QOS model, you can easilty see the three QOS =
mdoels RSVP, IntServ, and DiffServ are in synch in all respects. There =
are no ambigutiies to deal with either with traffic descriptions or or =
any QOS parameters because they have standardized the basic QOS =
characteristics for which there are proper mapping can be done. In fact, =
it is happening since their inception. NSIS QOS signaling is jus doing =
fine with these three QOS models. Do you see what has been the basic =
objective of NSIS QOS signaling? Is it not that the NSIS QOS signaling =
is the SECOND vesrion of RSVP with more "scalability" for end-to-end QOS =
signaling?

[RRR] If you folks forget these basic objectives of NSIS QOS signaling =
and want create unnecessary problems bring the so-called hard-coded some =
implementation-specific QOS parameters with any fundamental QOS =
properties like Y.1541 and DSTE, how can you be "objective" in solving =
the QOS interoperability problem for e2e QOS in NSIS for which it is =
being standardized?


> I think that the the situation where "We have today parameters and
> signalling sequence diagram defined differently in each standards-body
> especially for different access networks" is a problem, not a good
> thing, and the way to deal with it is not to import every one of those
> into NSIS, but rather to map "down" to the particular domain's model =
in
> individual QNEs. If they in turn need to do some private signaling =
among
> themselves (e.g. DOCSIS, 802.11e) that's their business.

Agree.


[RRR] It is not true when it is gemeralized with the things mixing =
between "apples" and "oranges." Kindly see the technical explanations as =
ststed above.


> I also disagree with the notion that by choosing a 4-parameter traffic
> model for describing traffic sources somehow means that you are
> constrained to an INTSERV QOS model in terms of the QNEs and their =
RMFs
> that actually admit and handle the traffic. The QOSM (to my way of
> thinking) defined by what constraints and handling directives can be
> accommodated for the stated traffic description, not that it imposes a
> traffic model or description on the source.

Yes.

[RRR] Can you provide an example taking the RSVP, IntServ, and DiffServ =
and show where are the problems and we can understand more clearly?


> Understood. However, what I read into this is that RMD is a QoS
> signaling architecture of its own and the desire is to co-mingle it =
with
> NSIS. That runs counter to the goal of universal interoperability, =
IMO.

That matches nearly exactly my opionion which I expressed once at
the first interim meeting in Helsinki back in 2002. I said that this
doesn't need to be standardized in NSIS in the first place, because
it's intra-domain edge-to-edge signaling where you can use any
proprietary protocol. I felt that NSIS would tackle also more the
inter-domain issues in QoS signaling (at least when reading the old
charter).

[RRR] I do not see this as it is explaned here. The basic idea of NSIS =
QOS signaling protocol has been end-to-end signaling. NSIS charter is =
clear about this. No only this, NSIS has clear code points how each =
application will be able to signal its desired QOS in NSIS upper layer. =
The specific application sitting in both ends of the network will be =
able to communicate transparently for providing the end-to-end QOS. As I =
explained to Ronald last time, NSIS is not just e2e network layer =
signaling.

[RRR] If NSIS is NOT developing the e2e QOS between the application =
connected to the two ends of the network, there will probably no need =
for another protocol like NSIS QOS signaling because the traditional QOS =
signaling with RSVP/IntServ in the access network and DiffServ in the =
backbone network is doing just fine.

[RRR] After all these years of hard works when NSIS signaling is robust =
enough to provide the end-to-end QOS signaling, and QSPEC templates that =
are quite OK for the RSVP, IntServ, and DiffServ QOS models for =
interoperability, why you two folks want to turn the wheels back?

[RRR] Can you two folks provide specific clarifications why you think =
that the so-called QOS models like Y.1541, 3GPP/Y.1451 and DSTE have the =
smae properties like those of RSVP, IntServ, and DiffServ? If not, why =
do you two folks just say the the obvious truth that these Y.1541, =
3GPP/Y.1541, and DSTE are NOT be considered as QOS models?

[RRR] If you two folks want to generalize mixing apples and oranges =
together, there will be no e2e QOS or no QOS interoperability, and NSIS =
QOS signaling will be of little for using for any practical purpose.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



------_=_NextPart_001_01C700F3.12BDB7F8
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<HTML dir=3Dltr><HEAD><TITLE>Re: [NSIS] Proposed Updates to the QSPEC =
Draft</TITLE>=0A=
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dunicode">=0A=
<META content=3D"MSHTML 6.00.2900.2963" name=3DGENERATOR><BASE =
href=3Dhttps://mclmail.saic.com/exchange/royrr/Drafts/RE:%20%5BNSIS%5D%20=
Proposed%20Updates%20to%20the%20QSPEC%20Draft.EML/1_text.htm></HEAD>=0A=
<BODY>=0A=
<DIV id=3DidOWAReplyText44284 dir=3Dltr>=0A=
<DIV dir=3Dltr><FONT face=3DArial color=3D#000000 size=3D2>Inline =
[RRR]</FONT></DIV></DIV>=0A=
<DIV dir=3Dltr><BR>=0A=
<HR tabIndex=3D-1>=0A=
<FONT face=3DTahoma size=3D2><B>From:</B> nsis-bounces@ietf.org on =
behalf of Roland Bless<BR><B>Sent:</B> Sat 11/4/2006 4:15 =
PM<BR><B>To:</B> David R Oran<BR><B>Cc:</B> Lars Westberg; NSIS =
Group<BR><B>Subject:</B> Re: [NSIS] Proposed Updates to the QSPEC =
Draft<BR></FONT><BR></DIV>=0A=
<DIV>=0A=
<P><FONT size=3D2>Hi,<BR><BR>David R Oran wrote:<BR>&gt; IESG). My =
assessment (again - let me emphasize that it's a personal<BR>&gt; view) =
was that the lack of closure over the parameters in fact =
revealed<BR>&gt; some deeper architectural difficulties, especially =
around QOS models and<BR>&gt; the role they play in the NSIS NSLP and =
QSPEC design.<BR><BR>Exactly.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] Now it reveals the fact that there is no socpe =
for interoperability in using different QOS models. If it is so, did =
Dave and Roland analyze the root cause of this interoperability =
problem?</FONT></P>=0A=
<P><FONT size=3D2>[RRR] If you folks do not analyze for each QOS model =
and want to generalize for all QOS models, you are NOT doing your =
technical analysis correctly and are sending the WRONG message to the =
entire NSIS technical community mixing apples and oragnes as if all QOS =
models have the same properties.</FONT></P>=0A=
<P><FONT size=3D2>[RRR} Did you folks know that RSVP, DiffServ, and =
IntServ QOS models interowrk among themselves using QSPEC template with =
NSIS QOS signaling messages. If so, why are you not ponting out =
this?</FONT></P>=0A=
<P><FONT size=3D2>[RRR} Is it not the case the so-called QOS models like =
Y.1451, 3GPP/Y.1541, and DSTE that do NOT qualify to be QOS models in =
the first place are the root cause of all problems? Why do you not speak =
out for these cases specifically? What is your problem?<BR><BR>&gt; I =
happen to fundamentally disagree with the notion that different =
QoS<BR>&gt; models require different traffic descriptions by the traffic =
source. The<BR>&gt; traffic generation process is what it is, and =
defining different QoS<BR>&gt; models that preserve the exact source =
traffic generation pattern at the<BR>&gt; receiver to differing degrees =
of accuracy is perfectly appropriate.<BR>&gt; However, if you alter the =
source traffic description in the middle<BR>&gt; through stacking or =
remapping this property is lost.<BR><BR>Yep.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] Again, it is a too general statement. Let us =
taken an example and analyze for each QOS model and examine where and =
how each QOS model fails. If we do not do this, there is no clear =
conlusion in support of this to "generic" statement.</FONT></P>=0A=
<P><FONT size=3D2>[RRR] If you analyze each QOS model, you can easilty =
see the three QOS mdoels RSVP, IntServ, and DiffServ are in synch in all =
respects. There are no ambigutiies to deal with either with traffic =
descriptions or or any QOS parameters because they have standardized the =
basic QOS characteristics for which there are proper mapping can be =
done. In fact, it is happening since their inception. NSIS QOS signaling =
is jus doing fine with these three QOS models. Do you see what has been =
the basic objective of NSIS QOS signaling? Is it not that the NSIS QOS =
signaling is the SECOND vesrion of RSVP with more "scalability" for =
end-to-end QOS signaling?</FONT></P>=0A=
<P><FONT size=3D2>[RRR] If you folks forget these basic objectives of =
NSIS QOS signaling and want create unnecessary problems bring the =
so-called hard-coded some implementation-specific QOS parameters with =
any fundamental QOS properties like Y.1541 and DSTE, how can you be =
"objective" in solving the QOS interoperability problem for e2e QOS in =
NSIS for which it is being standardized?</FONT></P><FONT =
size=3D2></FONT></DIV>=0A=
<DIV><FONT size=3D2>=0A=
<P><BR>&gt; I think that the the situation where "We have today =
parameters and<BR>&gt; signalling sequence diagram defined differently =
in each standards-body<BR>&gt; especially for different access networks" =
is a problem, not a good<BR>&gt; thing, and the way to deal with it is =
not to import every one of those<BR>&gt; into NSIS, but rather to map =
"down" to the particular domain's model in<BR>&gt; individual QNEs. If =
they in turn need to do some private signaling among<BR>&gt; themselves =
(e.g. DOCSIS, 802.11e) that's their business.<BR><BR>Agree.<BR></P>=0A=
<P>[RRR] It is not true when it is gemeralized with the things mixing =
between "apples" and "oranges." Kindly see the technical explanations as =
ststed above.</P>=0A=
<P><BR>&gt; I also disagree with the notion that by choosing a =
4-parameter traffic<BR>&gt; model for describing traffic sources somehow =
means that you are<BR>&gt; constrained to an INTSERV QOS model in terms =
of the QNEs and their RMFs<BR>&gt; that actually admit and handle the =
traffic. The QOSM (to my way of<BR>&gt; thinking) defined by what =
constraints and handling directives can be<BR>&gt; accommodated for the =
stated traffic description, not that it imposes a<BR>&gt; traffic model =
or description on the source.<BR><BR>Yes.</P>=0A=
<P>[RRR] Can you provide an example taking the RSVP, IntServ, and =
DiffServ and show where are the problems and we can understand more =
clearly?<BR><BR><BR>&gt; Understood. However, what I read into this is =
that RMD is a QoS<BR>&gt; signaling architecture of its own and the =
desire is to co-mingle it with<BR>&gt; NSIS. That runs counter to the =
goal of universal interoperability, IMO.<BR><BR>That matches nearly =
exactly my opionion which I expressed once at<BR>the first interim =
meeting in Helsinki back in 2002. I said that this<BR>doesn't need to be =
standardized in NSIS in the first place, because<BR>it's intra-domain =
edge-to-edge signaling where you can use any<BR>proprietary protocol. I =
felt that NSIS would tackle also more the<BR>inter-domain issues in QoS =
signaling (at least when reading the old<BR>charter).</P>=0A=
<P>[RRR] I do not see this as it is explaned here. The basic idea of =
NSIS QOS signaling protocol has been end-to-end signaling. NSIS charter =
is clear about this. No only this, NSIS has clear code points how each =
application will be able to signal its desired QOS in NSIS upper layer. =
The specific application sitting in both ends of the network will be =
able to communicate transparently for providing the end-to-end QOS. As I =
explained to Ronald last time, NSIS is not just e2e network layer =
signaling.</P>=0A=
<P>[RRR] If NSIS is NOT developing the e2e QOS between the application =
connected to the two ends of the network, there will probably no need =
for another protocol like NSIS QOS signaling because the traditional QOS =
signaling with RSVP/IntServ in the access network and DiffServ in the =
backbone network is doing just fine.</P>=0A=
<P>[RRR] After all these years of hard works when NSIS signaling is =
robust enough to provide the end-to-end QOS signaling, and QSPEC =
templates that are quite OK for the RSVP, IntServ, and DiffServ QOS =
models for interoperability, why you two folks want to turn the wheels =
back?</P>=0A=
<P>[RRR] Can you two folks provide specific clarifications why you think =
that the so-called QOS models like Y.1541, 3GPP/Y.1451 and DSTE have the =
smae properties like those of RSVP, IntServ, and DiffServ? If not, why =
do you two folks just say the the obvious truth that these Y.1541, =
3GPP/Y.1541, and DSTE are&nbsp;NOT be considered as QOS models?</P>=0A=
<P>[RRR] If you two folks want to generalize mixing apples and oranges =
together, there will be no e2e QOS or no QOS interoperability, and NSIS =
QOS&nbsp;signaling will be of little for using for any practical =
purpose.<BR><BR>Regards,<BR>&nbsp;Roland<BR><BR><BR>_____________________=
__________________________<BR>nsis mailing list<BR>nsis@ietf.org<BR><A =
href=3D"https://www1.ietf.org/mailman/listinfo/nsis">https://www1.ietf.or=
g/mailman/listinfo/nsis</A><BR></P></FONT></DIV></BODY></HTML>
------_=_NextPart_001_01C700F3.12BDB7F8--


--===============0161499500==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============0161499500==--




From nsis-bounces@ietf.org Sun Nov 05 14:17:18 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgnTs-00030T-BL; Sun, 05 Nov 2006 14:16:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgnTq-00030O-Ok
	for nsis@ietf.org; Sun, 05 Nov 2006 14:16:30 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgnTm-0004NQ-Uk
	for nsis@ietf.org; Sun, 05 Nov 2006 14:16:30 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GgnTd-0006a1-Fy; Sun, 05 Nov 2006 20:16:20 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GgnTd-0007O4-27; Sun, 05 Nov 2006 20:16:17 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 31469B5AA;
	Sun,  5 Nov 2006 20:16:16 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GgnTb-000782-BO; Sun, 05 Nov 2006 20:16:16 +0100
Message-ID: <454E384F.50006@tm.uka.de>
Date: Sun, 05 Nov 2006 20:15:27 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.7) Gecko/20060909 Thunderbird/1.5.0.7 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
References: <001001c6fe6e$1b50b6d0$4c0d5982@dynamic.ewi.utwente.nl>	<F0FE7CE5-315B-43B0-97C3-901D63D6DE93@cisco.com>	<454B1268.4030405@ericsson.com>
	<8203F813-69C3-4BBE-ADF2-8DC009E0F87C@cisco.com>
	<454D02E6.2090505@tm.uka.de>
	<62D92A9A02BCC845B202323D49A48426E1CF04@0591-ITS-EXMP02.us.saic.com>
In-Reply-To: <62D92A9A02BCC845B202323D49A48426E1CF04@0591-ITS-EXMP02.us.saic.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -2.9 (--)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-1.1 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 963faf56c3a5b6715f0b71b66181e01a
Cc: Lars Westberg <Lars.westberg@ericsson.com>, David R Oran <oran@cisco.com>,
	NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Radhika,

I'm always a little bit puzzled by your postings, but
I'll try to provide answers inline nevertheless.

> [RRR] Now it reveals the fact that there is no socpe for
> interoperability in using different QOS models. If it is so, did Dave
> and Roland analyze the root cause of this interoperability problem?

I guess that it's not generally possible to map any arbitrary QoS
model to all others.

> [RRR] If you folks do not analyze for each QOS model and want to
> generalize for all QOS models, you are NOT doing your technical
> analysis correctly and are sending the WRONG message to the entire
> NSIS technical community mixing apples and oragnes as if all QOS
> models have the same properties.

Sorry, but I guess you are completely misled here.
The idea is to have something like a generic QSPEC which
allows to map to all models. Example:
An end-host signals a Generic QSPEC that
specifies Traffic source (token bucket)
and Constraints, e.g., packet loss or allowed deviation
from average throughput.
So if the first domain now supports Y.1541, you have to define a mapping
in this particular QoS model, e.g. you must choose the appropriate QoS
class to put this reservation in while satisfying the constraints.
If the next domain then supports a QoS model consisting of the
12 DiffServ classes according to RFC4594 this domain must map the
generic QSPEC request to a particular service class in its own domain.
The same procedure will be repeated until you hit the other end host.
However, each QoS model must define a mapping from the generic QSPEC
to its own QoS model (e.g. service classes).

> [RRR} Did you folks know that RSVP, DiffServ, and IntServ QOS models
> interowrk among themselves using QSPEC template with NSIS QOS
> signaling messages. If so, why are you not ponting out this?

Its difficult to put all QoS models into one QSPEC because you
must also define the appropriate mapping between all. This is
difficult because there are some service classes that cannot
be easily mapped to the classes of the another QoS model. Therefore
this huge set of mandatory parameters which leads to nowhere.
Its easier to use a generic QSPEC like the one proposed by Dave
and to let each QoS model map this to its own service classes.
This appears to be more viable to me than the previous approach.

> [RRR} Is it not the case the so-called QOS models like Y.1451,
> 3GPP/Y.1541, and DSTE that do NOT qualify to be QOS models in the
> first place are the root cause of all problems? Why do you not speak
> out for these cases specifically? What is your problem?

Y.1541 is a valid QoS model that a domain may provide. No problem
with this.

> [RRR] Again, it is a too general statement. Let us taken an example
> and analyze for each QOS model and examine where and how each QOS
> model fails. If we do not do this, there is no clear conlusion in
> support of this to "generic" statement.

It's not about failure of QoS models.

> [RRR] If you analyze each QOS model, you can easilty see the three
> QOS mdoels RSVP, IntServ, and DiffServ are in synch in all respects.
> There are no ambigutiies to deal with either with traffic
> descriptions or or any QOS parameters because they have standardized
> the basic QOS characteristics for which there are proper mapping can
> be done. In fact, it is happening since their inception. NSIS QOS
> signaling is jus doing fine with these three QOS models. Do you see
> what has been the basic objective of NSIS QOS signaling? Is it not
> that the NSIS QOS signaling is the SECOND vesrion of RSVP with more
> "scalability" for end-to-end QOS signaling?

Mapping between these models may be easier than between other models.

> [RRR] If you folks forget these basic objectives of NSIS QOS
> signaling and want create unnecessary problems bring the so-called
> hard-coded some implementation-specific QOS parameters with any
> fundamental QOS properties like Y.1541 and DSTE, how can you be
> "objective" in solving the QOS interoperability problem for e2e QOS
> in NSIS for which it is being standardized?

We are not proposing to throw out any existing QoS models.
If you want to use Y.1541, fine. However, you will not be
able to directly signal QoS in your Y.1541 specific manner
from end to end or between domains using different QoS models.
You should stick to the generic QSPEC instead.

>> I think that the the situation where "We have today parameters and 
>> signalling sequence diagram defined differently in each
>> standards-body especially for different access networks" is a
>> problem, not a good thing, and the way to deal with it is not to
>> import every one of those into NSIS, but rather to map "down" to
>> the particular domain's model in individual QNEs. If they in turn
>> need to do some private signaling among themselves (e.g. DOCSIS,
>> 802.11e) that's their business.
> 
> Agree.
> 
> 
> [RRR] It is not true when it is gemeralized with the things mixing
> between "apples" and "oranges." Kindly see the technical explanations
> as ststed above.

???

>> I also disagree with the notion that by choosing a 4-parameter
>> traffic model for describing traffic sources somehow means that you
>> are constrained to an INTSERV QOS model in terms of the QNEs and
>> their RMFs that actually admit and handle the traffic. The QOSM (to
>> my way of thinking) defined by what constraints and handling
>> directives can be accommodated for the stated traffic description,
>> not that it imposes a traffic model or description on the source.
> 
> Yes.
> 
> [RRR] Can you provide an example taking the RSVP, IntServ, and
> DiffServ and show where are the problems and we can understand more
> clearly?

I gave an example above.

>> Understood. However, what I read into this is that RMD is a QoS 
>> signaling architecture of its own and the desire is to co-mingle it
>> with NSIS. That runs counter to the goal of universal
>> interoperability, IMO.
> 
> That matches nearly exactly my opionion which I expressed once at the
> first interim meeting in Helsinki back in 2002. I said that this 
> doesn't need to be standardized in NSIS in the first place, because 
> it's intra-domain edge-to-edge signaling where you can use any 
> proprietary protocol. I felt that NSIS would tackle also more the 
> inter-domain issues in QoS signaling (at least when reading the old 
> charter).
> 
> [RRR] I do not see this as it is explaned here. The basic idea of
> NSIS QOS signaling protocol has been end-to-end signaling. NSIS
> charter is clear about this. No only this, NSIS has clear code points

Yes and that's what I'm saying. So NSIS is about end-to-end QoS
signaling. However, what you signal in your own domain may use
NSIS or something completely different.

> how each application will be able to signal its desired QOS in NSIS
> upper layer. The specific application sitting in both ends of the
> network will be able to communicate transparently for providing the
> end-to-end QOS. As I explained to Ronald last time, NSIS is not just
> e2e network layer signaling.

It's not? No you also can signal end-to-edge, edge-to-edge,
middle-to-end, end-to-end whatever...

> [RRR] If NSIS is NOT developing the e2e QOS between the application
> connected to the two ends of the network, there will probably no need
> for another protocol like NSIS QOS signaling because the traditional
> QOS signaling with RSVP/IntServ in the access network and DiffServ in
> the backbone network is doing just fine.

No it's not, that's why NSIS was initiated.
Probably you should go back and read RFCs 3726, 4094, and 3583.

> [RRR] After all these years of hard works when NSIS signaling is
> robust enough to provide the end-to-end QOS signaling, and QSPEC
> templates that are quite OK for the RSVP, IntServ, and DiffServ QOS
> models for interoperability, why you two folks want to turn the
> wheels back?

No, we don't want to turn wheels back. We want to move the work forward.

> [RRR] Can you two folks provide specific clarifications why you think
> that the so-called QOS models like Y.1541, 3GPP/Y.1451 and DSTE have
> the smae properties like those of RSVP, IntServ, and DiffServ? If
> not, why do you two folks just say the the obvious truth that these
> Y.1541, 3GPP/Y.1541, and DSTE are NOT be considered as QOS models?

Using Y.1541 is just fine. No problem.

> [RRR] If you two folks want to generalize mixing apples and oranges
> together, there will be no e2e QOS or no QOS interoperability, and
> NSIS QOS signaling will be of little for using for any practical
> purpose.

That's jsut the goal. We are trying to avoid ending up with
a non-interoperable e2e QoS signaling.

Roland (not Ronald) ;^)

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sun Nov 05 14:30:25 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgngG-0007f7-6D; Sun, 05 Nov 2006 14:29:20 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgngF-0007f2-1g
	for nsis@ietf.org; Sun, 05 Nov 2006 14:29:19 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgngB-0005Pk-BM
	for nsis@ietf.org; Sun, 05 Nov 2006 14:29:19 -0500
Received: from sbz-30.cs.helsinki.fi (sbz-30.cs.helsinki.fi [128.214.9.98])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Sun, 05 Nov 2006 21:29:12 +0200
	id 000AFF33.454E3B88.00002293
Date: Sun, 5 Nov 2006 21:29:11 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: nsis@ietf.org
Subject: Re: [NSIS] jerry, dave,	cornelia's proposal very confusing and not
	solving problem
In-Reply-To: <FC60B943-6D5D-4726-8C5C-ECC1B63738DE@cisco.com>
Message-ID: <Pine.LNX.4.64.0611052122560.11651@sbz-30.cs.Helsinki.FI>
References: <hvOvd46k.1162535221.3187200.karagian@ewi.utwente.nl>
	<Pine.LNX.4.64.0611031047160.7180@sbz-31.cs.Helsinki.FI>
	<FC60B943-6D5D-4726-8C5C-ECC1B63738DE@cisco.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d6b246023072368de71562c0ab503126
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi,

Just a note about the QSPEC discussion. We should make it clear what is a 
QSPEC issue, and what is an NSLP issue.

1. The lists of parameters, optional/mandatory, and how a QNE processes 
  them is a QSPEC issue.

2. Stacking is an NSLP feature.

Thus, it would be good, if the current discussions are structured 
(e.g. today at the meeting at 3pm) based on 

A) Issues with the current proposal for QSPEC

B) Issues with the current proposal for QoS NSLP

Jukka

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sun Nov 05 23:09:46 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Ggvmw-0002Wa-OH; Sun, 05 Nov 2006 23:08:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Ggvmv-0002WV-OW
	for nsis@ietf.org; Sun, 05 Nov 2006 23:08:45 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Ggvmt-0002fQ-FD
	for nsis@ietf.org; Sun, 05 Nov 2006 23:08:45 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Mon, 06 Nov 2006 06:08:42 +0200
	id 000AFD08.454EB54A.00006E99
Date: Mon, 6 Nov 2006 06:08:42 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: nsis@ietf.org
Message-ID: <Pine.LNX.4.64.0611060606020.28899@sbz-31.cs.Helsinki.FI>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 30ac594df0e66ffa5a93eb4c48bcb014
Subject: [NSIS] Question about " QSPEC Proc."
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi Jerry et al.,

I was reading the QSPEC spec with Roland, and we were both wondering what 
the " QSPEC Proc." field in the QSPEC Common Header is? Isn't this 
something that should be actually in the API between the NSLP and RMF, not 
encoded inside the QSPEC itself?

Jukka

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sun Nov 05 23:13:38 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Ggvr3-0005KY-8v; Sun, 05 Nov 2006 23:13:01 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Ggvr1-0005IO-It
	for nsis@ietf.org; Sun, 05 Nov 2006 23:12:59 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Ggvr0-0003B0-0a
	for nsis@ietf.org; Sun, 05 Nov 2006 23:12:59 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Ggvqn-000050-1O; Mon, 06 Nov 2006 05:12:45 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Ggvqm-0003ph-MH; Mon, 06 Nov 2006 05:12:44 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 44692B3B8;
	Mon,  6 Nov 2006 05:12:44 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Ggvqj-00083W-Bj; Mon, 06 Nov 2006 05:12:44 +0100
Message-ID: <454EB5E2.2010009@tm.uka.de>
Date: Mon, 06 Nov 2006 05:11:14 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.7) Gecko/20060909 Thunderbird/1.5.0.7 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 057ebe9b96adec30a7efb2aeda4c26a4
Cc: David R Oran <oran@cisco.com>, nsis@ietf.org, "Kappler,
	Cornelia" <cornelia.kappler@siemens.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry and all,

after todays discussion, please see my comments below.

Ash, Gerald R (Jerry), ALABS wrote:
> Source traffic Description:
>  TMOD-1 (mandatory to include)
>  TMOD-2 (optional to include)
> 
> Constraints (optional to include):
>  Path Latency
>  Path Jitter
>  Path PLR
>  Path PER
>  Y.1541 class (shorthand notation for above 4 parameters)
>  Slack term
>  Priority (Preemption, Defending, Admission, RPH Priority)
>  DSTE class
> 
> Handling directives (optional to include):
>  excess treatment
> 
> Traffic Classifiers (optional to include):
>  PHB Class (note that PHB class is set by downstream QNE to tell QNI how
> to mark traffic to ensure it will be treated as committed by admission
> control)
> Please review and comment on the proposed changes.

That's a step in the right direction.
I'd like to go a little bit further and propose the following
(after some more discussion with Jukka):
1. Remove even the QoS model specific stuff from the current QSPEC,
   e.g. Y.1541, DSTE, DiffServ classes/PHB-IDs, IntServ classes etc.

2. Call the new proposal "Generic QSPEC" or "E2E-QSPEC" which
   must be understood by any QoS model (that's why I call it generic).
   and expresses what end-to-end QoS requirements a QNI has.
   This is necessary in order to have a clear distinction from the
   QOSM-specific QSPECs as introduced next.

3. Defining QOSM-specific QSPECs if the E2E-QSPEC is not sufficient for
   admission control or must be treated in a special way.
   BUT: each QoS model must "preserve" the E2E-QSPEC and define how
   to map it to its own QoS model (e.g. which service class should be
   used in order to fulfill an E2E-QSPEC), and, if defined, how the
   QOSM-specific QSPEC is built.
   This QOSM-specific QSPEC would be created by ingress nodes and
   terminated by egress nodes of a domain.
   For example we will have an RMD-QSPEC to signal PHR specific stuff
   or a Y1541-QSPEC for signaling the chosen Y.1541 class and other
   specific parameters. So internal nodes may perform easier
   admission control by only looking at these parameters that were
   set by the edge node and ignoring what's in the E2E-QSPEC.

4. Remove QSPEC stacking from QoS-NSLP. We will have only
   one QSPEC object which is passed to the RMF. But due to the
   requirement that the E2E-QSPEC must be preserved it may be
   necessary for an edge node to "encapsulate" the E2E-QSPEC
   in the QOSM-specific QSPEC. So an RMD-QSPEC may simply include
   the E2E-QSPEC, but mainly for carrying to the egress node
   transparently, i.e. internal nodes will ignore the content of
   the E2E-QSPEC and only pay attention to the RMD-QSPEC-specific
   stuff (aggregate bandwith etc.). The edge nodes may interpret
   and modify the E2E-QSPEC as required (e.g. set QoS available etc.).
   The egress node will pass only the E2E-QSPEC in the RESERVE
   to the next domain. Furthermore we will have the possibility
   to introduce a type of local QSPECs (non-standardized) that
   do not have to be registered at IANA.
   A specific QSPEC may also carry more than the E2E-QSPEC, e.g.,
   as Dave mentioned some additional L2-QSPEC.
   I guess Jukka and I have identified a few advantages for
   QoS NSLP processing if we do it this way. QoS NSLP will
   basically only specify a QSPEC common header that contains an
   ID for the QSPEC-type.

This would boil down to the following:
- the E2E-QSPEC must be understood by every QoS model, thus
  it is always signaled inter-domain.
- the E2E-QSPEC may be remapped to some internal QoS model
  parameters at the edge of a domain. These parameters may
  be signaled internally via some QOSM-specific QSPEC.
  Semantics and content of the E2E-QSPEC must be preserved
  tough.
- The QoS NSLP implementation doesn't have to care about
  stacked QSPECs or which QSPEC object is actually relevant
  or of interest for the RMF, so we can easily support
  different QOSMs in a QNE providing different RMFs.
- QoS model specific stuff is removed from the current
  QSPEC template document and goes into the QoS model
  definition documents.
- The new approach allows to use and deploy new QoS models
  easily without changing the overall E2E-QSPEC definition.

Example for two different domains using their own QOSM-QSPEC
which will simply include QOSM-specific parameters as well
as the E2E-QSPEC.

                     RMD                        Y.1541
               +-------------+             +-------------+
    E2E-QSPEC /   RMD-QSPEC   \ E2E-QSPEC /  Y.1541-QSPEC \ E2E-QSPEC
QNI---------->QNE--->QNE--->QNE---------->QNE--->QNE--->QNE---------->QNR
              \               /           \               /
               +-------------+             +-------------+
Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sun Nov 05 23:46:32 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgwMi-0006v5-PP; Sun, 05 Nov 2006 23:45:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgwMh-0006uc-Lx
	for nsis@ietf.org; Sun, 05 Nov 2006 23:45:43 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgwMg-0006ue-4X
	for nsis@ietf.org; Sun, 05 Nov 2006 23:45:43 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Mon, 06 Nov 2006 06:45:41 +0200
	id 000AFCDA.454EBDF5.000056A7
Date: Mon, 6 Nov 2006 06:45:41 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: nsis@ietf.org
Subject: Re: [NSIS] Proposed Updates to the QSPEC Draft
In-Reply-To: <454EB5E2.2010009@tm.uka.de>
Message-ID: <Pine.LNX.4.64.0611060639470.28899@sbz-31.cs.Helsinki.FI>
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F0F@KCCLUST06EVS1.ugd.att.com>
	<454EB5E2.2010009@tm.uka.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 67c1ea29f88502ef6a32ccec927970f0
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


After long discussions with Roland, I see the benefits in Roland's ideas, 
and support this proposal.

Jukka

On Mon, 6 Nov 2006, Roland Bless wrote:

> Hi Jerry and all,
> 
> after todays discussion, please see my comments below.
> 
> Ash, Gerald R (Jerry), ALABS wrote:
> > Source traffic Description:
> >  TMOD-1 (mandatory to include)
> >  TMOD-2 (optional to include)
> > 
> > Constraints (optional to include):
> >  Path Latency
> >  Path Jitter
> >  Path PLR
> >  Path PER
> >  Y.1541 class (shorthand notation for above 4 parameters)
> >  Slack term
> >  Priority (Preemption, Defending, Admission, RPH Priority)
> >  DSTE class
> > 
> > Handling directives (optional to include):
> >  excess treatment
> > 
> > Traffic Classifiers (optional to include):
> >  PHB Class (note that PHB class is set by downstream QNE to tell QNI how
> > to mark traffic to ensure it will be treated as committed by admission
> > control)
> > Please review and comment on the proposed changes.
> 
> That's a step in the right direction.
> I'd like to go a little bit further and propose the following
> (after some more discussion with Jukka):
> 1. Remove even the QoS model specific stuff from the current QSPEC,
>    e.g. Y.1541, DSTE, DiffServ classes/PHB-IDs, IntServ classes etc.
> 
> 2. Call the new proposal "Generic QSPEC" or "E2E-QSPEC" which
>    must be understood by any QoS model (that's why I call it generic).
>    and expresses what end-to-end QoS requirements a QNI has.
>    This is necessary in order to have a clear distinction from the
>    QOSM-specific QSPECs as introduced next.
> 
> 3. Defining QOSM-specific QSPECs if the E2E-QSPEC is not sufficient for
>    admission control or must be treated in a special way.
>    BUT: each QoS model must "preserve" the E2E-QSPEC and define how
>    to map it to its own QoS model (e.g. which service class should be
>    used in order to fulfill an E2E-QSPEC), and, if defined, how the
>    QOSM-specific QSPEC is built.
>    This QOSM-specific QSPEC would be created by ingress nodes and
>    terminated by egress nodes of a domain.
>    For example we will have an RMD-QSPEC to signal PHR specific stuff
>    or a Y1541-QSPEC for signaling the chosen Y.1541 class and other
>    specific parameters. So internal nodes may perform easier
>    admission control by only looking at these parameters that were
>    set by the edge node and ignoring what's in the E2E-QSPEC.
> 
> 4. Remove QSPEC stacking from QoS-NSLP. We will have only
>    one QSPEC object which is passed to the RMF. But due to the
>    requirement that the E2E-QSPEC must be preserved it may be
>    necessary for an edge node to "encapsulate" the E2E-QSPEC
>    in the QOSM-specific QSPEC. So an RMD-QSPEC may simply include
>    the E2E-QSPEC, but mainly for carrying to the egress node
>    transparently, i.e. internal nodes will ignore the content of
>    the E2E-QSPEC and only pay attention to the RMD-QSPEC-specific
>    stuff (aggregate bandwith etc.). The edge nodes may interpret
>    and modify the E2E-QSPEC as required (e.g. set QoS available etc.).
>    The egress node will pass only the E2E-QSPEC in the RESERVE
>    to the next domain. Furthermore we will have the possibility
>    to introduce a type of local QSPECs (non-standardized) that
>    do not have to be registered at IANA.
>    A specific QSPEC may also carry more than the E2E-QSPEC, e.g.,
>    as Dave mentioned some additional L2-QSPEC.
>    I guess Jukka and I have identified a few advantages for
>    QoS NSLP processing if we do it this way. QoS NSLP will
>    basically only specify a QSPEC common header that contains an
>    ID for the QSPEC-type.
> 
> This would boil down to the following:
> - the E2E-QSPEC must be understood by every QoS model, thus
>   it is always signaled inter-domain.
> - the E2E-QSPEC may be remapped to some internal QoS model
>   parameters at the edge of a domain. These parameters may
>   be signaled internally via some QOSM-specific QSPEC.
>   Semantics and content of the E2E-QSPEC must be preserved
>   tough.
> - The QoS NSLP implementation doesn't have to care about
>   stacked QSPECs or which QSPEC object is actually relevant
>   or of interest for the RMF, so we can easily support
>   different QOSMs in a QNE providing different RMFs.
> - QoS model specific stuff is removed from the current
>   QSPEC template document and goes into the QoS model
>   definition documents.
> - The new approach allows to use and deploy new QoS models
>   easily without changing the overall E2E-QSPEC definition.
> 
> Example for two different domains using their own QOSM-QSPEC
> which will simply include QOSM-specific parameters as well
> as the E2E-QSPEC.
> 
>                      RMD                        Y.1541
>                +-------------+             +-------------+
>     E2E-QSPEC /   RMD-QSPEC   \ E2E-QSPEC /  Y.1541-QSPEC \ E2E-QSPEC
> QNI---------->QNE--->QNE--->QNE---------->QNE--->QNE--->QNE---------->QNR
>               \               /           \               /
>                +-------------+             +-------------+
> Regards,
>  Roland
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 06 00:17:10 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgwqJ-0008Ar-IH; Mon, 06 Nov 2006 00:16:19 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgwqI-00089o-1P
	for nsis@ietf.org; Mon, 06 Nov 2006 00:16:18 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgwqF-0005FY-B7
	for nsis@ietf.org; Mon, 06 Nov 2006 00:16:18 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GgwqD-0000v3-4U; Mon, 06 Nov 2006 06:16:14 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GgwqC-00042l-Rl; Mon, 06 Nov 2006 06:16:12 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 5976FB3B6;
	Mon,  6 Nov 2006 06:16:12 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GgwqB-00085c-4o; Mon, 06 Nov 2006 06:16:12 +0100
Message-ID: <454EC4BA.7020006@tm.uka.de>
Date: Mon, 06 Nov 2006 06:14:34 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.7) Gecko/20060909 Thunderbird/1.5.0.7 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: "Hancock, Robert" <robert.hancock@roke.co.uk>
References: <A632AD91CF90F24A87C42F6B96ADE5C50157A97C@rsys005a.comm.ad.roke.co.uk>
In-Reply-To: <A632AD91CF90F24A87C42F6B96ADE5C50157A97C@rsys005a.comm.ad.roke.co.uk>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 082a9cbf4d599f360ac7f815372a6a15
Cc: nsis@ietf.org
Subject: [NSIS] Re: MA-Hello request/response processing (was: RE: GIST MA
	Keep-alive/Liveliness)
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hancock, Robert schrieb:
> Hi roland,
> 
> Some good questions. I think the facility is useful (see below)
> and so we need to keep it, but a small protocol change is 

I'm not saying that it's not useful. I just wanted a little
more rationale for it.

> Basically I think that since it is up to the originating node
> to decide whether to make such a liveness test, we don't need
> to be totally prescriptive about how to go about it. After all,

That's ok.

> a conformant node can just decide never to do such a test in
> the first place. What I think does need to be stated is that
> you can use the MA-Hello to test liveness and Die if the test
> fails. Text is added in 6.4 to say just that.

Yep. But the proposed text just mentions a mismatching
Hello-ID as failure. I guess that it's more likely that
the testing node wants to set also a timeout. The value
of the latter may be a local decision.

> This is the crucial point. I've proposed adding a Hello-ID
> to the MA-Hello message. It is set to a non-zero value in
> request (R=1) messages, and echoed in any reply; otherwise
> it is set to zero. Note that replies must have R=0. No-one
> ever has to generate R=1 messages, but if you get one you 
> have to send a response with R=0 including the Hello-ID.
> (I think this meets the matching requirement safely.)

Ok. So receiving an MA-Hello with R=1 but Hello-ID=0
is considered to be a protocol error (object value error?).

> I think of it as a diagnostic. You don't have to invoke it,
> but it is useful to define normatively how peers will respond
> to it. At the moment, it's a simple way to check that the

That's fine.

> GIST layer at the peer is still functioning without having
> to send some made-up data (which might interact with an 
> NSLP). It might even be used mainly by user tools (a GIST
> ping). It can also be used to test the whether outstanding 
> messages have been delivered to the other end of an MA, which
> can be useful in some circumstances.

I'd like to see this or similar text in the spec as rationale.

> The currently proposed text is in the issue tracker at
> http://nsis.srmr.co.uk/cgi-bin/roundup.cgi/nsis-ntlp-issues/issue157

Ok for me if you also include the explanatory text such as:
The request/response exchange is essentially a diagnostic, which a node
may invoke whenever it wants (including never). It is up to the node to
decide when to invoke it and how, and what to do with the result.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 06 02:48:57 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GgzDC-0005GL-EE; Mon, 06 Nov 2006 02:48:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GgzDB-0005G7-1q
	for nsis@ietf.org; Mon, 06 Nov 2006 02:48:05 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GgzD8-0007Bn-Fm
	for nsis@ietf.org; Mon, 06 Nov 2006 02:48:04 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GgzD6-0000VG-LM; Mon, 06 Nov 2006 08:48:01 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GgzD6-0005jZ-BI; Mon, 06 Nov 2006 08:48:00 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 4B8C5B4EF;
	Mon,  6 Nov 2006 08:47:59 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GgzD3-0008G0-Na; Mon, 06 Nov 2006 08:47:58 +0100
Message-ID: <454EE830.7040209@tm.uka.de>
Date: Mon, 06 Nov 2006 08:45:52 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.7) Gecko/20060909 Thunderbird/1.5.0.7 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: "Hancock, Robert" <robert.hancock@roke.co.uk>
Subject: Re: [NSIS] GIST issue on MA-Hold-Time
References: <A632AD91CF90F24A87C42F6B96ADE5C50157A97E@rsys005a.comm.ad.roke.co.uk>
In-Reply-To: <A632AD91CF90F24A87C42F6B96ADE5C50157A97E@rsys005a.comm.ad.roke.co.uk>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 82c9bddb247d9ba4471160a9a865a5f3
Cc: nsis <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Robert,

Hancock, Robert wrote:
> I am not sure how to do this in an interoperable way without 
> quite a bit of extra complexity (some more message rounds). I'd
> prefer just to emphasise that it's silly to use small values.
> Any text suggestions?

So you suppose that rejecting too small values would not
be a good choice? For instance, if a node sends an MA-Hold-Time
of 20 (ms), this node will usually get the traffic, so it
cannot be misused as an attack. However, if we allow to reject an
MA-Hold-Time value, we probably must add a mechanism that
doubles the MA-Hold-Time. Did you mean this with "some more message
rounds"? I'm not convinced that it makes sense.

>> * The MA-Hold-Time must be repeated in the first Confirm message
>>   due to the need for having this value available even if late
>>   state installation is used.
>>   => What should be done if the MA-Hold-Time is missing?
>>   Sending back an Object-Type error with mandatory object missing is
>>   probably the right thing to do.
> 
> Done.
> 
>>   Furthermore, the specification in 5.1 of the Confirm message
>>   is confusing in this respect and should be clarified.
>>        Confirm = Common-Header
>>                  Message-Routing-Information
>>                  Session-Identification
>>                  Network-Layer-Information
>>                  [ Responder-Cookie
>>                   [ Stack-Proposal
>>                      [ Stack-Configuration-Data ] ] ]
>>                  [ NSLP-Data ]
>>   While the example in fig.4 sheds more lighty on how it is
>>   intended, one would think that Stack-Proposal _and_ SCD
>>   of the Responder are repeated in the Confirm (so I guess
>>   nesting with brackets must be fixed). Therefore,
>>   clarification is needed that only the R-Node SP is repeated
>>   (w/o R-Node SCD)
>>   and the Q-Node SCD with MA-Hold-Time is mandatory in the
>>   very first Confirm. 
> 
> I think that the ABNF is correct here; the real problem is that
> ABNF is not a good syntax for handling 'conditionally mandatory'
> information elements. More discussion is in the issue tracker.
> Also, I don't really want to qualify which SP is included 
> (since this makes the ABNF harder to validate). Hopefully the
> new text before the ABNF in 5.1 makes this clearer.

Ok.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 06 05:21:45 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gh1b7-0007Rx-FN; Mon, 06 Nov 2006 05:20:57 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gh1b5-0007QB-1R
	for nsis@ietf.org; Mon, 06 Nov 2006 05:20:55 -0500
Received: from rsys002x.roke.co.uk ([193.118.201.109])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gh1TD-0005Un-Et
	for nsis@ietf.org; Mon, 06 Nov 2006 05:12:52 -0500
Received: from rsys005a.comm.ad.roke.co.uk (rsys005a [193.118.193.85])
	by rsys002x.roke.co.uk (8.13.1/8.13.1) with ESMTP id kA6ACafr016056;
	Mon, 6 Nov 2006 10:12:36 GMT
Received: from victor ([193.118.192.66]) by rsys005a.comm.ad.roke.co.uk with
	Microsoft SMTPSVC(6.0.3790.1830); Mon, 6 Nov 2006 10:12:35 +0000
From: "Robert Hancock" <robert.hancock@roke.co.uk>
To: "'Roland Bless'" <bless@tm.uka.de>
Subject: RE: [NSIS] GIST issue on MA-Hold-Time
Date: Mon, 6 Nov 2006 02:12:54 -0800
Message-ID: <00de01c7018c$213a0c10$0ffa12ac@comm.ad.roke.co.uk>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook, Build 10.0.6626
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Importance: Normal
In-Reply-To: <454EE830.7040209@tm.uka.de>
X-OriginalArrivalTime: 06 Nov 2006 10:12:35.0970 (UTC)
	FILETIME=[14D1EE20:01C7018C]
X-MailScanner-roke-co-uk: Found to be clean
X-MailScanner-roke-co-uk-SpamCheck: 
X-MailScanner-From: robert.hancock@roke.co.uk
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d185fa790257f526fedfd5d01ed9c976
Cc: 'nsis' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

hi roland,

> Hi Robert,
> 
> Hancock, Robert wrote:
> > I am not sure how to do this in an interoperable way without
> > quite a bit of extra complexity (some more message rounds). I'd
> > prefer just to emphasise that it's silly to use small values.
> > Any text suggestions?
> 
> So you suppose that rejecting too small values would not
> be a good choice? 

I did think of allowing a node (MAY) to send back Object Value Error
and drop the MA if the Hold-Time is 'too small'. I suppose that this
would then put the obligation on the initiator to have another go
with a sensible value, without putting a lot more complexity in the
protocol spec. I'd still like to put a bound on when this was
allowed - say for times < 1 second?

> For instance, if a node sends an 
> MA-Hold-Time of 20 (ms), this node will usually get the 
> traffic, so it cannot be misused as an attack. However, if we 
> allow to reject an MA-Hold-Time value, we probably must add a 
> mechanism that doubles the MA-Hold-Time. Did you mean this 
> with "some more message rounds"? I'm not convinced that it 
> makes sense.
> 
> >> * The MA-Hold-Time must be repeated in the first Confirm message
> >>   due to the need for having this value available even if late
> >>   state installation is used.
> >>   => What should be done if the MA-Hold-Time is missing?
> >>   Sending back an Object-Type error with mandatory object 
> missing is
> >>   probably the right thing to do.
> > 
> > Done.
> > 
> >>   Furthermore, the specification in 5.1 of the Confirm message
> >>   is confusing in this respect and should be clarified.
> >>        Confirm = Common-Header
> >>                  Message-Routing-Information
> >>                  Session-Identification
> >>                  Network-Layer-Information
> >>                  [ Responder-Cookie
> >>                   [ Stack-Proposal
> >>                      [ Stack-Configuration-Data ] ] ]
> >>                  [ NSLP-Data ]
> >>   While the example in fig.4 sheds more lighty on how it is
> >>   intended, one would think that Stack-Proposal _and_ SCD
> >>   of the Responder are repeated in the Confirm (so I guess
> >>   nesting with brackets must be fixed). Therefore,
> >>   clarification is needed that only the R-Node SP is repeated
> >>   (w/o R-Node SCD)
> >>   and the Q-Node SCD with MA-Hold-Time is mandatory in the
> >>   very first Confirm.
> > 
> > I think that the ABNF is correct here; the real problem is 
> that ABNF 
> > is not a good syntax for handling 'conditionally mandatory' 
> > information elements. More discussion is in the issue 
> tracker. Also, I 
> > don't really want to qualify which SP is included (since this makes 
> > the ABNF harder to validate). Hopefully the new text before 
> the ABNF 
> > in 5.1 makes this clearer.
> 
> Ok.

cool.

cheers,

r.

> 
> Regards,
>  Roland
> 
> 


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 06 12:31:44 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gh8JS-0000hU-Us; Mon, 06 Nov 2006 12:31:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gh8JR-0000hL-Lj
	for nsis@ietf.org; Mon, 06 Nov 2006 12:31:09 -0500
Received: from rsys002x.roke.co.uk ([193.118.201.109])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gh8JF-0007Vn-Q9
	for nsis@ietf.org; Mon, 06 Nov 2006 12:31:09 -0500
Received: from rsys005a.comm.ad.roke.co.uk (rsys005a [193.118.193.85])
	by rsys002x.roke.co.uk (8.13.1/8.13.1) with ESMTP id kA6HUoLj015350;
	Mon, 6 Nov 2006 17:30:51 GMT
Received: from victor ([193.118.192.66]) by rsys005a.comm.ad.roke.co.uk with
	Microsoft SMTPSVC(6.0.3790.1830); Mon, 6 Nov 2006 17:30:49 +0000
From: "Robert Hancock" <robert.hancock@roke.co.uk>
To: "'Roland Bless'" <bless@tm.uka.de>
Date: Mon, 6 Nov 2006 09:31:08 -0800
Message-ID: <002001c701c9$59c33c90$8b448182@comm.ad.roke.co.uk>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook, Build 10.0.6626
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <454EC4BA.7020006@tm.uka.de>
Importance: Normal
X-OriginalArrivalTime: 06 Nov 2006 17:30:50.0016 (UTC)
	FILETIME=[4D4AF200:01C701C9]
X-MailScanner-roke-co-uk: Found to be clean
X-MailScanner-roke-co-uk-SpamCheck: 
X-MailScanner-From: robert.hancock@roke.co.uk
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 32b73d73e8047ed17386f9799119ce43
Cc: nsis@ietf.org
Subject: [NSIS] RE: MA-Hello request/response processing (was: RE: GIST MA
	Keep-alive/Liveliness)
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland,

thanks for the inputs. I've made some changes to the proposed text;
please check out 
http://nsis.srmr.co.uk/cgi-bin/roundup.cgi/nsis-ntlp-issues/issue157
and see what you think.

cheers,

robert h.

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: 05 November 2006 21:15
> To: Hancock, Robert
> Cc: nsis@ietf.org
> Subject: Re: MA-Hello request/response processing (was: RE: 
> GIST MA Keep-alive/Liveliness)
> 
> 
> Hancock, Robert schrieb:
> > Hi roland,
> > 
> > Some good questions. I think the facility is useful (see 
> below) and so 
> > we need to keep it, but a small protocol change is
> 
> I'm not saying that it's not useful. I just wanted a little 
> more rationale for it.
> 
> > Basically I think that since it is up to the originating node to 
> > decide whether to make such a liveness test, we don't need to be 
> > totally prescriptive about how to go about it. After all,
> 
> That's ok.
> 
> > a conformant node can just decide never to do such a test 
> in the first 
> > place. What I think does need to be stated is that you can use the 
> > MA-Hello to test liveness and Die if the test fails. Text 
> is added in 
> > 6.4 to say just that.
> 
> Yep. But the proposed text just mentions a mismatching
> Hello-ID as failure. I guess that it's more likely that
> the testing node wants to set also a timeout. The value
> of the latter may be a local decision.
> 
> > This is the crucial point. I've proposed adding a Hello-ID
> > to the MA-Hello message. It is set to a non-zero value in request 
> > (R=1) messages, and echoed in any reply; otherwise it is 
> set to zero. 
> > Note that replies must have R=0. No-one ever has to generate R=1 
> > messages, but if you get one you have to send a response with R=0 
> > including the Hello-ID. (I think this meets the matching 
> requirement 
> > safely.)
> 
> Ok. So receiving an MA-Hello with R=1 but Hello-ID=0
> is considered to be a protocol error (object value error?).
> 
> > I think of it as a diagnostic. You don't have to invoke it, 
> but it is 
> > useful to define normatively how peers will respond to it. At the 
> > moment, it's a simple way to check that the
> 
> That's fine.
> 
> > GIST layer at the peer is still functioning without having
> > to send some made-up data (which might interact with an
> > NSLP). It might even be used mainly by user tools (a GIST
> > ping). It can also be used to test the whether outstanding 
> > messages have been delivered to the other end of an MA, which
> > can be useful in some circumstances.
> 
> I'd like to see this or similar text in the spec as rationale.
> 
> > The currently proposed text is in the issue tracker at 
> > http://nsis.srmr.co.uk/cgi-bin/roundup.cgi/nsis-ntlp-issues/issue157
> 
> Ok for me if you also include the explanatory text such as:
> The request/response exchange is essentially a diagnostic, 
> which a node may invoke whenever it wants (including never). 
> It is up to the node to decide when to invoke it and how, and 
> what to do with the result.
> 
> Regards,
>  Roland
> 
> 


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 07 11:22:59 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhThx-0001q8-IY; Tue, 07 Nov 2006 11:21:53 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhThx-0001pl-0O
	for nsis@ietf.org; Tue, 07 Nov 2006 11:21:53 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GhThr-00059Q-H2
	for nsis@ietf.org; Tue, 07 Nov 2006 11:21:52 -0500
Received: from webmail.cs.utwente.nl (janus.ewi.utwente.nl [130.89.10.26])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with SMTP id kA7GLTH0022646
	for <nsis@ietf.org>; Tue, 7 Nov 2006 17:21:33 +0100 (MET)
Received: from 130.129.32.143 (auth. user karagian@imap2.cs.utwente.nl)
	by webmail.cs.utwente.nl with HTTP; Tue, 07 Nov 2006 16:21:28 +0000
To: nsis@ietf.org
Date: Tue, 07 Nov 2006 16:21:27 +0000
X-Mailer: IlohaMail/0.8.13 (On: webmail.cs.utwente.nl)
Message-ID: <zEW2gRVz.1162916487.9969660.karagian@ewi.utwente.nl>
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
Bounce-To: "Georgios Karagiannis" <karagian@cs.utwente.nl>
MIME-Version: 1.0 
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: 0 () 
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Tue, 07 Nov 2006 17:21:34 +0100 (MET)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 67c1ea29f88502ef6a32ccec927970f0
Subject: [NSIS] conensus of sundays meeting and rolands proposal
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all

I thought that during the meeting of Sunday we reached the following
three points of consensus, on the following:

Consensus_1):
Use the QSpec template draft as is but change the following:

The QSPEC1/QSPEC2 sematincs are replaced by Daves proposal for traffic
source description, where a generic (end-to-end) QSPEC has to be
interpreted by QNEs.
However, the additions/modifcations to this proposal are related to the
fact that a consensus was reached where the local/private QOS models can
use stacking and that they can map/interpret the traffic source
description into their own local/private QSPEC and ignore the end-to-end
QSPEC.
The mapping/interpreatation procedure will be specified in each QoS model.
In this consnsus we also agreed that it should be possible that more than
two QSPECs should be able to be satcked, where at least one oif them is
the end-to-end QSPEC.

The way of how the stacked objects are processed should then be decribed
in the QOS-NSLP draft.

The consequence of the above on the QOS-NSLP, QSPEC tempalet draft and
QOS models is the following:

* The QSPEC template draft will not use anymore the notion of
QSPC1/QSPEc2 objects and instead use the traffic source description
proposed by Dave Oran.

* All the other QSPEC parameers  that currently are described in the
QSPEC template draft have to be considered as optional to implement by
the QNEs.

* The R and Q fags are eliminated since remapping is not allowed.

* The mapping/interpreattion of the traffic source description is
described in the QoS models and not in the QSPEC template draft.

* The QSPEC template draft has to decribe how a QOS model can stack the
QSPEC objects, which is a similar description that we have now in the
QSPEC templated draft. It has to empasize how the stacked QSPEc objest
are interpreted by the QOS models.

* Further the features/procedures/etc. described in the QSPEC template
draft remain the same.


Consensus_2)
The QoS-NSLP remains the same as is now with the difference that more
than two QSPEC objects can be stacked. Note that only one is a real
QSPEC object, i.e., end-to-end QSPEC and the rest are private/local
QSPEC objects meaning that the two types of QSPEc objects will have tio
use different object IDs.

Consensus_3)
The goal of the QOS model, from the point of view of what it can specify
remains the same as now, but the QOS model will not use remapping,
however the QoS model has to describe how the end-to-end QSPEC (traffic
source description) has to be interpreted/mapped by the QNEs that are
supporting the QOS model.

This means that the QoS model can specify how the traffic source
description can be mapped into a local/private QSPEC. By using e.g.,
stacaking the end-to-end QSPEC can be then ignored.

------------Rolands proposal:

Now Roland has sent a new proposal that mainly agrees with the three
points of reached consensus, but it has two differences:

Difference_1:
The stacking concept is supported but the implementation of how the
stacking is specified and implemented is different.
Only one QSPEc is used. Thus any QoS model interprets only one QSPEC
object, which it can be either the end-to-end QSPEC or a local/private
QSPEC.

The local/private QSPEC can provide the concept of stacking by e.g.,
encapsulating the end-to-end QSPEC and providing means/ways to ignore
the end-to-end QSPEC. I suppose that if multiple levels are stacking are
used then a QOS model could ignore all the paramteres that might be
stacked by another QoS model (which is not the end-to-end QoS model) but
which arer not relevant for the local/private QoS model.
Then method/way of how teh end-to-end QSPEC has to be interpreted/mapped
has to be described in the QOS model specification.
In this way stacking is supproted, but it will not be anymore specified
in the QOS-NSLP, but in the QSPEC template draft and the QoS models.
Note that Roland does not mention in his proposal that the stacking
procedure has to be specified in the QSPEC template draft, but in my
opinuion this is necessary.

Difference_2:
Roland proposes to remove the description of all QoS model related
parameters from the QSPEC templated  draft. QoS model specific
parameters care the Y.1541, DSTE, PHB classes, etc..

I have a question for Roland:
Is my explanation of the three points of consensus, i.e., Consensus_1,
Consensus_2 and Consensus_3 in agreement with wath we have decided on
Sunday?

Do you agree that my explantion related to your proposal and to the 2
differences between the three points of consensus and your proposal is
the right one?

If yes, I can agree with Difference_1 only if we make sure that the
implementation of the stacking concept keeps its semantics. This also
means that the specification of how stacking is accomplished can be
removed from the QOS-NSLP draft, but it has to be specified/explained in
the QSPEC template draft and in the QOS models.

Regarding Difference_2 I can only agree if the editors of the QSEPC
template draft agree with the removal of all QoS model specific QSPEC
parameter descriptions from the QSPEC template draft.

Best Regards,
Georgios

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 07 12:17:23 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhUYc-0005cY-Q4; Tue, 07 Nov 2006 12:16:18 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhUYa-0005X3-Ju; Tue, 07 Nov 2006 12:16:16 -0500
Received: from smtp2.dei.uc.pt ([193.137.203.231] helo=smtp.dei.uc.pt)
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1GhUYU-0005pH-To; Tue, 07 Nov 2006 12:16:16 -0500
Received: from zoing (dhcp70-95.ietf67.org [130.129.70.95])
	(authenticated bits=0)
	by smtp.dei.uc.pt (8.13.8/8.13.8) with ESMTP id kA7HF2Rx010125
	(version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO);
	Tue, 7 Nov 2006 17:15:17 GMT
Message-Id: <200611071715.kA7HF2Rx010125@smtp.dei.uc.pt>
From: =?iso-8859-1?Q?Lu=EDs_Cordeiro?= <cordeiro@dei.uc.pt>
To: <nsis-imp@ietf.org>, <nsis@ietf.org>
Date: Tue, 7 Nov 2006 17:15:00 -0000
MIME-Version: 1.0
X-Mailer: Microsoft Office Outlook, Build 11.0.5510
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: AccCkEHhKhagnea7RymV4KQ8aRxvcQ==
X-FCTUC-DEI-SIC-MailScanner-Information: Please contact helpdesk@dei.uc.pt for
	more information
X-FCTUC-DEI-SIC-MailScanner: Found to be clean
X-FCTUC-DEI-SIC-MailScanner-SpamCheck: not spam (whitelisted),
	SpamAssassin (score=-4.398, required 3, autolearn=not spam,
	ALL_TRUSTED -1.80, BAYES_00 -2.60, HTML_MESSAGE 0.00)
X-FCTUC-DEI-SIC-MailScanner-From: cordeiro@dei.uc.pt
X-Spam-Status: No
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a8041eca2a724d631b098c15e9048ce9
Cc: 
Subject: [NSIS] University of Coimbra NSIS implementation release
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0130093865=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============0130093865==
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_01C9_01C70290.4E577890"

This is a multi-part message in MIME format.

------=_NextPart_000_01C9_01C70290.4E577890
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hi all,

=20

The new versions of the University of Coimbra JAVA NSIS implementation =
are
now available in our site: http://nsis.dei.uc.pt =
<http://nsis.dei.uc.pt/>=20

=20

You can find the implementation of the following draft versions:

            - GIST v11

            - QoS NSLP v11

            - NATFW NSLP v12

=20

As soon as possible, a NSLP framework will be available to ease the
implementation of new NSLPs using our GIST implementation and GIST/NSLP
interface.

=20

If you need any information or help about these implementations please
contact me.

=20

Best Regards,

Lu=EDs Cordeiro

=20


------=_NextPart_000_01C9_01C70290.4E577890
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<html xmlns:v=3D"urn:schemas-microsoft-com:vml" =
xmlns:o=3D"urn:schemas-microsoft-com:office:office" =
xmlns:w=3D"urn:schemas-microsoft-com:office:word" =
xmlns:st1=3D"urn:schemas-microsoft-com:office:smarttags" =
xmlns=3D"http://www.w3.org/TR/REC-html40">

<head>
<meta http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<meta name=3DGenerator content=3D"Microsoft Word 11 (filtered medium)">
<o:SmartTagType =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
 name=3D"PlaceName"/>
<o:SmartTagType =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
 name=3D"PlaceType"/>
<o:SmartTagType =
namespaceuri=3D"urn:schemas-microsoft-com:office:smarttags"
 name=3D"place"/>
<!--[if !mso]>
<style>
st1\:*{behavior:url(#default#ieooui) }
</style>
<![endif]-->
<style>
<!--
 /* Style Definitions */
 p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0cm;
	margin-bottom:.0001pt;
	font-size:12.0pt;
	font-family:"Times New Roman";}
a:link, span.MsoHyperlink
	{color:blue;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{color:purple;
	text-decoration:underline;}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:Arial;
	color:windowtext;}
@page Section1
	{size:595.3pt 841.9pt;
	margin:70.85pt 3.0cm 70.85pt 3.0cm;}
div.Section1
	{page:Section1;}
-->
</style>

</head>

<body lang=3DPT link=3Dblue vlink=3Dpurple>

<div class=3DSection1>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Hi all,<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>The new versions of the <st1:place =
w:st=3D"on"><st1:PlaceType
 w:st=3D"on">University</st1:PlaceType> of <st1:PlaceName =
w:st=3D"on">Coimbra JAVA
  NSIS</st1:PlaceName></st1:place> implementation are now available in =
our
site: <a href=3D"http://nsis.dei.uc.pt/" =
title=3D"blocked::http://nsis.dei.uc.pt/">http://nsis.dei.uc.pt</a><o:p><=
/o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>You can find the implementation of the =
following draft
versions:<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;
- GIST v11<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;
- QoS NSLP v11<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp=
;&nbsp;&nbsp;&nbsp;
- NATFW NSLP v12<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>As soon as possible, a NSLP framework will be
available to ease the implementation of new NSLPs using our GIST =
implementation
and GIST/NSLP interface.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>If you need any information or help about =
these
implementations please contact me.<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Best Regards,<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'>Lu=EDs Cordeiro<o:p></o:p></span></font></p>

<p class=3DMsoNormal><font size=3D2 face=3DArial><span lang=3DEN-GB =
style=3D'font-size:
10.0pt;font-family:Arial'><o:p>&nbsp;</o:p></span></font></p>

</div>

</body>

</html>

------=_NextPart_000_01C9_01C70290.4E577890--



--===============0130093865==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============0130093865==--





From nsis-bounces@ietf.org Tue Nov 07 12:28:30 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhUjS-0005uk-6l; Tue, 07 Nov 2006 12:27:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhUjQ-0005or-PG
	for nsis@ietf.org; Tue, 07 Nov 2006 12:27:28 -0500
Received: from mclmx2.mail.saic.com ([149.8.64.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GhUjO-00081j-52
	for nsis@ietf.org; Tue, 07 Nov 2006 12:27:28 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx2.mail.saic.com; Tue, 7 Nov 2006 12:27:16 -0500
Received: from 0015-ITS-EXBH01.us.saic.com ([10.43.229.18])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006110712271520647 ; Tue, 07 Nov 2006 12:27:16 -0500
Received: from 0591-ITS-EXMP02.us.saic.com ([10.42.81.248]) by
	0015-ITS-EXBH01.us.saic.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Tue, 7 Nov 2006 12:27:15 -0500
x-mimeole: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] conensus of sundays meeting and rolands proposal
Date: Tue, 7 Nov 2006 12:27:15 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF07@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] conensus of sundays meeting and rolands proposal
Thread-Index: AccCiUHYWcZ98kxJQ1Gvs07M1jsZUgABzCdm
References: <zEW2gRVz.1162916487.9969660.karagian@ewi.utwente.nl>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: "Georgios Karagiannis" <karagian@cs.utwente.nl>,
    <nsis@ietf.org>
X-OriginalArrivalTime: 07 Nov 2006 17:27:15.0731 (UTC)
	FILETIME=[F7FB6630:01C70291]
X-Spam-Score: 0.3 (/)
X-Scan-Signature: 40161b1d86420e0807d771943d981d25
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1153522673=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1153522673==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C70291.F7B3E02E"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C70291.F7B3E02E
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hi, Georgios and all:
=20
It is MANDATORY, as Ronald and Jukka, "to remove all QoS model specific =
QSPEC parameter descriptions from the QSPEC template draft." There must =
NOT be any "if" and "when."=20
=20
Before going any step further in order to make progress related to this =
"generic" standard-track RFC draft, we have to take this first mandatory =
step.=20
=20
Otherwise, we go back to the circle of arguments and conter-arguments of =
interworking.=20
=20
Best regards,
Radhika
=20
________________________________

From: nsis-bounces@ietf.org on behalf of Georgios Karagiannis
Sent: Tue 11/7/2006 11:21 AM
To: nsis@ietf.org
Subject: [NSIS] conensus of sundays meeting and rolands proposal



Hi all

I thought that during the meeting of Sunday we reached the following
three points of consensus, on the following:

Consensus_1):
Use the QSpec template draft as is but change the following:

The QSPEC1/QSPEC2 sematincs are replaced by Daves proposal for traffic
source description, where a generic (end-to-end) QSPEC has to be
interpreted by QNEs.
However, the additions/modifcations to this proposal are related to the
fact that a consensus was reached where the local/private QOS models can
use stacking and that they can map/interpret the traffic source
description into their own local/private QSPEC and ignore the end-to-end
QSPEC.
The mapping/interpreatation procedure will be specified in each QoS =
model.
In this consnsus we also agreed that it should be possible that more =
than
two QSPECs should be able to be satcked, where at least one oif them is
the end-to-end QSPEC.

The way of how the stacked objects are processed should then be decribed
in the QOS-NSLP draft.

The consequence of the above on the QOS-NSLP, QSPEC tempalet draft and
QOS models is the following:

* The QSPEC template draft will not use anymore the notion of
QSPC1/QSPEc2 objects and instead use the traffic source description
proposed by Dave Oran.

* All the other QSPEC parameers  that currently are described in the
QSPEC template draft have to be considered as optional to implement by
the QNEs.

* The R and Q fags are eliminated since remapping is not allowed.

* The mapping/interpreattion of the traffic source description is
described in the QoS models and not in the QSPEC template draft.

* The QSPEC template draft has to decribe how a QOS model can stack the
QSPEC objects, which is a similar description that we have now in the
QSPEC templated draft. It has to empasize how the stacked QSPEc objest
are interpreted by the QOS models.

* Further the features/procedures/etc. described in the QSPEC template
draft remain the same.


Consensus_2)
The QoS-NSLP remains the same as is now with the difference that more
than two QSPEC objects can be stacked. Note that only one is a real
QSPEC object, i.e., end-to-end QSPEC and the rest are private/local
QSPEC objects meaning that the two types of QSPEc objects will have tio
use different object IDs.

Consensus_3)
The goal of the QOS model, from the point of view of what it can specify
remains the same as now, but the QOS model will not use remapping,
however the QoS model has to describe how the end-to-end QSPEC (traffic
source description) has to be interpreted/mapped by the QNEs that are
supporting the QOS model.

This means that the QoS model can specify how the traffic source
description can be mapped into a local/private QSPEC. By using e.g.,
stacaking the end-to-end QSPEC can be then ignored.

------------Rolands proposal:

Now Roland has sent a new proposal that mainly agrees with the three
points of reached consensus, but it has two differences:

Difference_1:
The stacking concept is supported but the implementation of how the
stacking is specified and implemented is different.
Only one QSPEc is used. Thus any QoS model interprets only one QSPEC
object, which it can be either the end-to-end QSPEC or a local/private
QSPEC.

The local/private QSPEC can provide the concept of stacking by e.g.,
encapsulating the end-to-end QSPEC and providing means/ways to ignore
the end-to-end QSPEC. I suppose that if multiple levels are stacking are
used then a QOS model could ignore all the paramteres that might be
stacked by another QoS model (which is not the end-to-end QoS model) but
which arer not relevant for the local/private QoS model.
Then method/way of how teh end-to-end QSPEC has to be interpreted/mapped
has to be described in the QOS model specification.
In this way stacking is supproted, but it will not be anymore specified
in the QOS-NSLP, but in the QSPEC template draft and the QoS models.
Note that Roland does not mention in his proposal that the stacking
procedure has to be specified in the QSPEC template draft, but in my
opinuion this is necessary.

Difference_2:
Roland proposes to remove the description of all QoS model related
parameters from the QSPEC templated  draft. QoS model specific
parameters care the Y.1541, DSTE, PHB classes, etc..

I have a question for Roland:
Is my explanation of the three points of consensus, i.e., Consensus_1,
Consensus_2 and Consensus_3 in agreement with wath we have decided on
Sunday?

Do you agree that my explantion related to your proposal and to the 2
differences between the three points of consensus and your proposal is
the right one?

If yes, I can agree with Difference_1 only if we make sure that the
implementation of the stacking concept keeps its semantics. This also
means that the specification of how stacking is accomplished can be
removed from the QOS-NSLP draft, but it has to be specified/explained in
the QSPEC template draft and in the QOS models.

Regarding Difference_2 I can only agree if the editors of the QSEPC
template draft agree with the removal of all QoS model specific QSPEC
parameter descriptions from the QSPEC template draft.

Best Regards,
Georgios




------_=_NextPart_001_01C70291.F7B3E02E
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<HTML dir=3Dltr><HEAD><TITLE>[NSIS] conensus of sundays meeting and =
rolands proposal</TITLE>=0A=
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dunicode">=0A=
<META content=3D"MSHTML 6.00.2900.2963" name=3DGENERATOR><BASE =
href=3Dhttps://mclmail.saic.com/exchange/royrr/Drafts/RE:%20%5BNSIS%5D%20=
conensus%20of%20sundays%20meeting%20and%20rolands%20proposal.EML/1_text.h=
tm></HEAD>=0A=
<BODY>=0A=
<DIV id=3DidOWAReplyText83401 dir=3Dltr>=0A=
<DIV dir=3Dltr><FONT color=3D#000000 size=3D2>Hi, Georgios and =
all:</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2>It&nbsp;is MANDATORY, as Ronald and Jukka, =
"to remove </FONT><FONT size=3D2>all QoS model specific QSPEC parameter =
descriptions from the QSPEC template draft." There must NOT be any "if" =
and "when." </FONT></DIV><FONT size=3D2></FONT></DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2>Before going any step further in order to =
make progress related to this "generic" standard-track RFC draft, we =
have to take this first mandatory step. </FONT></DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2>Otherwise, we go back to the circle of =
arguments and conter-arguments of interworking. </FONT></DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2>Best regards,</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT size=3D2>Radhika</FONT></DIV>=0A=
<DIV dir=3Dltr>&nbsp;</DIV>=0A=
<DIV dir=3Dltr>=0A=
<HR tabIndex=3D-1>=0A=
<FONT face=3DTahoma size=3D2><B>From:</B> nsis-bounces@ietf.org on =
behalf of Georgios Karagiannis<BR><B>Sent:</B> Tue 11/7/2006 11:21 =
AM<BR><B>To:</B> nsis@ietf.org<BR><B>Subject:</B> [NSIS] conensus of =
sundays meeting and rolands proposal<BR></FONT><BR></DIV>=0A=
<DIV>=0A=
<P><FONT size=3D2>Hi all<BR><BR>I thought that during the meeting of =
Sunday we reached the following<BR>three points of consensus, on the =
following:<BR><BR>Consensus_1):<BR>Use the QSpec template draft as is =
but change the following:<BR><BR>The QSPEC1/QSPEC2 sematincs are =
replaced by Daves proposal for traffic<BR>source description, where a =
generic (end-to-end) QSPEC has to be<BR>interpreted by QNEs.<BR>However, =
the additions/modifcations to this proposal are related to the<BR>fact =
that a consensus was reached where the local/private QOS models =
can<BR>use stacking and that they can map/interpret the traffic =
source<BR>description into their own local/private QSPEC and ignore the =
end-to-end<BR>QSPEC.<BR>The mapping/interpreatation procedure will be =
specified in each QoS model.<BR>In this consnsus we also agreed that it =
should be possible that more than<BR>two QSPECs should be able to be =
satcked, where at least one oif them is<BR>the end-to-end =
QSPEC.<BR><BR>The way of how the stacked objects are processed should =
then be decribed<BR>in the QOS-NSLP draft.<BR><BR>The consequence of the =
above on the QOS-NSLP, QSPEC tempalet draft and<BR>QOS models is the =
following:<BR><BR>* The QSPEC template draft will not use anymore the =
notion of<BR>QSPC1/QSPEc2 objects and instead use the traffic source =
description<BR>proposed by Dave Oran.<BR><BR>* All the other QSPEC =
parameers&nbsp; that currently are described in the<BR>QSPEC template =
draft have to be considered as optional to implement by<BR>the =
QNEs.<BR><BR>* The R and Q fags are eliminated since remapping is not =
allowed.<BR><BR>* The mapping/interpreattion of the traffic source =
description is<BR>described in the QoS models and not in the QSPEC =
template draft.<BR><BR>* The QSPEC template draft has to decribe how a =
QOS model can stack the<BR>QSPEC objects, which is a similar description =
that we have now in the<BR>QSPEC templated draft. It has to empasize how =
the stacked QSPEc objest<BR>are interpreted by the QOS models.<BR><BR>* =
Further the features/procedures/etc. described in the QSPEC =
template<BR>draft remain the same.<BR><BR><BR>Consensus_2)<BR>The =
QoS-NSLP remains the same as is now with the difference that =
more<BR>than two QSPEC objects can be stacked. Note that only one is a =
real<BR>QSPEC object, i.e., end-to-end QSPEC and the rest are =
private/local<BR>QSPEC objects meaning that the two types of QSPEc =
objects will have tio<BR>use different object =
IDs.<BR><BR>Consensus_3)<BR>The goal of the QOS model, from the point of =
view of what it can specify<BR>remains the same as now, but the QOS =
model will not use remapping,<BR>however the QoS model has to describe =
how the end-to-end QSPEC (traffic<BR>source description) has to be =
interpreted/mapped by the QNEs that are<BR>supporting the QOS =
model.<BR><BR>This means that the QoS model can specify how the traffic =
source<BR>description can be mapped into a local/private QSPEC. By using =
e.g.,<BR>stacaking the end-to-end QSPEC can be then =
ignored.<BR><BR>------------Rolands proposal:<BR><BR>Now Roland has sent =
a new proposal that mainly agrees with the three<BR>points of reached =
consensus, but it has two differences:<BR><BR>Difference_1:<BR>The =
stacking concept is supported but the implementation of how =
the<BR>stacking is specified and implemented is different.<BR>Only one =
QSPEc is used. Thus any QoS model interprets only one QSPEC<BR>object, =
which it can be either the end-to-end QSPEC or a =
local/private<BR>QSPEC.<BR><BR>The local/private QSPEC can provide the =
concept of stacking by e.g.,<BR>encapsulating the end-to-end QSPEC and =
providing means/ways to ignore<BR>the end-to-end QSPEC. I suppose that =
if multiple levels are stacking are<BR>used then a QOS model could =
ignore all the paramteres that might be<BR>stacked by another QoS model =
(which is not the end-to-end QoS model) but<BR>which arer not relevant =
for the local/private QoS model.<BR>Then method/way of how teh =
end-to-end QSPEC has to be interpreted/mapped<BR>has to be described in =
the QOS model specification.<BR>In this way stacking is supproted, but =
it will not be anymore specified<BR>in the QOS-NSLP, but in the QSPEC =
template draft and the QoS models.<BR>Note that Roland does not mention =
in his proposal that the stacking<BR>procedure has to be specified in =
the QSPEC template draft, but in my<BR>opinuion this is =
necessary.<BR><BR>Difference_2:<BR>Roland proposes to remove the =
description of all QoS model related<BR>parameters from the QSPEC =
templated&nbsp; draft. QoS model specific<BR>parameters care the Y.1541, =
DSTE, PHB classes, etc..<BR><BR>I have a question for Roland:<BR>Is my =
explanation of the three points of consensus, i.e., =
Consensus_1,<BR>Consensus_2 and Consensus_3 in agreement with wath we =
have decided on<BR>Sunday?<BR><BR>Do you agree that my explantion =
related to your proposal and to the 2<BR>differences between the three =
points of consensus and your proposal is<BR>the right one?<BR><BR>If =
yes, I can agree with Difference_1 only if we make sure that =
the<BR>implementation of the stacking concept keeps its semantics. This =
also<BR>means that the specification of how stacking is accomplished can =
be<BR>removed from the QOS-NSLP draft, but it has to be =
specified/explained in<BR>the QSPEC template draft and in the QOS =
models.<BR><BR>Regarding Difference_2 I can only agree if the editors of =
the QSEPC<BR>template draft agree with the removal of all QoS model =
specific QSPEC<BR>parameter descriptions from the QSPEC template =
draft.<BR><BR>Best =
Regards,<BR>Georgios<BR><BR></FONT></P></DIV></BODY></HTML>
------_=_NextPart_001_01C70291.F7B3E02E--


--===============1153522673==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1153522673==--




From nsis-bounces@ietf.org Tue Nov 07 17:43:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhZeW-0003dr-Vk; Tue, 07 Nov 2006 17:42:44 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhZeV-0003cw-7k
	for nsis@ietf.org; Tue, 07 Nov 2006 17:42:43 -0500
Received: from mail121.messagelabs.com ([216.82.241.195])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GhZeR-00011V-H9
	for nsis@ietf.org; Tue, 07 Nov 2006 17:42:43 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-11.tower-121.messagelabs.com!1162939358!12309708!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 352 invoked from network); 7 Nov 2006 22:42:38 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-11.tower-121.messagelabs.com with SMTP;
	7 Nov 2006 22:42:38 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA7Mgac3009470; 
	Tue, 7 Nov 2006 17:42:37 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA7MgRJ7009358; Tue, 7 Nov 2006 17:42:29 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="----_=_NextPart_001_01C702BD.EEA2C5B0"
Date: Tue, 7 Nov 2006 16:41:57 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F4B@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F43@KCCLUST06EVS1.ugd.att.com>
X-MS-Has-Attach: yes
X-MS-TNEF-Correlator: 
Thread-Topic: Proposed Updates to the QSPEC Draft
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQAVmeSEAACVQV8A==
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9d54f0d83a3613448292c2f9f09a7e10
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>,
	Cornelia Kappler <cornelia.kappler@gmx.net>,
	David R Oran <oran@cisco.com>, "Kappler,
	Cornelia" <cornelia.kappler@siemens.com>
Subject: [NSIS] RE: Proposed Updates to the QSPEC Draft
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

------_=_NextPart_001_01C702BD.EEA2C5B0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi All,

A number of us met on Sunday to discuss the proposed updates to the
QSPEC draft.  The discussion focused on the issues raised on the list in
the last few days regarding the proposal (see below), i.e., that it
should be possible to 'translate' the TMOD parameter to bandwidth within
a local domain (e.g., RMD domain) and that QSPEC stacking should be
retained. =20

We agreed that stacking would be a desirable capability in some domains
(e.g., RMD) to simplify control plane processing.  Stacking also allows
a TMOD parameter to be 'translated' to a bandwidth equivalent TMOD
parameter within a local domain.

Attached are draft slides that I intend to discuss at the NSIS meeting
on Thursday AM.  The agreements that we made are pretty much summarized
on slide 8.

Please let us know of any comments on the slides.

Thanks,
Jerry=20

-----Original Message-----
From: Ash, Gerald R (Jerry), ALABS=20
Sent: Wednesday, November 01, 2006 5:13 PM
To: nsis@ietf.org
Cc: Ash, Gerald R (Jerry), ALABS; 'David R Oran'; 'Kappler, Cornelia';
'Cornelia Kappler'
Subject: Proposed Updates to the QSPEC Draft

Hi All,

Per John's suggestion below we've discussed Dave Oran's suggestions at
length and have the following proposal. =20

We think the proposed changes address many concerns raised on the list
over the last few months and will require some rather basic changes in
the QSPEC document
http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.

We appreciate Dave's suggestions and careful review of the document.

Here is a summary of the proposed changes:

1. The notion of QOSMs will be played down in the updated draft.
Language to the effect that 'QNSLP/QSPEC can signal for different QOSMs
across multiple domains' will be replaced by the notion that QNSLP/QSPEC
allows different elements on the path to implement different data plane
QoS mechanisms that meets the constraints included in the QSPEC.  A QOSM
describes common capabilities among QNEs to act consistently when
requested to admit traffic and in treating admitted traffic.  A QOSM ID
need not be defined or signaled and a QNE need not support any
particular QOSM.  A 'QOSM specification' would become a 'QNSLP/QSPEC
applicability statement', which documents how a QNE/RMF interprets and
treats the various elements in a QSPEC. =20

An updated QOSM definition is as follows:

QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
constraints and traffic handling directives a QNE implementing that QOSM
is capable of supporting and how resources will be managed by the RMF in
such a QNE.

2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
semantics:
A) The source traffic description is mandatory to include by the QNI and
mandatory to interpret by downstream QNEs.  The traffic description is
specified by a traffic model (TMOD) parameter consisting of 4
sub-parameters:
- rate (r)
- bucket size (b)
- peak rate (p)
- minimum policed unit (m)
The TMOD parameter is a mathematically complete way to describe a
traffic source.  Examples will be given for setting TMOD to correspond
to specific cases, e.g. where only bandwidth is available (r=3Dp; b and =
m
set to default values) and other cases (e.g., TCP).  As such, the
bandwidth parameter is eliminated.
B) All other QSPEC parameters are optional to include by the QNI, and
may be either mandatory or optional to interpret by downstream QNEs.
That is, the QNI can explicitly specify for each constraint parameter
whether it wishes admission control to succeed or fail if that
constraint cannot be met.  A downstream QNE distinguishes whether it is
failure to understand the parameter or failure to meet the constraint
that causes the admission control failure (more detail on this below).
C) Additional QSPEC parameters can be defined in updates to the QSPEC
document (or alternatively specified in QOSM applicability statement
documents).  There is already an IANA registry established for
additional QSPEC parameters.

3. The concept of remapping QSPEC parameters is eliminated.  This
concept is ill defined and has caused much confusion (there are few if
any sensible remappings).  Rather we define 'interpret' a QSPEC
parameter to only mean what we now call 'strictly interpret' (i.e., must
conform to the RFCs defining the parameter and procedures) and eliminate
the remap option part of 'interpret'.

4. Since remapping is eliminated, the concept of stacking QSPECs is also
eliminated.  Discussion of tunneling of QNSLP messages is eliminated
since there are no impacts on tunneling required by QSPEC.

5. Current QSPEC flags are modified as follows:
A) The QNI sets the M flag for each QSPEC parameter it populates that it
intends must be interpreted or the reservation fails.  Currently the M
flag is statically set for every QSPEC parameter.
B) If the M flag is set, a downstream QNE must interpret the parameter
or the reservation fails.  If a QNE does not support the parameter with
the M flag set, it sets the N flag and rejects the reservation.  If the
QNE supports the parameter but cannot meet the parameter, it sets the E
flag and rejects the reservation.=20
C) If the M flag is not set, a downstream QNE should interpret the
parameter.  If a QNE does not support the parameter, it sets the N flag
and can optionally accept or reject the reservation.  If a QNE supports
the parameter but cannot meet the parameter, it sets the E flag and
optional accepts or rejects the reservation.
D) The R (remapped parameter) flag and Q (non QOSM) flag are eliminated.

6. Summary of proposed QSPEC parameters:

Source traffic Description:
 TMOD-1 (mandatory to include)
 TMOD-2 (optional to include)

Constraints (optional to include):
 Path Latency
 Path Jitter
 Path PLR
 Path PER
 Y.1541 class (shorthand notation for above 4 parameters)
 Slack term
 Priority (Preemption, Defending, Admission, RPH Priority)
 DSTE class

Handling directives (optional to include):
 excess treatment

Traffic Classifiers (optional to include):
 PHB Class (note that PHB class is set by downstream QNE to tell QNI how
to mark traffic to ensure it will be treated as committed by admission
control)

Eliminated:
 bandwidth
 Ctot, Dtot, Csum, Dsum

7. Editorial changes:
A) rewrite description of read-only and read-write parameters to make
the rules more clear:=20
- which parameters can be ro/rw (function of object in which parameter
included)
- can only increase rw values and not delete, etc.
B) eliminate the distinction of 'control' and 'qos description'
parameters and replace with suggested taxonomy, as described above
C) clarify discussion of packet outcomes from Y.1540 and place in
appropriate sections
D) clarify which parameters can be included in <QoS Desired> <QoS
Available>, <QoS Reserved>, or <QoS Minimum>

Please review and comment on the proposed changes.

Thanks,
Jerry, Dave, Cornelia


-----Original Message-----
From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]=20
Sent: Monday, October 23, 2006 2:59 AM
To: nsis@ietf.org
Subject: [NSIS] Where we stand with the draft-ietf-nsis-qspec draft

Hi all,

We have had a very large thread on the current
draft-ietf-nsis-qspec-12.txt draft.  We don't have a clear consensus, so
I'd like to try to summarize the discussions.

Several people feel that the current draft contains too many mandatory
parameters and mapping some QoS Models to others is mixing
apples-to-oranges.  The feeling is that interoperability is hard to
achieve and therefore we should not try to gain interoperability.  Jukka
Manner's message on 17th of October proposed a way forward.
http://www1.ietf.org/mail-archive/web/nsis/current/msg06919.html

However, several (other) people feel that interoperability is important
and is needed. Giving up on interoperability would not be a good thing,
and that we are very close to a solution.  Dave Oran proposed a
mechanism to reach a solution on the 19th October:
http://www1.ietf.org/mail-archive/web/nsis/current/msg06962.html

I have asked Jerry and Dave to make a proposal along the lines of Dave's
message above, and the working group can review this, to see if we can
gain consensus on this.  I think it would be sufficient to send some
rough text to the mailing list, and if this is OK, then we can revise
the draft.

One open question to Dave/Jerry about the proposal is, in the event that
the source cannot be described a mathematically complete way, is there
any value (or does it make any sense at all) in sending an incomplete
qspec, or potentially "wildcarding" sending some of the mandatory
parameters.  I'm wondering if some sources / deployments may have
certain parameters that are relatively well defined, but not all are so
well defined.

thanks,
John

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

------_=_NextPart_001_01C702BD.EEA2C5B0
Content-Type: application/octet-stream; name="vgs.ietf67.nsis.qspec.110906.pdf"
Content-Transfer-Encoding: base64
Content-Description: vgs.ietf67.nsis.qspec.110906.pdf
Content-Disposition: attachment;
	filename="vgs.ietf67.nsis.qspec.110906.pdf"
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------_=_NextPart_001_01C702BD.EEA2C5B0
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

------_=_NextPart_001_01C702BD.EEA2C5B0--




From nsis-bounces@ietf.org Tue Nov 07 20:02:25 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhboZ-0005nO-PR; Tue, 07 Nov 2006 20:01:15 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhboY-0005mX-G4
	for nsis@ietf.org; Tue, 07 Nov 2006 20:01:14 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GhboX-0002Nk-09
	for nsis@ietf.org; Tue, 07 Nov 2006 20:01:14 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-14.tower-146.messagelabs.com!1162947671!531971!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 20875 invoked from network); 8 Nov 2006 01:01:11 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-14.tower-146.messagelabs.com with SMTP;
	8 Nov 2006 01:01:11 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA811AE7004904; 
	Tue, 7 Nov 2006 20:01:10 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA810vvw004817; Tue, 7 Nov 2006 20:01:03 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="----_=_NextPart_001_01C702D1.3DEA60C4"
Date: Tue, 7 Nov 2006 19:00:11 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F4F@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F4B@KCCLUST06EVS1.ugd.att.com>
X-MS-Has-Attach: yes
X-MS-TNEF-Correlator: 
Thread-Topic: Proposed Updates to the QSPEC Draft
Thread-Index: Acb2cMiplW5+UcujTsGCzloRBxObRAGwSgBQAVmeSEAACVQV8AAE134Q
From: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>
To: <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d7d3f294c642eb492c2d44336474db48
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>,
	Cornelia Kappler <cornelia.kappler@gmx.net>,
	David R Oran <oran@cisco.com>, "Kappler,
	Cornelia" <cornelia.kappler@siemens.com>
Subject: [NSIS] RE: Proposed Updates to the QSPEC Draft
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

------_=_NextPart_001_01C702D1.3DEA60C4
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Hi All,

A number of us met on Sunday to discuss the proposed updates to the
QSPEC draft.  The discussion focused on the issues raised on the list in
the last few days regarding the proposal (see below), i.e., that it
should be possible to 'translate' the TMOD parameter to bandwidth within
a local domain (e.g., RMD domain) and that QSPEC stacking should be
retained. =20

We agreed that stacking would be a desirable capability in some domains
(e.g., RMD) to simplify control plane processing.  Stacking also allows
a TMOD parameter to be 'translated' to a bandwidth equivalent TMOD
parameter within a local domain.

Attached are draft slides that I intend to discuss at the NSIS meeting
on Thursday AM.  The agreements that we made are pretty much summarized
on slide 8.

Please let us know of any comments on the slides.

Thanks,
Jerry=20

-----Original Message-----
From: Ash, Gerald R (Jerry), ALABS=20
Sent: Wednesday, November 01, 2006 5:13 PM
To: nsis@ietf.org
Cc: Ash, Gerald R (Jerry), ALABS; 'David R Oran'; 'Kappler, Cornelia';
'Cornelia Kappler'
Subject: Proposed Updates to the QSPEC Draft

Hi All,

Per John's suggestion below we've discussed Dave Oran's suggestions at
length and have the following proposal. =20

We think the proposed changes address many concerns raised on the list
over the last few months and will require some rather basic changes in
the QSPEC document
http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.

We appreciate Dave's suggestions and careful review of the document.

Here is a summary of the proposed changes:

1. The notion of QOSMs will be played down in the updated draft.
Language to the effect that 'QNSLP/QSPEC can signal for different QOSMs
across multiple domains' will be replaced by the notion that QNSLP/QSPEC
allows different elements on the path to implement different data plane
QoS mechanisms that meets the constraints included in the QSPEC.  A QOSM
describes common capabilities among QNEs to act consistently when
requested to admit traffic and in treating admitted traffic.  A QOSM ID
need not be defined or signaled and a QNE need not support any
particular QOSM.  A 'QOSM specification' would become a 'QNSLP/QSPEC
applicability statement', which documents how a QNE/RMF interprets and
treats the various elements in a QSPEC. =20

An updated QOSM definition is as follows:

QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
constraints and traffic handling directives a QNE implementing that QOSM
is capable of supporting and how resources will be managed by the RMF in
such a QNE.

2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
semantics:
A) The source traffic description is mandatory to include by the QNI and
mandatory to interpret by downstream QNEs.  The traffic description is
specified by a traffic model (TMOD) parameter consisting of 4
sub-parameters:
- rate (r)
- bucket size (b)
- peak rate (p)
- minimum policed unit (m)
The TMOD parameter is a mathematically complete way to describe a
traffic source.  Examples will be given for setting TMOD to correspond
to specific cases, e.g. where only bandwidth is available (r=3Dp; b and =
m
set to default values) and other cases (e.g., TCP).  As such, the
bandwidth parameter is eliminated.
B) All other QSPEC parameters are optional to include by the QNI, and
may be either mandatory or optional to interpret by downstream QNEs.
That is, the QNI can explicitly specify for each constraint parameter
whether it wishes admission control to succeed or fail if that
constraint cannot be met.  A downstream QNE distinguishes whether it is
failure to understand the parameter or failure to meet the constraint
that causes the admission control failure (more detail on this below).
C) Additional QSPEC parameters can be defined in updates to the QSPEC
document (or alternatively specified in QOSM applicability statement
documents).  There is already an IANA registry established for
additional QSPEC parameters.

3. The concept of remapping QSPEC parameters is eliminated.  This
concept is ill defined and has caused much confusion (there are few if
any sensible remappings).  Rather we define 'interpret' a QSPEC
parameter to only mean what we now call 'strictly interpret' (i.e., must
conform to the RFCs defining the parameter and procedures) and eliminate
the remap option part of 'interpret'.

4. Since remapping is eliminated, the concept of stacking QSPECs is also
eliminated.  Discussion of tunneling of QNSLP messages is eliminated
since there are no impacts on tunneling required by QSPEC.

5. Current QSPEC flags are modified as follows:
A) The QNI sets the M flag for each QSPEC parameter it populates that it
intends must be interpreted or the reservation fails.  Currently the M
flag is statically set for every QSPEC parameter.
B) If the M flag is set, a downstream QNE must interpret the parameter
or the reservation fails.  If a QNE does not support the parameter with
the M flag set, it sets the N flag and rejects the reservation.  If the
QNE supports the parameter but cannot meet the parameter, it sets the E
flag and rejects the reservation.=20
C) If the M flag is not set, a downstream QNE should interpret the
parameter.  If a QNE does not support the parameter, it sets the N flag
and can optionally accept or reject the reservation.  If a QNE supports
the parameter but cannot meet the parameter, it sets the E flag and
optional accepts or rejects the reservation.
D) The R (remapped parameter) flag and Q (non QOSM) flag are eliminated.

6. Summary of proposed QSPEC parameters:

Source traffic Description:
 TMOD-1 (mandatory to include)
 TMOD-2 (optional to include)

Constraints (optional to include):
 Path Latency
 Path Jitter
 Path PLR
 Path PER
 Y.1541 class (shorthand notation for above 4 parameters)
 Slack term
 Priority (Preemption, Defending, Admission, RPH Priority)
 DSTE class

Handling directives (optional to include):
 excess treatment

Traffic Classifiers (optional to include):
 PHB Class (note that PHB class is set by downstream QNE to tell QNI how
to mark traffic to ensure it will be treated as committed by admission
control)

Eliminated:
 bandwidth
 Ctot, Dtot, Csum, Dsum

7. Editorial changes:
A) rewrite description of read-only and read-write parameters to make
the rules more clear:=20
- which parameters can be ro/rw (function of object in which parameter
included)
- can only increase rw values and not delete, etc.
B) eliminate the distinction of 'control' and 'qos description'
parameters and replace with suggested taxonomy, as described above
C) clarify discussion of packet outcomes from Y.1540 and place in
appropriate sections
D) clarify which parameters can be included in <QoS Desired> <QoS
Available>, <QoS Reserved>, or <QoS Minimum>

Please review and comment on the proposed changes.

Thanks,
Jerry, Dave, Cornelia

------_=_NextPart_001_01C702D1.3DEA60C4
Content-Type: application/octet-stream; name="vgs.ietf67.nsis.qspec.110906.pdf"
Content-Transfer-Encoding: base64
Content-Description: vgs.ietf67.nsis.qspec.110906.pdf
Content-Disposition: attachment;
	filename="vgs.ietf67.nsis.qspec.110906.pdf"
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------_=_NextPart_001_01C702D1.3DEA60C4
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

------_=_NextPart_001_01C702D1.3DEA60C4--




From nsis-bounces@ietf.org Wed Nov 08 05:46:04 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhkvI-0000Yn-DB; Wed, 08 Nov 2006 05:44:48 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhkvG-0000TD-SB
	for nsis@ietf.org; Wed, 08 Nov 2006 05:44:46 -0500
Received: from david.siemens.de ([192.35.17.14])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GhkvD-0001yy-Dy
	for nsis@ietf.org; Wed, 08 Nov 2006 05:44:46 -0500
Received: from mail2.siemens.de (localhost [127.0.0.1])
	by david.siemens.de (8.12.6/8.12.6) with ESMTP id kA8AiYtb028850;
	Wed, 8 Nov 2006 11:44:35 +0100
Received: from mchp771a.ww002.siemens.net (mchp771a.ww002.siemens.net
	[139.25.131.189])
	by mail2.siemens.de (8.12.6/8.12.6) with ESMTP id kA8AhqkB002716;
	Wed, 8 Nov 2006 11:44:34 +0100
Received: from blns00fa.ww002.siemens.net ([147.54.45.102]) by
	mchp771a.ww002.siemens.net with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 8 Nov 2006 11:44:15 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: AW: [NSIS] Question about " QSPEC Proc."
Date: Wed, 8 Nov 2006 11:44:13 +0100
Message-ID: <DAA882AD01598448BB6FAB8C4D96F9DB02A7E31E@blns00fa.ww002.siemens.net>
In-Reply-To: <Pine.LNX.4.64.0611060606020.28899@sbz-31.cs.Helsinki.FI>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Question about " QSPEC Proc."
Thread-Index: AccBWZMilfo5Bl9KTsikfJPNc+BAwgBw66uw
From: "Kappler, Cornelia" <cornelia.kappler@siemens.com>
To: "Jukka MJ Manner" <jmanner@cs.Helsinki.FI>, <nsis@ietf.org>
X-OriginalArrivalTime: 08 Nov 2006 10:44:15.0262 (UTC)
	FILETIME=[D5B663E0:01C70322]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9ed51c9d1356100bce94f1ae4ec616a9
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jukka, =20

[Just as a reminder for those not so familiar with the details of the =
QSPEC:]=20
The "QSPEC procedure" ID is included in a QoS NSLP message (currently in =
the QSPEC header) to identify the expected message sequence (e.g. =
QUERY-RESERVE-RESPONSE) and the QoS Objects to include in each message =
(e.g. QoS Available into QUERY, QoS Desired in RESERVE and QoS Reserved =
in RESPONSE). The QSPEC Procedure ID is included by the QNE initiating a =
signaling session to inform the receiving QNE about the message it is =
supposed to generate in reply.

I am not sure where the decision is taken about which QSPEC procedure to =
use; presumably in the application or the RMF. It must be included in a =
message, but I am not sure where the best place is (QoS NSLP or QSPEC). =
It may have to be passed over the (non.standardized) API between QoS =
NSLP and QSPEC.

Cornelia

> -----Urspr=FCngliche Nachricht-----
> Von: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI]=20
> Gesendet: Montag, 6. November 2006 05:09
> An: nsis@ietf.org
> Betreff: [NSIS] Question about " QSPEC Proc."
>=20
>=20
> Hi Jerry et al.,
>=20
> I was reading the QSPEC spec with Roland, and we were both=20
> wondering what=20
> the " QSPEC Proc." field in the QSPEC Common Header is? Isn't this=20
> something that should be actually in the API between the NSLP=20
> and RMF, not=20
> encoded inside the QSPEC itself?
>=20
> Jukka
>=20
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 08 12:00:57 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhqmF-0001Dq-Er; Wed, 08 Nov 2006 11:59:51 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhqmD-0001Dj-Kb
	for nsis@ietf.org; Wed, 08 Nov 2006 11:59:49 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GhqmC-0000ho-BO
	for nsis@ietf.org; Wed, 08 Nov 2006 11:59:49 -0500
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GhqmA-0003qE-Rc; Wed, 08 Nov 2006 17:59:47 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id AFDB6B5B0;
	Wed,  8 Nov 2006 17:59:46 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GhqmA-0003Mz-5p; Wed, 08 Nov 2006 17:59:46 +0100
Message-ID: <45520C93.6070901@tm.uka.de>
Date: Wed, 08 Nov 2006 17:57:55 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.7) Gecko/20060909 Thunderbird/1.5.0.7 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: "Kappler, Cornelia" <cornelia.kappler@siemens.com>
Subject: Re: AW: [NSIS] Question about " QSPEC Proc."
References: <DAA882AD01598448BB6FAB8C4D96F9DB02A7E31E@blns00fa.ww002.siemens.net>
In-Reply-To: <DAA882AD01598448BB6FAB8C4D96F9DB02A7E31E@blns00fa.ww002.siemens.net>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.2 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9182cfff02fae4f1b6e9349e01d62f32
Cc: Jukka MJ Manner <jmanner@cs.Helsinki.FI>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Cornelia,

Kappler, Cornelia schrieb:
> [Just as a reminder for those not so familiar with the details of the
> QSPEC:] The "QSPEC procedure" ID is included in a QoS NSLP message
> (currently in the QSPEC header) to identify the expected message
> sequence (e.g. QUERY-RESERVE-RESPONSE) and the QoS Objects to include
> in each message (e.g. QoS Available into QUERY, QoS Desired in
> RESERVE and QoS Reserved in RESPONSE). The QSPEC Procedure ID is
> included by the QNE initiating a signaling session to inform the
> receiving QNE about the message it is supposed to generate in reply.
> 
> I am not sure where the decision is taken about which QSPEC procedure
> to use; presumably in the application or the RMF. It must be included
> in a message, but I am not sure where the best place is (QoS NSLP or
> QSPEC). It may have to be passed over the (non.standardized) API
> between QoS NSLP and QSPEC.

Maybe I don't get the point completely, but I would expect
that it is an internal API issue (currently we don't have any
QoS NSLP API), so QoS-NSLP would usually indicate the message type
(QUERY, RESERVE, RESPONSE) to the RMF/Application.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 08 12:17:17 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Ghr1r-0002O9-Qm; Wed, 08 Nov 2006 12:15:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Ghr1o-0002Nj-7R
	for nsis@ietf.org; Wed, 08 Nov 2006 12:15:56 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Ghr1j-0002qA-NO
	for nsis@ietf.org; Wed, 08 Nov 2006 12:15:56 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Wed, 08 Nov 2006 19:15:46 +0200
	id 000AFDE7.455210C2.00003B5D
Date: Wed, 8 Nov 2006 19:15:46 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: "Kappler, Cornelia" <cornelia.kappler@siemens.com>
Subject: Re: AW: [NSIS] Question about " QSPEC Proc."
In-Reply-To: <DAA882AD01598448BB6FAB8C4D96F9DB02A7E31E@blns00fa.ww002.siemens.net>
Message-ID: <Pine.LNX.4.64.0611081913070.12350@sbz-31.cs.Helsinki.FI>
References: <DAA882AD01598448BB6FAB8C4D96F9DB02A7E31E@blns00fa.ww002.siemens.net>
Mime-Version: 1.0
Content-Type: multipart/mixed; boundary="=_courier-15212-1163006151-0001-2"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f60d0f7806b0c40781eee6b9cd0b2135
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

This is a MIME-formatted message.  If you see this text it means that your
E-mail software does not support MIME-formatted messages.

--=_courier-15212-1163006151-0001-2
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable
X-Mime-Autoconverted: from 8bit to quoted-printable by courier 0.47


Hi,

The NSLP knows what messaging is going on, it doesn't need to be specifi=
ed 
in some additional tags. Therefore I thought that this information can b=
e 
passed between the RMF through an API. But I guess you could encode that=
 
into the QSPEC by the QNI/QNR, if there is some good reason.

Jukka

On Wed, 8 Nov 2006, Kappler, Cornelia wrote:

> Hi Jukka,  
> 
> [Just as a reminder for those not so familiar with the details of the =
QSPEC:] 
> The "QSPEC procedure" ID is included in a QoS NSLP message (currently =
in the QSPEC header) to identify the expected message sequence (e.g. QUE=
RY-RESERVE-RESPONSE) and the QoS Objects to include in each message (e.g=
. QoS Available into QUERY, QoS Desired in RESERVE and QoS Reserved in R=
ESPONSE). The QSPEC Procedure ID is included by the QNE initiating a sig=
naling session to inform the receiving QNE about the message it is suppo=
sed to generate in reply.
> 
> I am not sure where the decision is taken about which QSPEC procedure =
to use; presumably in the application or the RMF. It must be included in=
 a message, but I am not sure where the best place is (QoS NSLP or QSPEC=
). It may have to be passed over the (non.standardized) API between QoS =
NSLP and QSPEC.
> 
> Cornelia
> 
> > -----Urspr=FCngliche Nachricht-----
> > Von: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI] 
> > Gesendet: Montag, 6. November 2006 05:09
> > An: nsis@ietf.org
> > Betreff: [NSIS] Question about " QSPEC Proc."
> > 
> > 
> > Hi Jerry et al.,
> > 
> > I was reading the QSPEC spec with Roland, and we were both 
> > wondering what 
> > the " QSPEC Proc." field in the QSPEC Common Header is? Isn't this 
> > something that should be actually in the API between the NSLP 
> > and RMF, not 
> > encoded inside the QSPEC itself?
> > 
> > Jukka
> > 
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
> > 
> 
--=_courier-15212-1163006151-0001-2
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--=_courier-15212-1163006151-0001-2--




From nsis-bounces@ietf.org Wed Nov 08 13:11:55 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhrtP-0004js-BI; Wed, 08 Nov 2006 13:11:19 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhrtN-0004i7-0P
	for nsis@ietf.org; Wed, 08 Nov 2006 13:11:17 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Ghrpe-0002Ks-5r
	for nsis@ietf.org; Wed, 08 Nov 2006 13:07:27 -0500
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GhrpY-0007ST-Rv; Wed, 08 Nov 2006 19:07:23 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id AB72C854A;
	Wed,  8 Nov 2006 19:07:20 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GhrpY-0003R6-37; Wed, 08 Nov 2006 19:07:20 +0100
Message-ID: <45521C6A.5010103@tm.uka.de>
Date: Wed, 08 Nov 2006 19:05:30 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.7) Gecko/20060909 Thunderbird/1.5.0.7 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] conensus of sundays meeting and rolands proposal
References: <zEW2gRVz.1162916487.9969660.karagian@ewi.utwente.nl>
In-Reply-To: <zEW2gRVz.1162916487.9969660.karagian@ewi.utwente.nl>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -2.9 (--)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-1.1 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 10d3e4e3c32e363f129e380e644649be
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

Georgios Karagiannis wrote:

> Now Roland has sent a new proposal that mainly agrees with the three
> points of reached consensus, but it has two differences:
> 
> Difference_1:
> The stacking concept is supported but the implementation of how the
> stacking is specified and implemented is different.
> Only one QSPEc is used. Thus any QoS model interprets only one QSPEC
> object, which it can be either the end-to-end QSPEC or a local/private
> QSPEC.

Not quite. QoS-NSLP carries only one QSPEC object. It doesn't
mean that only one QSPEC is used or only one QSPEC is interpreted.
So if the X-QoS Model wants to carry also the E2E-QSPEC in its
own X-QSPEC it may interpret or not interpret the included E2E-QSPEC.
That's defined in the X-QoS Model/X-QSPEC

> The local/private QSPEC can provide the concept of stacking by e.g.,
> encapsulating the end-to-end QSPEC and providing means/ways to ignore
> the end-to-end QSPEC. I suppose that if multiple levels are stacking are
> used then a QOS model could ignore all the paramteres that might be
> stacked by another QoS model (which is not the end-to-end QoS model) but
> which arer not relevant for the local/private QoS model.

Yep.

> Then method/way of how teh end-to-end QSPEC has to be interpreted/mapped
> has to be described in the QOS model specification.
> In this way stacking is supproted, but it will not be anymore specified
> in the QOS-NSLP, but in the QSPEC template draft and the QoS models.

Yep.

> Note that Roland does not mention in his proposal that the stacking
> procedure has to be specified in the QSPEC template draft, but in my
> opinuion this is necessary.

That actually depends. The current QSPEC template draft actually
mixes the template and the interop/Generic QSPEC things. One could
split it into two documents:
- one describing common QSPEC definition aspects (template)
- one the E2E-QSPEC (as an application of the previous draft)

> Difference_2:
> Roland proposes to remove the description of all QoS model related
> parameters from the QSPEC templated  draft. QoS model specific
> parameters care the Y.1541, DSTE, PHB classes, etc..

Yep.

> I have a question for Roland:
> Is my explanation of the three points of consensus, i.e., Consensus_1,
> Consensus_2 and Consensus_3 in agreement with wath we have decided on
> Sunday?

I'm not really sure that we reached _clear_ consensus there.
I guess that I understood your problems with the new proposal better.

> Do you agree that my explantion related to your proposal and to the 2
> differences between the three points of consensus and your proposal is
> the right one?
> 
> If yes, I can agree with Difference_1 only if we make sure that the
> implementation of the stacking concept keeps its semantics. This also
> means that the specification of how stacking is accomplished can be
> removed from the QOS-NSLP draft, but it has to be specified/explained in
> the QSPEC template draft and in the QOS models.

As I said: whether a QoS model needs something like stacking or not is
up to its own decision.

> Regarding Difference_2 I can only agree if the editors of the QSEPC
> template draft agree with the removal of all QoS model specific QSPEC
> parameter descriptions from the QSPEC template draft.

My idea behind the proposal was to have a clearer separation between
different models. Otherwise if a QSPEC crosses several domains each
QoS model may attach its own parameters, maybe only optional. In this
case I would like to see a scope flag as Jukka suggested so that
QOSM-specific parameters should probably be removed before passing on
to the next domain. Otherwise I don't know what an RMD-domain would
gain from seeing any Y.1541-specific parameters.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 08 13:23:07 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Ghs41-0001xv-52; Wed, 08 Nov 2006 13:22:17 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Ghs40-0001xq-Ia
	for nsis@ietf.org; Wed, 08 Nov 2006 13:22:16 -0500
Received: from mgw-ext14.nokia.com ([131.228.20.173])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Ghs3z-0005CL-5o
	for nsis@ietf.org; Wed, 08 Nov 2006 13:22:16 -0500
Received: from esebh108.NOE.Nokia.com (esebh108.ntc.nokia.com [172.21.143.145])
	by mgw-ext14.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA8ILkPM010201; Wed, 8 Nov 2006 20:22:12 +0200
Received: from esebh102.NOE.Nokia.com ([172.21.138.183]) by
	esebh108.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 8 Nov 2006 20:22:03 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh102.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 8 Nov 2006 20:22:03 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Wed, 8 Nov 2006 20:22:02 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED059E@esebe199.NOE.Nokia.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Presenters - slides please
Thread-Index: AccDYsj+VNN5nTDqTKK2egbRKtMiNA==
From: <john.loughney@nokia.com>
To: <nsis@ietf.org>
X-OriginalArrivalTime: 08 Nov 2006 18:22:03.0546 (UTC)
	FILETIME=[CA15E7A0:01C70362]
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 2870a44b67ee17965ce5ad0177e150f4
Cc: stiemerling@netlab.nec.de
Subject: [NSIS] Presenters - slides please
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Presenters, please send your slides to me by 5PM, so I can post them
before the meeting.

John

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 08 17:34:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GhvzK-0005f2-Ik; Wed, 08 Nov 2006 17:33:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GhvzI-0005em-LJ
	for nsis@ietf.org; Wed, 08 Nov 2006 17:33:40 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GhvzA-0003Ie-Ee
	for nsis@ietf.org; Wed, 08 Nov 2006 17:33:40 -0500
Received: from webmail.cs.utwente.nl (janus.ewi.utwente.nl [130.89.10.26])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with SMTP id kA8MWirH005205;
	Wed, 8 Nov 2006 23:32:45 +0100 (MET)
Received: from 130.129.32.143 (auth. user karagian@imap2.cs.utwente.nl)
	by webmail.cs.utwente.nl with HTTP; Wed, 08 Nov 2006 22:32:44 +0000
To: "Ash, Gerald R (Jerry), NPE" <gash@att.com>,
	"nsis@ietf.org" <nsis@ietf.org>
Subject: [NSIS] RE: Proposed Updates to the QSPEC Draft
Date: Wed, 08 Nov 2006 22:32:43 +0000
X-Mailer: IlohaMail/0.8.13 (On: webmail.cs.utwente.nl)
Message-ID: <86V4zTmq.1163025163.7700040.karagian@ewi.utwente.nl>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F4F@KCCLUST06EVS1.ugd.att.com>
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
Bounce-To: "Georgios Karagiannis" <karagian@cs.utwente.nl>
MIME-Version: 1.0 
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: 0.3 () FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Wed, 08 Nov 2006 23:32:47 +0100 (MET)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 48472a944c87678fcfe8db15ffecdfff
Cc: "Ash, Gerald R \(Jerry\), NPE" <gash@att.com>, "Kappler,
	Cornelia" <cornelia.kappler@siemens.com>, David R Oran <oran@cisco.com>,
	Cornelia Kappler <cornelia.kappler@gmx.net>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

I think taht the slides encompass most of the points of the reached
consensus, except the fact that we talked about using stacking of more
than one QSPEC objects, where only one of them is the end-to-end QSPEC
and the rest are private/local QSPECs that are belonging to
hierearchical nested QoS models.
Could you please also check the 3 points of consensus that I have
summarized in the e-mail that I have sent yestrday on the nsis mailing
list.

Best Regards,
Georgios


On 11/8/2006, "Ash, Gerald R (Jerry), NPE" <gash@att.com> wrote:

>Hi All,
>
>A number of us met on Sunday to discuss the proposed updates to the
>QSPEC draft.  The discussion focused on the issues raised on the list in
>the last few days regarding the proposal (see below), i.e., that it
>should be possible to 'translate' the TMOD parameter to bandwidth within
>a local domain (e.g., RMD domain) and that QSPEC stacking should be
>retained. =20
>
>We agreed that stacking would be a desirable capability in some domains
>(e.g., RMD) to simplify control plane processing.  Stacking also allows
>a TMOD parameter to be 'translated' to a bandwidth equivalent TMOD
>parameter within a local domain.
>
>Attached are draft slides that I intend to discuss at the NSIS meeting
>on Thursday AM.  The agreements that we made are pretty much summarized
>on slide 8.
>
>Please let us know of any comments on the slides.
>
>Thanks,
>Jerry=20
>
>-----Original Message-----
>From: Ash, Gerald R (Jerry), ALABS=20
>Sent: Wednesday, November 01, 2006 5:13 PM
>To: nsis@ietf.org
>Cc: Ash, Gerald R (Jerry), ALABS; 'David R Oran'; 'Kappler, Cornelia';
>'Cornelia Kappler'
>Subject: Proposed Updates to the QSPEC Draft
>
>Hi All,
>
>Per John's suggestion below we've discussed Dave Oran's suggestions at
>length and have the following proposal. =20
>
>We think the proposed changes address many concerns raised on the list
>over the last few months and will require some rather basic changes in
>the QSPEC document
>http://www.ietf.org/internet-drafts/draft-ietf-nsis-qspec-12.txt.
>
>We appreciate Dave's suggestions and careful review of the document.
>
>Here is a summary of the proposed changes:
>
>1. The notion of QOSMs will be played down in the updated draft.
>Language to the effect that 'QNSLP/QSPEC can signal for different QOSMs
>across multiple domains' will be replaced by the notion that QNSLP/QSPEC
>allows different elements on the path to implement different data plane
>QoS mechanisms that meets the constraints included in the QSPEC.  A QOSM
>describes common capabilities among QNEs to act consistently when
>requested to admit traffic and in treating admitted traffic.  A QOSM ID
>need not be defined or signaled and a QNE need not support any
>particular QOSM.  A 'QOSM specification' would become a 'QNSLP/QSPEC
>applicability statement', which documents how a QNE/RMF interprets and
>treats the various elements in a QSPEC. =20
>
>An updated QOSM definition is as follows:
>
>QoS Model (QOSM): A method to achieve QoS for a traffic flow, e.g.,
>IntServ Controlled Load.  A QOSM specifies what sub-set of QSPEC QoS
>constraints and traffic handling directives a QNE implementing that QOSM
>is capable of supporting and how resources will be managed by the RMF in
>such a QNE.
>
>2. The QSPEC1/QSPEC2 semantics will be replaced with Dave's a) --> d)
>semantics:
>A) The source traffic description is mandatory to include by the QNI and
>mandatory to interpret by downstream QNEs.  The traffic description is
>specified by a traffic model (TMOD) parameter consisting of 4
>sub-parameters:
>- rate (r)
>- bucket size (b)
>- peak rate (p)
>- minimum policed unit (m)
>The TMOD parameter is a mathematically complete way to describe a
>traffic source.  Examples will be given for setting TMOD to correspond
>to specific cases, e.g. where only bandwidth is available (r=3Dp; b and m
>set to default values) and other cases (e.g., TCP).  As such, the
>bandwidth parameter is eliminated.
>B) All other QSPEC parameters are optional to include by the QNI, and
>may be either mandatory or optional to interpret by downstream QNEs.
>That is, the QNI can explicitly specify for each constraint parameter
>whether it wishes admission control to succeed or fail if that
>constraint cannot be met.  A downstream QNE distinguishes whether it is
>failure to understand the parameter or failure to meet the constraint
>that causes the admission control failure (more detail on this below).
>C) Additional QSPEC parameters can be defined in updates to the QSPEC
>document (or alternatively specified in QOSM applicability statement
>documents).  There is already an IANA registry established for
>additional QSPEC parameters.
>
>3. The concept of remapping QSPEC parameters is eliminated.  This
>concept is ill defined and has caused much confusion (there are few if
>any sensible remappings).  Rather we define 'interpret' a QSPEC
>parameter to only mean what we now call 'strictly interpret' (i.e., must
>conform to the RFCs defining the parameter and procedures) and eliminate
>the remap option part of 'interpret'.
>
>4. Since remapping is eliminated, the concept of stacking QSPECs is also
>eliminated.  Discussion of tunneling of QNSLP messages is eliminated
>since there are no impacts on tunneling required by QSPEC.
>
>5. Current QSPEC flags are modified as follows:
>A) The QNI sets the M flag for each QSPEC parameter it populates that it
>intends must be interpreted or the reservation fails.  Currently the M
>flag is statically set for every QSPEC parameter.
>B) If the M flag is set, a downstream QNE must interpret the parameter
>or the reservation fails.  If a QNE does not support the parameter with
>the M flag set, it sets the N flag and rejects the reservation.  If the
>QNE supports the parameter but cannot meet the parameter, it sets the E
>flag and rejects the reservation.=20
>C) If the M flag is not set, a downstream QNE should interpret the
>parameter.  If a QNE does not support the parameter, it sets the N flag
>and can optionally accept or reject the reservation.  If a QNE supports
>the parameter but cannot meet the parameter, it sets the E flag and
>optional accepts or rejects the reservation.
>D) The R (remapped parameter) flag and Q (non QOSM) flag are eliminated.
>
>6. Summary of proposed QSPEC parameters:
>
>Source traffic Description:
> TMOD-1 (mandatory to include)
> TMOD-2 (optional to include)
>
>Constraints (optional to include):
> Path Latency
> Path Jitter
> Path PLR
> Path PER
> Y.1541 class (shorthand notation for above 4 parameters)
> Slack term
> Priority (Preemption, Defending, Admission, RPH Priority)
> DSTE class
>
>Handling directives (optional to include):
> excess treatment
>
>Traffic Classifiers (optional to include):
> PHB Class (note that PHB class is set by downstream QNE to tell QNI how
>to mark traffic to ensure it will be treated as committed by admission
>control)
>
>Eliminated:
> bandwidth
> Ctot, Dtot, Csum, Dsum
>
>7. Editorial changes:
>A) rewrite description of read-only and read-write parameters to make
>the rules more clear:=20
>- which parameters can be ro/rw (function of object in which parameter
>included)
>- can only increase rw values and not delete, etc.
>B) eliminate the distinction of 'control' and 'qos description'
>parameters and replace with suggested taxonomy, as described above
>C) clarify discussion of packet outcomes from Y.1540 and place in
>appropriate sections
>D) clarify which parameters can be included in <QoS Desired> <QoS
>Available>, <QoS Reserved>, or <QoS Minimum>
>
>Please review and comment on the proposed changes.
>
>Thanks,
>Jerry, Dave, Cornelia

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 09:31:14 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiAv2-0005hR-5P; Thu, 09 Nov 2006 09:30:16 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiAv0-0005hK-Vm
	for nsis@ietf.org; Thu, 09 Nov 2006 09:30:14 -0500
Received: from kyoto.netlab.nec.de ([195.37.70.21])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiAuy-0008BF-MX
	for nsis@ietf.org; Thu, 09 Nov 2006 09:30:14 -0500
Received: from [127.0.0.1] (kobe.netlab.nec.de [195.37.70.60])
	by kyoto.netlab.nec.de (Postfix) with ESMTP id 42FB713CF82
	for <nsis@ietf.org>; Thu,  9 Nov 2006 15:32:40 +0100 (CET)
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Transfer-Encoding: 7bit
Message-Id: <51256569-AF34-419C-A4EF-0D605871E0D6@netlab.nec.de>
Content-Type: text/plain; charset=US-ASCII; format=flowed
To: nsis <nsis@ietf.org>
From: Martin Stiemerling <stiemerling@netlab.nec.de>
Date: Thu, 9 Nov 2006 06:30:07 -0800
X-Mailer: Apple Mail (2.752.2)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6d62ab47271805379d7172ee693a45db
Subject: [NSIS] Slides for session available
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

here

https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=67

   Martin

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 11:26:30 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiCj0-00074x-TT; Thu, 09 Nov 2006 11:25:58 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiCiz-00074s-Rl
	for nsis@ietf.org; Thu, 09 Nov 2006 11:25:57 -0500
Received: from mgw-ext14.nokia.com ([131.228.20.173])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiCiy-0003eJ-AE
	for nsis@ietf.org; Thu, 09 Nov 2006 11:25:57 -0500
Received: from esebh108.NOE.Nokia.com (esebh108.ntc.nokia.com [172.21.143.145])
	by mgw-ext14.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kA9GPn0S020000; Thu, 9 Nov 2006 18:25:55 +0200
Received: from esebh103.NOE.Nokia.com ([172.21.143.33]) by
	esebh108.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 9 Nov 2006 18:25:52 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh103.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 9 Nov 2006 18:25:52 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Thu, 9 Nov 2006 18:25:52 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED05C6@esebe199.NOE.Nokia.com>
In-Reply-To: <51256569-AF34-419C-A4EF-0D605871E0D6@netlab.nec.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: New, Improved slides for session available
Thread-Index: AccEDJ5KEe8wME/FTOGTmd+19wdKAgADvqxA
From: <john.loughney@nokia.com>
To: <stiemerling@netlab.nec.de>, <nsis@ietf.org>
X-OriginalArrivalTime: 09 Nov 2006 16:25:52.0230 (UTC)
	FILETIME=[B9454060:01C7041B]
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 6d62ab47271805379d7172ee693a45db
Cc: 
Subject: [NSIS] New, Improved slides for session available
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

More slides are available.

https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=3D6=
7


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 13:37:21 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiElD-00079T-9Y; Thu, 09 Nov 2006 13:36:23 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiElB-00078j-Ul
	for nsis@ietf.org; Thu, 09 Nov 2006 13:36:21 -0500
Received: from difra-computer.de ([81.169.157.103] helo=www.difra-computer.de)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiEl9-0002sl-L5
	for nsis@ietf.org; Thu, 09 Nov 2006 13:36:21 -0500
Received: from win2003.dickmann2003.home (p548A536B.dip.t-dialin.net
	[84.138.83.107])
	by www.difra-computer.de (Postfix) with ESMTP id 26C883E402C
	for <nsis@ietf.org>; Thu,  9 Nov 2006 19:34:26 +0100 (CET)
Received: from [127.0.0.1] ([192.168.28.2]) by win2003.dickmann2003.home with
	Microsoft SMTPSVC(6.0.3790.1830); Thu, 9 Nov 2006 19:24:42 +0100
Message-ID: <455374E0.2080007@christian-dickmann.de>
Date: Thu, 09 Nov 2006 13:35:12 -0500
From: Christian Dickmann <mail@christian-dickmann.de>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: nsis@ietf.org
Content-Type: text/plain; charset=ISO-8859-15; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 09 Nov 2006 18:24:42.0518 (UTC)
	FILETIME=[5340DB60:01C7042C]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: cf4fa59384e76e63313391b70cd0dd25
Subject: [NSIS] GIST Issue 157
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Robert,

I just read issue 157 
(http://nsis.srmr.co.uk/cgi-bin/roundup.cgi/nsis-ntlp-issues/issue157). 
The text says "If sent with R=0 and null Hello-ID, it indicates that a 
node wishes to keep a messaging association open.", so a ID=0 resets the 
MA keep-alive timer. However, in the non-zero ID the text does not 
explicitly state whether the keep-alive timer is reset or not. Can you 
clarify this?
The question is: If I want to keep the MA open AND test the liveness do 
I have to send both types of MA-Hello?

Regards,
Christian Dickmann



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 14:00:40 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiF80-0003yx-LS; Thu, 09 Nov 2006 13:59:56 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiF7z-0003yG-Mz
	for nsis@ietf.org; Thu, 09 Nov 2006 13:59:55 -0500
Received: from sj-iport-5.cisco.com ([171.68.10.87])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiF7y-0006c0-A1
	for nsis@ietf.org; Thu, 09 Nov 2006 13:59:55 -0500
Received: from sj-dkim-4.cisco.com ([171.71.179.196])
	by sj-iport-5.cisco.com with ESMTP; 09 Nov 2006 10:59:52 -0800
Received: from sj-core-1.cisco.com (sj-core-1.cisco.com [171.71.177.237])
	by sj-dkim-4.cisco.com (8.12.11/8.12.11) with ESMTP id kA9IxqT7023896
	for <nsis@ietf.org>; Thu, 9 Nov 2006 10:59:52 -0800
Received: from imail.cisco.com (sjc12-sbr-sw3-3f5.cisco.com [172.19.96.182])
	by sj-core-1.cisco.com (8.12.10/8.12.6) with ESMTP id kA9Ixqio019081
	for <nsis@ietf.org>; Thu, 9 Nov 2006 10:59:52 -0800 (PST)
Received: from [130.129.69.123] ([10.21.88.85])
	by imail.cisco.com (8.12.11/8.12.10) with ESMTP id kA9IkPsq027199
	for <nsis@ietf.org>; Thu, 9 Nov 2006 10:46:27 -0800
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Transfer-Encoding: 7bit
Message-Id: <5E7777BA-D9C1-4E4E-90E1-187F09A6236B@cisco.com>
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
To: NSIS Group <nsis@ietf.org>
From: David R Oran <oran@cisco.com>
Date: Thu, 9 Nov 2006 13:59:40 -0500
X-Mailer: Apple Mail (2.752.2)
DKIM-Signature: a=rsa-sha1; q=dns; l=2134; t=1163098792; x=1163962792;
	c=relaxed/simple; s=sjdkim4002;
	h=Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version;
	d=cisco.com; i=oran@cisco.com;
	z=From:David=20R=20Oran=20<oran@cisco.com>
	|Subject:Alternatives=20for=20supporting=20domain-local=20admission=20
	control=20operation=20in=20NSLP/QSPEC |Sender:;
	X=v=3Dcisco.com=3B=20h=3DAsnLtv42gftunNOCS9kWkQ811yQ=3D;
	b=lKNqOFrFUuzKUp2cYVYVWaGm+lk7Ol6no6pT4YWh82zMeUiQC5mRHYqNtVAH2Ec0wYip8RIP
	zMwUdZvjoRj8etBOmDTvVDDePKhf8yOQMFHEX972sY7HMELaYl9oDCEk;
DKIM-Signature: v=0.5; a=rsa-sha256; q=dns/txt; l=2134; t=1163097990;
	x=1163961990; c=relaxed/simple; s=oregon;
	h=To:Content-Type:From:Subject:Content-Transfer-Encoding:MIME-Version; 
	d=cisco.com; i=oran@cisco.com;
	z=From:=20David=20R=20Oran=20<oran@cisco.com>
	|Subject:=20Alternatives=20for=20supporting=20domain-local=20admission=20
	control=20operation=20in=20NSLP/QSPEC |Sender:=20
	|To:=20NSIS=20Group=20<nsis@ietf.org>;
	bh=v446qKTvW64mNoJuHe9rLKEPRVVpbloHWIcscB4oVsA=;
	b=ZXg8GXZLpg4e5M5Jyc91JA0xPEc7QQhGDP1n1HvJt6EFvY+FL+jvTrtefc2fxf4cdcA0A6tm
	spNVK1wFreayRdcdMhvzM3J9jXTXJaL0yra3Qlli68jJAs2RqDwp7hS2;
Authentication-Results: sj-dkim-4; header.From=oran@cisco.com; dkim=pass (si
	g from cisco.com/sjdkim4002 verified; ); 
	header.From=oran@cisco.com; dkim=pass (
	sig from cisco.com/oregon verified; ); 
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 21c69d3cfc2dd19218717dbe1d974352
Subject: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Some of this I raised in a private exchange with the QSPEC authors as  
we were polishing our proposal for the QSPEC redesign. John Loughney  
asked me to clean it up and send it to the whole list, so here it is.

There seems to be a strong working group consensus to support the  
ability to have the edge-nodes at entry/exit of a domain perform a  
"full" NSLP-style admission control function, based on the standard  
QSPEC, yet allow the interior nodes to have both a simpler/customized  
data plane, *and* a simpler admission control capability in their RMF  
which is based on either a subset of the QSPEC parameters, or some  
private parameters, or both.

When Jerry/Cornelia/Dave proposed eliminating stacking, this was the  
piece of functionality on which we got strong opposition, so that  
indicates the WG consensus I noted above.

In thinking about what protocol designs could meet the above  
requirements, I discerned three. One is essentially identical to the  
former/existing QSPEC stacking functionality we wanted to eliminate.  
A second is a proposal by Roland Bless. The third is to not handle  
this in QSPECS as all, but by recursing at the NSLP level.

In more detail, the three alternatives are:

1. Do complete QSPEC stacking, and deal with the issues around what  
constraints have to be met by the stacked QSPEC, whether the QSPEC  
lower in the "stack" is opacified, and what the rules are for  
pushing, swapping, and popping.

2. Allow domain-parameters to be added to a QSPEC on domain entry and  
removed on exit. In this option one has to describe whether anything  
is opacified and which existing QSPEC parameters get "re-interpreted"  
if any.

3. Recurse at the NSLP level, have a separate domain-private NSLP  
session, and that would contain a domain-private set of paramteres  
instead of the end-to-end QSPEC. I'll note that this approach allows  
(but not requires) the two signaling state machines (end-end and edge- 
edge through interior nodes) to be decoupled.

We already have (3) for other purposes, like aggregation and tunneling.

Dave.


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 15:45:06 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiGl8-0002aY-97; Thu, 09 Nov 2006 15:44:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiGl6-0002ZJ-Ik
	for nsis@ietf.org; Thu, 09 Nov 2006 15:44:24 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiGl4-00079G-QS
	for nsis@ietf.org; Thu, 09 Nov 2006 15:44:24 -0500
Received: from sbz-30.cs.helsinki.fi (sbz-30.cs.helsinki.fi [128.214.9.98])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Thu, 09 Nov 2006 22:39:12 +0200
	id 000AFDA1.455391F0.00001069
Date: Thu, 9 Nov 2006 22:39:11 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: nsis@ietf.org
Message-ID: <Pine.LNX.4.64.0611092120100.20021@sbz-30.cs.Helsinki.FI>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 9466e0365fc95844abaf7c3f15a05c7d
Subject: [NSIS] QoS NSLP: summary object
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Based on John's request, here is recap of the summary object mentioned 
today. After Roland's comment on the mic, there are actually two use 
cases:

1. Allow a QNE to refresh several sessions in one NSLP refreshing RESERVE 
message. 

2. Allow a ANE to tear down several sessions in one NSLP tearing 
RESERVE message.

Both of these save bytes and processing compared to doing it independently 
for each session. They can be implemented outside GIST by keeping track of 
the SII_HANDLE provided by GIST (however, you can only know the handle 
after you have received a message from the per, e.g., a RESPONSE). If a 
QNE has sessions whose next peer has the same SII_HANDLE, it can use a new 
SESSION_LIST where it indicates SID+RSN pairs of sessions it wants to 
refresh or tear down. The receiving QNE, in case of unknown SID, will 
reply with a RESPONSE carrying the same object to indicate which session 
references where not understood. If a confirmation was requested from the 
QNR for a tear, and it must send an individual RESPONSE for each 
terminated session, as usual - the summary refresh could hit a QNR with 
multiple sessions being terminated.

We would like to add this to the NSLP spec. Any comments, arguments, 
feedback? I'll wait for a week, then add this unless I hear arguments 
against.

Jukka

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 16:31:44 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiHUJ-00012H-8v; Thu, 09 Nov 2006 16:31:07 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiHUI-00012B-H9
	for nsis@ietf.org; Thu, 09 Nov 2006 16:31:06 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiHUH-0006xo-3Q
	for nsis@ietf.org; Thu, 09 Nov 2006 16:31:06 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GiHU9-0001H8-L6; Thu, 09 Nov 2006 22:30:58 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GiHU6-0000IT-2k; Thu, 09 Nov 2006 22:30:54 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 613C7B604;
	Thu,  9 Nov 2006 22:30:53 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GiHU4-00050B-Q9; Thu, 09 Nov 2006 22:30:53 +0100
Message-ID: <45539D8E.70906@tm.uka.de>
Date: Thu, 09 Nov 2006 22:28:46 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.8) Gecko/20061025 Thunderbird/1.5.0.8 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
Subject: Re: [NSIS] QoS NSLP: summary object
References: <Pine.LNX.4.64.0611092120100.20021@sbz-30.cs.Helsinki.FI>
In-Reply-To: <Pine.LNX.4.64.0611092120100.20021@sbz-30.cs.Helsinki.FI>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.1 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.3 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 50a516d93fd399dc60588708fd9a3002
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jukka and all,

Jukka MJ Manner wrote:
> Based on John's request, here is recap of the summary object mentioned 
> today. After Roland's comment on the mic, there are actually two use 
> cases:
> 
> 1. Allow a QNE to refresh several sessions in one NSLP refreshing RESERVE 
> message. 
> 
> 2. Allow a ANE to tear down several sessions in one NSLP tearing 
> RESERVE message.
> 
> Both of these save bytes and processing compared to doing it independently 
> for each session. They can be implemented outside GIST by keeping track of 
> the SII_HANDLE provided by GIST (however, you can only know the handle 
> after you have received a message from the per, e.g., a RESPONSE). If a 

Not quite. After you received a GIST response. GIST will report it
upwards to the NSLP. However, you care about refresh or tear only
after the initial setup worked. So I don't see any problems here.
Furthermore, it is an optimization, so basically no problem if it
is not applicable in some rare cases.

> QNE has sessions whose next peer has the same SII_HANDLE, it can use a new 
> SESSION_LIST where it indicates SID+RSN pairs of sessions it wants to 
> refresh or tear down. The receiving QNE, in case of unknown SID, will 

Yes, and as we just discussed in private, it would be good to have
an SID-List only object (SIDs w/o RSN), too.

> reply with a RESPONSE carrying the same object to indicate which session 
> references where not understood. If a confirmation was requested from the 
> QNR for a tear, and it must send an individual RESPONSE for each 
> terminated session, as usual - the summary refresh could hit a QNR with 
> multiple sessions being terminated.

In this case we would get a summary error back.

> We would like to add this to the NSLP spec. Any comments, arguments, 
> feedback? I'll wait for a week, then add this unless I hear arguments 
> against.

Obviously I'm in favor of having that. However, we must be careful
about covering the new error cases and handle them properly.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 16:48:55 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiHkh-0000GM-TC; Thu, 09 Nov 2006 16:48:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiHkg-0000Fp-V0
	for nsis@ietf.org; Thu, 09 Nov 2006 16:48:02 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiHkf-0000c2-Dd
	for nsis@ietf.org; Thu, 09 Nov 2006 16:48:02 -0500
Received: from sbz-30.cs.helsinki.fi (sbz-30.cs.helsinki.fi [128.214.9.98])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Thu, 09 Nov 2006 23:47:56 +0200
	id 000AFE04.4553A20C.00002B2B
Date: Thu, 9 Nov 2006 23:47:54 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: Roland Bless <bless@tm.uka.de>
Subject: Re: [NSIS] QoS NSLP: summary object
In-Reply-To: <45539D8E.70906@tm.uka.de>
Message-ID: <Pine.LNX.4.64.0611092342100.23234@sbz-30.cs.Helsinki.FI>
References: <Pine.LNX.4.64.0611092120100.20021@sbz-30.cs.Helsinki.FI>
	<45539D8E.70906@tm.uka.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4b800b1eab964a31702fa68f1ff0e955
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi,
two comments back. :)

Jukka

On Thu, 9 Nov 2006, Roland Bless wrote:

> Hi Jukka and all,
> 
> Jukka MJ Manner wrote:
> > Based on John's request, here is recap of the summary object mentioned 
> > today. After Roland's comment on the mic, there are actually two use 
> > cases:
> > 
> > 1. Allow a QNE to refresh several sessions in one NSLP refreshing RESERVE 
> > message. 
> > 
> > 2. Allow a ANE to tear down several sessions in one NSLP tearing 
> > RESERVE message.
> > 
> > Both of these save bytes and processing compared to doing it independently 
> > for each session. They can be implemented outside GIST by keeping track of 
> > the SII_HANDLE provided by GIST (however, you can only know the handle 
> > after you have received a message from the per, e.g., a RESPONSE). If a 
> 
> Not quite. After you received a GIST response. GIST will report it
> upwards to the NSLP. However, you care about refresh or tear only
> after the initial setup worked. So I don't see any problems here.
> Furthermore, it is an optimization, so basically no problem if it
> is not applicable in some rare cases.

[1] This note comes from looking at the GIST API. In order to send a 
summary refresh, you need to have compared the GIST SII_HANDLE before. And 
you will only get to know the SII_HANDLE for a session if you receive the 
RESPONSE. It is not a problem in the general case, just a note.

> 
> > QNE has sessions whose next peer has the same SII_HANDLE, it can use a new 
> > SESSION_LIST where it indicates SID+RSN pairs of sessions it wants to 
> > refresh or tear down. The receiving QNE, in case of unknown SID, will 
> 
> Yes, and as we just discussed in private, it would be good to have
> an SID-List only object (SIDs w/o RSN), too.

[2] Actualy... may be not. You need the SID AND the RSN for both tear and 
refresh. Thus, you need to send both anyway. RSN is a mandatory object for 
all RESERVE messages. You need to have the RSN in synch if you are tearing 
otherwise you will tear a different state than the previous QNE.

> 
> > reply with a RESPONSE carrying the same object to indicate which session 
> > references where not understood. If a confirmation was requested from the 
> > QNR for a tear, and it must send an individual RESPONSE for each 
> > terminated session, as usual - the summary refresh could hit a QNR with 
> > multiple sessions being terminated.
> 
> In this case we would get a summary error back.
> 
> > We would like to add this to the NSLP spec. Any comments, arguments, 
> > feedback? I'll wait for a week, then add this unless I hear arguments 
> > against.
> 
> Obviously I'm in favor of having that. However, we must be careful
> about covering the new error cases and handle them properly.
> 
> Regards,
>  Roland
> 
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 16:56:33 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiHsK-0005kF-FR; Thu, 09 Nov 2006 16:55:56 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiHsJ-0005jY-6P
	for nsis@ietf.org; Thu, 09 Nov 2006 16:55:55 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiHsH-00020N-M0
	for nsis@ietf.org; Thu, 09 Nov 2006 16:55:55 -0500
Received: from webmail.cs.utwente.nl (janus.ewi.utwente.nl [130.89.10.26])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with SMTP id kA9LtjfP020914;
	Thu, 9 Nov 2006 22:55:45 +0100 (MET)
Received: from 130.129.32.143 (auth. user karagian@imap2.cs.utwente.nl)
	by webmail.cs.utwente.nl with HTTP; Thu, 09 Nov 2006 21:55:44 +0000
To: "David R Oran" <oran@cisco.com>, "NSIS Group" <nsis@ietf.org>
Subject: Re: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
Date: Thu, 09 Nov 2006 21:55:44 +0000
X-Mailer: IlohaMail/0.8.13 (On: webmail.cs.utwente.nl)
Message-ID: <cKrDyA4e.1163109344.1969670.karagian@ewi.utwente.nl>
In-Reply-To: <5E7777BA-D9C1-4E4E-90E1-187F09A6236B@cisco.com>
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
Bounce-To: "Georgios Karagiannis" <karagian@cs.utwente.nl>
MIME-Version: 1.0 
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: 0.3 () FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Thu, 09 Nov 2006 22:55:46 +0100 (MET)
X-Spam-Score: 0.1 (/)
X-Scan-Signature: a8a20a483a84f747e56475e290ee868e
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all

I really do not understand this proposal.

Note that option 2 is not exactly what Roland proposed. Actually option 2
should say that a QOS model is able to craeate its own local/private
QSPEC and that tis local/private QSPEC can encapsulate the parameters of
the end-to-end QSPEC. Additionally the local/private QSPEc can then
piggiback its own local QSPEC parameters. In this case the parameters of
the end-to-end QSPEC parameters can be ignored.

Please also note that the discussion on stacking it is only related to
options 1 and 2.
Point 3 is a separate issue that is related to the possibility that is
given to QOS models to use/initiate their own session local/edge
-to-edge session. Note that NSIS should support edge-to-edge scenarios
and therefore this point should not be a point of consensus. Please also
note that the edge-toedge support has been introduced by this WG 3/4
years ago in the QOS-NSLP work. Note that the local (edge-to-edge)
scenario could map the end-to-end QSPEC traffic source description to
the local traffic description.

Thus regarding stacking what we have to decide is whether we use option 1
or option 2, i.e., Rolands proposal.

Please note that the motivation to use encapsulation/piggibacking  of
local QSPEc parameters in end-to-end QSPECs is to increase the
flexibility, modularity and the signaling performance. The flexibility
and modularity
increase can be achieved by providing the possibility to any existing
or future QoS model applied in a local QoS NSLP aware domain to use
and stack any required local QoS model information on QoS NSLP
signaling messages passing through the local domain (see requirement
5.4.2 in RFC3726).

For example, suppose that a QOS model uses parallel sessions (an
end-to-end session and an edge-to-edge (local) session), then the QOS
model can piggiback QoS model (local) specific control information into
the end-to-end RESPONSE message that is sent from an egress node to an
ingress node. If that will not be allowed than the egress node would
need to generate a completely new local RESPONSE message in order to
send the same QoS model specific control information.

Thus if I look to this proposal I think that we need need both otion 3
(parallel session support) and option 2, which is Rolands proposal that
encapsulates end-to-end QSPEC into a local QSPEC and gives the
possibility to the QoS model to specify which QSPEC parameters can be
ignored.

Best Regards,
Georgios

On 11/9/2006, "David R Oran" <oran@cisco.com> wrote:

>Some of this I raised in a private exchange with the QSPEC authors as
>we were polishing our proposal for the QSPEC redesign. John Loughney
>asked me to clean it up and send it to the whole list, so here it is.
>
>There seems to be a strong working group consensus to support the
>ability to have the edge-nodes at entry/exit of a domain perform a
>"full" NSLP-style admission control function, based on the standard
>QSPEC, yet allow the interior nodes to have both a simpler/customized
>data plane, *and* a simpler admission control capability in their RMF
>which is based on either a subset of the QSPEC parameters, or some
>private parameters, or both.
>
>When Jerry/Cornelia/Dave proposed eliminating stacking, this was the
>piece of functionality on which we got strong opposition, so that
>indicates the WG consensus I noted above.
>
>In thinking about what protocol designs could meet the above
>requirements, I discerned three. One is essentially identical to the
>former/existing QSPEC stacking functionality we wanted to eliminate.
>A second is a proposal by Roland Bless. The third is to not handle
>this in QSPECS as all, but by recursing at the NSLP level.
>
>In more detail, the three alternatives are:
>
>1. Do complete QSPEC stacking, and deal with the issues around what
>constraints have to be met by the stacked QSPEC, whether the QSPEC
>lower in the "stack" is opacified, and what the rules are for
>pushing, swapping, and popping.
>
>2. Allow domain-parameters to be added to a QSPEC on domain entry and
>removed on exit. In this option one has to describe whether anything
>is opacified and which existing QSPEC parameters get "re-interpreted"
>if any.
>
>3. Recurse at the NSLP level, have a separate domain-private NSLP
>session, and that would contain a domain-private set of paramteres
>instead of the end-to-end QSPEC. I'll note that this approach allows
>(but not requires) the two signaling state machines (end-end and edge-
>edge through interior nodes) to be decoupled.
>
>We already have (3) for other purposes, like aggregation and tunneling.
>
>Dave.
>
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 17:18:15 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiICr-0000Os-Ib; Thu, 09 Nov 2006 17:17:09 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiICp-0000Ne-M0
	for nsis@ietf.org; Thu, 09 Nov 2006 17:17:07 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiICo-0005NE-76
	for nsis@ietf.org; Thu, 09 Nov 2006 17:17:07 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GiICn-0002Bz-1Q; Thu, 09 Nov 2006 23:17:05 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GiICm-0005pa-Ks; Thu, 09 Nov 2006 23:17:04 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 0FE76B636;
	Thu,  9 Nov 2006 23:17:04 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GiICl-000520-Ck; Thu, 09 Nov 2006 23:17:03 +0100
Message-ID: <4553A860.40004@tm.uka.de>
Date: Thu, 09 Nov 2006 23:14:56 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.8) Gecko/20061025 Thunderbird/1.5.0.8 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: David R Oran <oran@cisco.com>
Subject: Re: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
References: <5E7777BA-D9C1-4E4E-90E1-187F09A6236B@cisco.com>
In-Reply-To: <5E7777BA-D9C1-4E4E-90E1-187F09A6236B@cisco.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.1 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.3 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d8ae4fd88fcaf47c1a71c804d04f413d
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Dave and all,

David R Oran wrote:
> In thinking about what protocol designs could meet the above
> requirements, I discerned three. One is essentially identical to the
> former/existing QSPEC stacking functionality we wanted to eliminate. A
> second is a proposal by Roland Bless. The third is to not handle this in
> QSPECS as all, but by recursing at the NSLP level.
> 
> In more detail, the three alternatives are:
> 
> 1. Do complete QSPEC stacking, and deal with the issues around what
> constraints have to be met by the stacked QSPEC, whether the QSPEC lower
> in the "stack" is opacified, and what the rules are for pushing,
> swapping, and popping.

> 2. Allow domain-parameters to be added to a QSPEC on domain entry and
> removed on exit. In this option one has to describe whether anything is
> opacified and which existing QSPEC parameters get "re-interpreted" if any.

It's actually not the _same_ QSPEC with added parameters,
but it would be a "new" QSPEC (type). So to give an
example in order to make it clear, an RMD-QSPEC would
look like this:
----------------------------------
| QSPEC-Type = RMD-QSPEC         |  Common QSPEC header
+================================+\
|  RMD-specific Parameter 1      | \
+--------------------------------+  \
|             ....               |   local QSPEC parameters
+--------------------------------+  /
|  RMD-specific Parameter n      | /
+--------------------------------+/
| +----------------------------+ |
| | QSPEC-Type = Generic QSPEC | |
| +============================+ |
| |                            | | Encapsulated "external" QSPEC
| |          ....              | |
| +----------------------------+ |
+--------------------------------+
The important difference to 1 is that QoS-NSLP only sees
and passes one QSPEC up to the RMF. The type of the
QSPEC thus may change within a domain. As written in my earlier
mail:
1. QNI signals its QoS requirements with its Generic QSPEC
2. The ingress edge QNE will map these to its own QoS model, e.g.
   choosing the corresponding DiffServ or Y.1541 class.
   The mapped parameter values may be carried within
   the above described local RMD-QSPEC.
3. The egress QNE will process the RMD-QSPEC and pass only
    the encapsulated Generic QSPEC to the next domain, that
    may perform its own mapping again possibly without requiring
    to transmit its own local QSPEC.

> 3. Recurse at the NSLP level, have a separate domain-private NSLP
> session, and that would contain a domain-private set of paramteres
> instead of the end-to-end QSPEC. I'll note that this approach allows
> (but not requires) the two signaling state machines (end-end and
> edge-edge through interior nodes) to be decoupled.

> We already have (3) for other purposes, like aggregation and tunneling.

But Georgios point was probably more to avoid having a
separate signaling and to piggyback additional information instead.
So this would be an optimization, not possible with approach 3 only.

My view is the following:
1 is more complicated than 2, at least for the QoS NSLP.
2 could be used in addition to 3.
I'm really open to hear cases were approach 2 fails badly
or is disadvantageous over approach 1.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 17:27:49 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiIMY-0000AP-N1; Thu, 09 Nov 2006 17:27:10 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiIMW-0000AJ-Q4
	for nsis@ietf.org; Thu, 09 Nov 2006 17:27:08 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiIMV-0006Ls-Cz
	for nsis@ietf.org; Thu, 09 Nov 2006 17:27:08 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GiIMU-00042o-09; Thu, 09 Nov 2006 23:27:06 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GiIMT-0006yz-NC; Thu, 09 Nov 2006 23:27:05 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 4A1EFB61E;
	Thu,  9 Nov 2006 23:27:05 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GiIMS-00052L-M2; Thu, 09 Nov 2006 23:27:05 +0100
Message-ID: <4553AABA.7020809@tm.uka.de>
Date: Thu, 09 Nov 2006 23:24:58 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.8) Gecko/20061025 Thunderbird/1.5.0.8 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
Subject: Re: [NSIS] QoS NSLP: summary object
References: <Pine.LNX.4.64.0611092120100.20021@sbz-30.cs.Helsinki.FI>
	<45539D8E.70906@tm.uka.de>
	<Pine.LNX.4.64.0611092342100.23234@sbz-30.cs.Helsinki.FI>
In-Reply-To: <Pine.LNX.4.64.0611092342100.23234@sbz-30.cs.Helsinki.FI>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.1 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.3 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8abaac9e10c826e8252866cbe6766464
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jukka,

Jukka MJ Manner wrote:
> [1] This note comes from looking at the GIST API. In order to send a 
> summary refresh, you need to have compared the GIST SII_HANDLE before. And 
> you will only get to know the SII_HANDLE for a session if you receive the 
> RESPONSE. It is not a problem in the general case, just a note.

That is not correct. You get the SII-Handle from GIST
as soon you got the _GIST_ Response back (NOT the QoS NSLP
RESPONSE) from the other peer (you should get a NetworkNotification
that passes the SII-handle). However,
you'll do summary refreshes/tears only for established sessions
anyway, so you'll have the SII-handle in all cases.

>>> QNE has sessions whose next peer has the same SII_HANDLE, it can use a new 
>>> SESSION_LIST where it indicates SID+RSN pairs of sessions it wants to 
>>> refresh or tear down. The receiving QNE, in case of unknown SID, will 
>> Yes, and as we just discussed in private, it would be good to have
>> an SID-List only object (SIDs w/o RSN), too.
> 
> [2] Actualy... may be not. You need the SID AND the RSN for both tear and 
> refresh. Thus, you need to send both anyway. RSN is a mandatory object for 
> all RESERVE messages. You need to have the RSN in synch if you are tearing 
> otherwise you will tear a different state than the previous QNE.

My thinking was that if you are tearing the session you maybe don't
care about it's current state anymore?

Cheers (you got the beer),
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 17:49:47 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiIhr-0002Au-Rn; Thu, 09 Nov 2006 17:49:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiIhq-00027J-H5
	for nsis@ietf.org; Thu, 09 Nov 2006 17:49:10 -0500
Received: from mail121.messagelabs.com ([216.82.241.195])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GiIhp-0000ht-70
	for nsis@ietf.org; Thu, 09 Nov 2006 17:49:10 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-12.tower-121.messagelabs.com!1163112547!10249933!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 380 invoked from network); 9 Nov 2006 22:49:07 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-12.tower-121.messagelabs.com with SMTP;
	9 Nov 2006 22:49:07 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kA9Mn7Rf016988; 
	Thu, 9 Nov 2006 17:49:07 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kA9Mn4LQ016978; Thu, 9 Nov 2006 17:49:04 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Date: Thu, 9 Nov 2006 16:49:03 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F62@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <4553A860.40004@tm.uka.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Thread-Index: AccETTBRKwUymAEBRBukYt/ILQ8d+wAA9+8Q
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: "Roland Bless" <bless@tm.uka.de>, "David R Oran" <oran@cisco.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a87a9cdae4ac5d3fbeee75cd0026d632
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland,

The picture for option 2 is very helpful.  Could you send an example
picture for option 1 for comparison?

Thanks,
Jerry=20

-----Original Message-----
From: Roland Bless [mailto:bless@tm.uka.de]=20
Sent: Thursday, November 09, 2006 5:15 PM
To: David R Oran
Cc: NSIS Group
Subject: Re: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

Hi Dave and all,

David R Oran wrote:
> In thinking about what protocol designs could meet the above
> requirements, I discerned three. One is essentially identical to the
> former/existing QSPEC stacking functionality we wanted to eliminate. A
> second is a proposal by Roland Bless. The third is to not handle this
in
> QSPECS as all, but by recursing at the NSLP level.
>=20
> In more detail, the three alternatives are:
>=20
> 1. Do complete QSPEC stacking, and deal with the issues around what
> constraints have to be met by the stacked QSPEC, whether the QSPEC
lower
> in the "stack" is opacified, and what the rules are for pushing,
> swapping, and popping.

> 2. Allow domain-parameters to be added to a QSPEC on domain entry and
> removed on exit. In this option one has to describe whether anything
is
> opacified and which existing QSPEC parameters get "re-interpreted" if
any.

It's actually not the _same_ QSPEC with added parameters,
but it would be a "new" QSPEC (type). So to give an
example in order to make it clear, an RMD-QSPEC would
look like this:
----------------------------------
| QSPEC-Type =3D RMD-QSPEC         |  Common QSPEC header
+=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D+\
|  RMD-specific Parameter 1      | \
+--------------------------------+  \
|             ....               |   local QSPEC parameters
+--------------------------------+  /
|  RMD-specific Parameter n      | /
+--------------------------------+/
| +----------------------------+ |
| | QSPEC-Type =3D Generic QSPEC | |
| =
+=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D+ |
| |                            | | Encapsulated "external" QSPEC
| |          ....              | |
| +----------------------------+ |
+--------------------------------+
The important difference to 1 is that QoS-NSLP only sees
and passes one QSPEC up to the RMF. The type of the
QSPEC thus may change within a domain. As written in my earlier
mail:
1. QNI signals its QoS requirements with its Generic QSPEC
2. The ingress edge QNE will map these to its own QoS model, e.g.
   choosing the corresponding DiffServ or Y.1541 class.
   The mapped parameter values may be carried within
   the above described local RMD-QSPEC.
3. The egress QNE will process the RMD-QSPEC and pass only
    the encapsulated Generic QSPEC to the next domain, that
    may perform its own mapping again possibly without requiring
    to transmit its own local QSPEC.

> 3. Recurse at the NSLP level, have a separate domain-private NSLP
> session, and that would contain a domain-private set of paramteres
> instead of the end-to-end QSPEC. I'll note that this approach allows
> (but not requires) the two signaling state machines (end-end and
> edge-edge through interior nodes) to be decoupled.

> We already have (3) for other purposes, like aggregation and
tunneling.

But Georgios point was probably more to avoid having a
separate signaling and to piggyback additional information instead.
So this would be an optimization, not possible with approach 3 only.

My view is the following:
1 is more complicated than 2, at least for the QoS NSLP.
2 could be used in addition to 3.
I'm really open to hear cases were approach 2 fails badly
or is disadvantageous over approach 1.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 18:20:23 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiJBB-0003KB-Ob; Thu, 09 Nov 2006 18:19:29 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiJBA-0003Jx-JC
	for nsis@ietf.org; Thu, 09 Nov 2006 18:19:28 -0500
Received: from smtp.polymtl.ca ([132.207.4.11])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiJB9-00049R-8m
	for nsis@ietf.org; Thu, 09 Nov 2006 18:19:28 -0500
Received: from localhost (imp-4-2.polymtl.ca [132.207.4.77])
	by smtp.polymtl.ca (8.13.6/8.13.6) with ESMTP id kA9NJJG0016575;
	Thu, 9 Nov 2006 18:19:19 -0500
Received: from dhcp35-130.ietf67.org (dhcp35-130.ietf67.org [130.129.35.130]) 
	by www.imp.polymtl.ca (IMP) with HTTP 
	for <stouea@mailhost.info.polymtl.ca>; Thu, 09 Nov 2006 18:16:35 -0500
Message-ID: <1163114195.4553b6d30af2b@www.imp.polymtl.ca>
Date: Thu, 09 Nov 2006 18:16:35 -0500
From: Stephane Ouellette <stouea@info.polymtl.ca>
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
References: <cKrDyA4e.1163109344.1969670.karagian@ewi.utwente.nl>
In-Reply-To: <cKrDyA4e.1163109344.1969670.karagian@ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
User-Agent: Internet Messaging Program (IMP) 3.2.3
X-Originating-IP: 130.129.35.130
X-Poly-FromMTA: (imp-4-2.polymtl.ca [132.207.4.77]) at Thu,
	9 Nov 2006 23:19:19 +0000
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 14582b0692e7f70ce7111d04db3781c8
Cc: David R Oran <oran@cisco.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
Reply-To: stephane.ouellette@polymtl.ca
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios, David and all,

  see my comments inline.

Stephane.


Quoting Georgios Karagiannis <karagian@cs.utwente.nl>:

> Hi all
>
> I really do not understand this proposal.
>
> Note that option 2 is not exactly what Roland proposed. Actually option 2
> should say that a QOS model is able to craeate its own local/private
> QSPEC and that tis local/private QSPEC can encapsulate the parameters of
> the end-to-end QSPEC. Additionally the local/private QSPEc can then
> piggiback its own local QSPEC parameters. In this case the parameters of
> the end-to-end QSPEC parameters can be ignored.
>


[STEPHANE] I read Roland's proposal and its description by Georgios is exactly
what I understood.

Roland's proposal can be found here:

http://www1.ietf.org/mail-archive/web/nsis/current/msg07158.html


> Please also note that the discussion on stacking it is only related to
> options 1 and 2.
> Point 3 is a separate issue that is related to the possibility that is
> given to QOS models to use/initiate their own session local/edge
> -to-edge session. Note that NSIS should support edge-to-edge scenarios
> and therefore this point should not be a point of consensus. Please also
> note that the edge-toedge support has been introduced by this WG 3/4
> years ago in the QOS-NSLP work. Note that the local (edge-to-edge)
> scenario could map the end-to-end QSPEC traffic source description to
> the local traffic description.
>
> Thus regarding stacking what we have to decide is whether we use option 1
> or option 2, i.e., Rolands proposal.
>

[STEPHANE] I also think that recursion is out of scope of this debate, as it is
always possible to create a NSIS session between the edge nodes of a domain to
emulate local QSPEC stacking. However, recursion increases the number of states
that must be managed by the edge nodes.


> Please note that the motivation to use encapsulation/piggibacking  of
> local QSPEc parameters in end-to-end QSPECs is to increase the
> flexibility, modularity and the signaling performance. The flexibility
> and modularity
> increase can be achieved by providing the possibility to any existing
> or future QoS model applied in a local QoS NSLP aware domain to use
> and stack any required local QoS model information on QoS NSLP
> signaling messages passing through the local domain (see requirement
> 5.4.2 in RFC3726).
>

[STEPHANE] I think that what the RFC states is unambiguous: reuse the existing
session to avoid creating unnecessary states in the edge nodes.


> For example, suppose that a QOS model uses parallel sessions (an
> end-to-end session and an edge-to-edge (local) session), then the QOS
> model can piggiback QoS model (local) specific control information into
> the end-to-end RESPONSE message that is sent from an egress node to an
> ingress node. If that will not be allowed than the egress node would
> need to generate a completely new local RESPONSE message in order to
> send the same QoS model specific control information.
>

[STEPHANE] That's right.


> Thus if I look to this proposal I think that we need need both otion 3
> (parallel session support) and option 2, which is Rolands proposal that
> encapsulates end-to-end QSPEC into a local QSPEC and gives the
> possibility to the QoS model to specify which QSPEC parameters can be
> ignored.
>
> Best Regards,
> Georgios
>

[STEPHANE]
I would vote for a combination of Roland's proposal and option 3.

Thanks for your attention,

Stephane.



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 18:52:49 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiJgk-0006UL-EX; Thu, 09 Nov 2006 18:52:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiJgj-0006UG-PZ
	for nsis@ietf.org; Thu, 09 Nov 2006 18:52:05 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiJgd-0007q8-GE
	for nsis@ietf.org; Thu, 09 Nov 2006 18:52:05 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GiJgL-000358-QH; Fri, 10 Nov 2006 00:51:48 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GiJgL-0001G5-BW; Fri, 10 Nov 2006 00:51:41 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id C7EC8B636;
	Fri, 10 Nov 2006 00:51:40 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GiJgK-00055Z-42; Fri, 10 Nov 2006 00:51:40 +0100
Message-ID: <4553BE8E.9080003@tm.uka.de>
Date: Fri, 10 Nov 2006 00:49:34 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.8) Gecko/20061025 Thunderbird/1.5.0.8 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Subject: Re: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F62@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F62@KCCLUST06EVS1.ugd.att.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Spam-Score: -4.1 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.3 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2beba50d0fcdeee5f091c59f204d4365
Cc: David R Oran <oran@cisco.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi,

Ash, Gerald R (Jerry), ALABS wrote:
> The picture for option 2 is very helpful.  Could you send an example
> picture for option 1 for comparison?

Sure. I thought that the backers of the stack proposal would have to do
it :-)
Stacking:
+--------------------------------+
| QSPEC-Type = RMD-QSPEC         |\
+================================+ \
|  RMD-specific Parameter 1      |  \
+--------------------------------+   1) RMD-QSPEC
|             ....               |  /
+--------------------------------+ /
|  RMD-specific Parameter n      |/
+--------------------------------+
+--------------------------------+\
| QSPEC-Type = Generic QSPEC     | \
+================================+  \
|                                |   2) Generic QSPEC
|             ....               |  /
|                                | /
+--------------------------------+/

> Thanks,
> Jerry 
> 
> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: Thursday, November 09, 2006 5:15 PM
> To: David R Oran
> Cc: NSIS Group
> Subject: Re: [NSIS] Alternatives for supporting domain-local admission
> controloperation in NSLP/QSPEC
> 
> Hi Dave and all,
> 
> David R Oran wrote:
>> In thinking about what protocol designs could meet the above
>> requirements, I discerned three. One is essentially identical to the
>> former/existing QSPEC stacking functionality we wanted to eliminate. A
>> second is a proposal by Roland Bless. The third is to not handle this
> in
>> QSPECS as all, but by recursing at the NSLP level.
>>
>> In more detail, the three alternatives are:
>>
>> 1. Do complete QSPEC stacking, and deal with the issues around what
>> constraints have to be met by the stacked QSPEC, whether the QSPEC
> lower
>> in the "stack" is opacified, and what the rules are for pushing,
>> swapping, and popping.
> 
>> 2. Allow domain-parameters to be added to a QSPEC on domain entry and
>> removed on exit. In this option one has to describe whether anything
> is
>> opacified and which existing QSPEC parameters get "re-interpreted" if
> any.
> 
> It's actually not the _same_ QSPEC with added parameters,
> but it would be a "new" QSPEC (type). So to give an
> example in order to make it clear, an RMD-QSPEC would
> look like this:
> ----------------------------------
> | QSPEC-Type = RMD-QSPEC         |  Common QSPEC header
> +================================+\
> |  RMD-specific Parameter 1      | \
> +--------------------------------+  \
> |             ....               |   local QSPEC parameters
> +--------------------------------+  /
> |  RMD-specific Parameter n      | /
> +--------------------------------+/
> | +----------------------------+ |
> | | QSPEC-Type = Generic QSPEC | |
> | +============================+ |
> | |                            | | Encapsulated "external" QSPEC
> | |          ....              | |
> | +----------------------------+ |
> +--------------------------------+
> The important difference to 1 is that QoS-NSLP only sees
> and passes one QSPEC up to the RMF. The type of the
> QSPEC thus may change within a domain. As written in my earlier
> mail:
> 1. QNI signals its QoS requirements with its Generic QSPEC
> 2. The ingress edge QNE will map these to its own QoS model, e.g.
>    choosing the corresponding DiffServ or Y.1541 class.
>    The mapped parameter values may be carried within
>    the above described local RMD-QSPEC.
> 3. The egress QNE will process the RMD-QSPEC and pass only
>     the encapsulated Generic QSPEC to the next domain, that
>     may perform its own mapping again possibly without requiring
>     to transmit its own local QSPEC.
> 
>> 3. Recurse at the NSLP level, have a separate domain-private NSLP
>> session, and that would contain a domain-private set of paramteres
>> instead of the end-to-end QSPEC. I'll note that this approach allows
>> (but not requires) the two signaling state machines (end-end and
>> edge-edge through interior nodes) to be decoupled.
> 
>> We already have (3) for other purposes, like aggregation and
> tunneling.
> 
> But Georgios point was probably more to avoid having a
> separate signaling and to piggyback additional information instead.
> So this would be an optimization, not possible with approach 3 only.
> 
> My view is the following:
> 1 is more complicated than 2, at least for the QoS NSLP.
> 2 could be used in addition to 3.
> I'm really open to hear cases were approach 2 fails badly
> or is disadvantageous over approach 1.
> 
> Regards,
>  Roland
> 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 


-- 
Dr. Roland Bless -- WWW: http://www.tm.uka.de/~bless
Institut für Telematik, Universität Karlsruhe (TH), Germany
Zirkel 2, D-76128 Karlsruhe -- Geb. 20.20, 3.OG, Raum 358
Tel.: +49 721 608-6413 Fax: +49 721 608-6789

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 21:01:22 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiLgq-0005Jy-WF; Thu, 09 Nov 2006 21:00:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiLgp-0005Gc-1z
	for nsis@ietf.org; Thu, 09 Nov 2006 21:00:19 -0500
Received: from smtp.mei.co.jp ([133.183.129.25])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiLZf-00070F-PQ
	for nsis@ietf.org; Thu, 09 Nov 2006 20:52:57 -0500
Received: from mail-gw.jp.panasonic.com (dodgers.mei.co.jp [157.8.1.150])
	by smtp.mei.co.jp (8.12.11.20060614/3.7W/jazz) with ESMTP id
	kAA1qifS005151; Fri, 10 Nov 2006 10:52:44 +0900 (JST)
Received: by mail-gw.jp.panasonic.com (8.11.6p2/3.7W/somlx1) with ESMTP id
	kAA1qkD12229; Fri, 10 Nov 2006 10:52:46 +0900 (JST)
Received: from epochmail.jp.panasonic.com (localhost [127.0.0.1])
	by mail.jp.panasonic.com (8.11.6p2/3.7W/astros) with ESMTP id
	kAA1qkj04185; Fri, 10 Nov 2006 10:52:46 +0900 (JST)
Received: by epochmail.jp.panasonic.com (8.12.11.20060308/3.7W/soml22) id
	kAA1qjed002671; Fri, 10 Nov 2006 10:52:45 +0900 (JST)
Received: from [10.179.71.70]
	by soml22.jp.panasonic.com (8.12.11.20060308/3.7W) with ESMTP id
	kAA1qh7n002613; Fri, 10 Nov 2006 10:52:43 +0900 (JST)
Date: Fri, 10 Nov 2006 10:52:44 +0900
From: Takako Sanda <sanda.takako@jp.panasonic.com>
To: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
Subject: Re: [NSIS] QoS NSLP: summary object
In-Reply-To: <Pine.LNX.4.64.0611092120100.20021@sbz-30.cs.Helsinki.FI>
References: <Pine.LNX.4.64.0611092120100.20021@sbz-30.cs.Helsinki.FI>
Message-Id: <20061110101541.18C9.SANDA.TAKAKO@jp.panasonic.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
X-Mailer: Becky! ver. 2.20 [ja]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 82c9bddb247d9ba4471160a9a865a5f3
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jukka,

On Thu, 09 Nov 2006 22:39:11 +0200 (EET)
Jukka MJ Manner <jmanner@cs.Helsinki.FI> wrote:

> 
> Based on John's request, here is recap of the summary object mentioned 
> today. After Roland's comment on the mic, there are actually two use 
> cases:
> 
> 1. Allow a QNE to refresh several sessions in one NSLP refreshing RESERVE 
> message. 
> 
> 2. Allow a ANE to tear down several sessions in one NSLP tearing 
> RESERVE message.
>

I agree to having these use cases.

But I am wondering how QNE decides to send summary refresh/teardown.

For example, in aggregation scenario, there may be the case that summary
refresh/teardown should not be done, i.e. some of the aggregated
sessions need to be refreshed/torn individually.
So, QNE may need to have some mechanisms to know if summary refresh is
required or not.

cheers,
Takako


> Both of these save bytes and processing compared to doing it independently 
> for each session. They can be implemented outside GIST by keeping track of 
> the SII_HANDLE provided by GIST (however, you can only know the handle 
> after you have received a message from the per, e.g., a RESPONSE). If a 
> QNE has sessions whose next peer has the same SII_HANDLE, it can use a new 
> SESSION_LIST where it indicates SID+RSN pairs of sessions it wants to 
> refresh or tear down. The receiving QNE, in case of unknown SID, will 
> reply with a RESPONSE carrying the same object to indicate which session 
> references where not understood. If a confirmation was requested from the 
> QNR for a tear, and it must send an individual RESPONSE for each 
> terminated session, as usual - the summary refresh could hit a QNR with 
> multiple sessions being terminated.
> 
> We would like to add this to the NSLP spec. Any comments, arguments, 
> feedback? I'll wait for a week, then add this unless I hear arguments 
> against.
> 
> Jukka
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 21:04:31 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiLkI-0000KO-NF; Thu, 09 Nov 2006 21:03:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiLkG-0000Ix-QZ
	for nsis@ietf.org; Thu, 09 Nov 2006 21:03:52 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiLkB-0000C6-Ax
	for nsis@ietf.org; Thu, 09 Nov 2006 21:03:52 -0500
Received: from sbz-12.cs.helsinki.fi (sbz-12.cs.helsinki.fi [128.214.9.76])
	(AUTH: PLAIN cs-relay, TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Fri, 10 Nov 2006 04:03:38 +0200
	id 000AFD5B.4553DDFA.000012C8
Received: by sbz-12.cs.helsinki.fi (Postfix, from userid 87)
	id 20CD914B797; Fri, 10 Nov 2006 04:03:35 +0200 (EET)
To: sanda.takako@jp.panasonic.com
Subject: Re: [NSIS] QoS NSLP: summary object
Received: from 130.129.69.128 (auth. user jmanner@courier)
	by iloha.cs.Helsinki.FI with HTTP; Fri, 10 Nov 2006 04:03:35 +0200
X-IlohaMail-Blah: jmanner@cs.helsinki.fi
X-IlohaMail-Method: mail() [mem]
X-IlohaMail-Dummy: moo
X-Mailer: IlohaMail/0.8.12+pk-10 (On: iloha.cs.Helsinki.FI)
In-Reply-To: <20061110101541.18C9.SANDA.TAKAKO@jp.panasonic.com>
Message-ID: <LwTGdung.1163124215.6590130.jmanner@cs.helsinki.fi>
From: jmanner@cs.helsinki.fi
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Date: Fri, 10 Nov 2006 04:03:35 +0200 (EET)
X-Spam-Score: 0.2 (/)
X-Scan-Signature: f60d0f7806b0c40781eee6b9cd0b2135
Cc: "nsis@ietf.org" <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi,
A simple example: session refresh between domains, your QNE knows it is
the edge. For the tear, a mobility scenario, where several flows change
paths and you want to tear old states.

Jukka

On 11/10/2006, "Takako Sanda" <sanda.takako@jp.panasonic.com> wrote:
> Hi Jukka,
>=20
> On Thu, 09 Nov 2006 22:39:11 +0200 (EET)
> Jukka MJ Manner <jmanner@cs.Helsinki.FI> wrote:
>=20
> >
> > Based on John's request, here is recap of the summary object mentioned
> > today. After Roland's comment on the mic, there are actually two use
> > cases:
> >
> > 1. Allow a QNE to refresh several sessions in one NSLP refreshing RESERVE
> > message.
> >
> > 2. Allow a ANE to tear down several sessions in one NSLP tearing
> > RESERVE message.
> >
>=20
> I agree to having these use cases.
>=20
> But I am wondering how QNE decides to send summary refresh/teardown.
>=20
> For example, in aggregation scenario, there may be the case that summary
> refresh/teardown should not be done, i.e. some of the aggregated
> sessions need to be refreshed/torn individually.
> So, QNE may need to have some mechanisms to know if summary refresh is
> required or not.
>=20
> cheers,
> Takako
>=20
>=20
> > Both of these save bytes and processing compared to doing it independentl=
y
> > for each session. They can be implemented outside GIST by keeping track o=
f
> > the SII_HANDLE provided by GIST (however, you can only know the handle
> > after you have received a message from the per, e.g., a RESPONSE). If a
> > QNE has sessions whose next peer has the same SII_HANDLE, it can use a ne=
w
> > SESSION_LIST where it indicates SID+RSN pairs of sessions it wants to
> > refresh or tear down. The receiving QNE, in case of unknown SID, will
> > reply with a RESPONSE carrying the same object to indicate which session
> > references where not understood. If a confirmation was requested from the
> > QNR for a tear, and it must send an individual RESPONSE for each
> > terminated session, as usual - the summary refresh could hit a QNR with
> > multiple sessions being terminated.
> >
> > We would like to add this to the NSLP spec. Any comments, arguments,
> > feedback? I'll wait for a week, then add this unless I hear arguments
> > against.
> >
> > Jukka
> >
> > _______________________________________________
> > nsis mailing list
> > nsis@ietf.org
> > https://www1.ietf.org/mailman/listinfo/nsis
>=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 09 21:23:59 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiM31-0003tI-K9; Thu, 09 Nov 2006 21:23:15 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiM30-0003sp-5J; Thu, 09 Nov 2006 21:23:14 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1GiM2x-0002HZ-SJ; Thu, 09 Nov 2006 21:23:14 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GiM2s-0007Lz-H1; Fri, 10 Nov 2006 03:23:06 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GiM2s-0002mt-5v; Fri, 10 Nov 2006 03:23:06 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 6AF9DB232;
	Fri, 10 Nov 2006 03:23:05 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GiM2p-0005Cz-Vf; Fri, 10 Nov 2006 03:23:04 +0100
Message-ID: <4553E20A.4070304@tm.uka.de>
Date: Fri, 10 Nov 2006 03:20:58 +0100
From: Roland Bless <bless@tm.uka.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; de;
	rv:1.8.0.8) Gecko/20061025 Thunderbird/1.5.0.8 Mnenhy/0.7.4.0
MIME-Version: 1.0
To: NSIS Group <nsis@ietf.org>, nsis-imp@ietf.org
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-15
Content-Transfer-Encoding: 7bit
X-Spam-Score: -4.1 (----)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.2 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7bac9cb154eb5790ae3b2913587a40de
Cc: 
Subject: [NSIS] Next Interop
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all,

I just wanted to suggest a date for the next NSIS Interop.
Next IETF is from 18th-23rd March 2007. So having it at
the beginning of March would be good (5.3.-9.3.2007).
Maybe the week before is also good.
Other suggestions?

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 10 02:40:48 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiQzk-0008QK-Pv; Fri, 10 Nov 2006 02:40:12 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiQzj-0008QE-6v
	for nsis@ietf.org; Fri, 10 Nov 2006 02:40:11 -0500
Received: from tcmail31.telekom.de ([217.6.95.238])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiQzh-0000xt-NX
	for nsis@ietf.org; Fri, 10 Nov 2006 02:40:11 -0500
Received: from S4DE8PSAAGT.mitte.t-com.de by tcmail31.telekom.de with ESMTP;
	Fri, 10 Nov 2006 08:40:08 +0100
Received: from S4DE8PSAAFQ.mitte.t-com.de ([10.151.180.5]) by
	S4DE8PSAAGT.mitte.t-com.de with Microsoft SMTPSVC(6.0.3790.1830); 
	Fri, 10 Nov 2006 08:40:08 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: Re: [NSIS] QoS NSLP: summary object
Date: Fri, 10 Nov 2006 08:40:07 +0100
Message-Id: <6439282641581441A36F7F6F83ED2ED2CBF94C@S4DE8PSAAFQ.mitte.t-com.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] QoS NSLP: summary object
Thread-Index: AccESPl/qdL1YxgPREymvK3I8teKlAAUDnoA
From: "Geib, Ruediger" <Ruediger.Geib@t-systems.com>
To: <jmanner@cs.Helsinki.FI>,
    <bless@tm.uka.de>
X-OriginalArrivalTime: 10 Nov 2006 07:40:08.0374 (UTC)
	FILETIME=[72148D60:01C7049B]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ff03b0075c3fc728d7d60a15b4ee1ad2
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Dear Jukka, dear Roland

as expressed by my mail

http://www1.ietf.org/mail-archive/web/nsis/current/msg04320.html

(see "3.2.5 Change to: Refresh options"),

I support proposal number 1 on refreshing several sessions=20
by a single Reserve.=20

The above mail also proposed changes to the RSN mechanism aiming to=20
simplify summary refresh, but these aren't applicable any more=20
(what I proposed back then on RSN handling would require=20
significant changes in NSLP - too late for that now).

Regards,

Rudiger

-----Original Message-----
From: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI]
Sent: Thursday, November 09, 2006 10:48 PM
To: Roland Bless
Cc: nsis@ietf.org
Subject: Re: [NSIS] QoS NSLP: summary object



Hi,
two comments back. :)

Jukka

On Thu, 9 Nov 2006, Roland Bless wrote:

> Hi Jukka and all,
>=20
> Jukka MJ Manner wrote:
> > Based on John's request, here is recap of the summary object
mentioned=20
> > today. After Roland's comment on the mic, there are actually two use

> > cases:
> >=20
> > 1. Allow a QNE to refresh several sessions in one NSLP refreshing
RESERVE=20
> > message.=20
> >=20
> > 2. Allow a ANE to tear down several sessions in one NSLP tearing=20
> > RESERVE message.
> >=20
> > Both of these save bytes and processing compared to doing it
independently=20
> > for each session. They can be implemented outside GIST by keeping
track of=20
> > the SII_HANDLE provided by GIST (however, you can only know the
handle=20
> > after you have received a message from the per, e.g., a RESPONSE).
If a=20
>=20
> Not quite. After you received a GIST response. GIST will report it
> upwards to the NSLP. However, you care about refresh or tear only
> after the initial setup worked. So I don't see any problems here.
> Furthermore, it is an optimization, so basically no problem if it
> is not applicable in some rare cases.

[1] This note comes from looking at the GIST API. In order to send a=20
summary refresh, you need to have compared the GIST SII_HANDLE before.
And=20
you will only get to know the SII_HANDLE for a session if you receive
the=20
RESPONSE. It is not a problem in the general case, just a note.

>=20
> > QNE has sessions whose next peer has the same SII_HANDLE, it can use
a new=20
> > SESSION_LIST where it indicates SID+RSN pairs of sessions it wants
to=20
> > refresh or tear down. The receiving QNE, in case of unknown SID,
will=20
>=20
> Yes, and as we just discussed in private, it would be good to have
> an SID-List only object (SIDs w/o RSN), too.

[2] Actualy... may be not. You need the SID AND the RSN for both tear
and=20
refresh. Thus, you need to send both anyway. RSN is a mandatory object
for=20
all RESERVE messages. You need to have the RSN in synch if you are
tearing=20
otherwise you will tear a different state than the previous QNE.

>=20
> > reply with a RESPONSE carrying the same object to indicate which
session=20
> > references where not understood. If a confirmation was requested
from the=20
> > QNR for a tear, and it must send an individual RESPONSE for each=20
> > terminated session, as usual - the summary refresh could hit a QNR
with=20
> > multiple sessions being terminated.
>=20
> In this case we would get a summary error back.
>=20
> > We would like to add this to the NSLP spec. Any comments, arguments,

> > feedback? I'll wait for a week, then add this unless I hear
arguments=20
> > against.
>=20
> Obviously I'm in favor of having that. However, we must be careful
> about covering the new error cases and handle them properly.
>=20
> Regards,
>  Roland
>=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 10 12:09:34 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiZs9-0006kH-6n; Fri, 10 Nov 2006 12:08:57 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiZs8-0006gc-0r
	for nsis@ietf.org; Fri, 10 Nov 2006 12:08:56 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiZs5-0000Ng-Gc
	for nsis@ietf.org; Fri, 10 Nov 2006 12:08:56 -0500
Received: from sbz-30.cs.helsinki.fi (sbz-30.cs.helsinki.fi [128.214.9.98])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Fri, 10 Nov 2006 19:08:52 +0200
	id 000AFD0C.4554B224.00005CE3
Date: Fri, 10 Nov 2006 19:08:52 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: NSIS Group <nsis@ietf.org>
Subject: Re: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
In-Reply-To: <4553A860.40004@tm.uka.de>
Message-ID: <Pine.LNX.4.64.0611101903510.14922@sbz-30.cs.Helsinki.FI>
References: <5E7777BA-D9C1-4E4E-90E1-187F09A6236B@cisco.com>
	<4553A860.40004@tm.uka.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 14582b0692e7f70ce7111d04db3781c8
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi all,

I'd very much like to see approach "2" in addition to approach "3".

In addition to what Roland discussed below, a reason to have the "local 
vs. e2e" QSPEC together inside a single QSPEC is that local QoS objects 
and piggybacking is a functionality of a particular QoS model. If a QoS 
Model needs those functionality, they should be in the RMF, not in the 
base protocol, the NSLP.

Jukka

On Thu, 9 Nov 2006, Roland Bless wrote:

> Hi Dave and all,
> 
> David R Oran wrote:
> > In thinking about what protocol designs could meet the above
> > requirements, I discerned three. One is essentially identical to the
> > former/existing QSPEC stacking functionality we wanted to eliminate. A
> > second is a proposal by Roland Bless. The third is to not handle this in
> > QSPECS as all, but by recursing at the NSLP level.
> > 
> > In more detail, the three alternatives are:
> > 
> > 1. Do complete QSPEC stacking, and deal with the issues around what
> > constraints have to be met by the stacked QSPEC, whether the QSPEC lower
> > in the "stack" is opacified, and what the rules are for pushing,
> > swapping, and popping.
> 
> > 2. Allow domain-parameters to be added to a QSPEC on domain entry and
> > removed on exit. In this option one has to describe whether anything is
> > opacified and which existing QSPEC parameters get "re-interpreted" if any.
> 
> It's actually not the _same_ QSPEC with added parameters,
> but it would be a "new" QSPEC (type). So to give an
> example in order to make it clear, an RMD-QSPEC would
> look like this:
> ----------------------------------
> | QSPEC-Type = RMD-QSPEC         |  Common QSPEC header
> +================================+\
> |  RMD-specific Parameter 1      | \
> +--------------------------------+  \
> |             ....               |   local QSPEC parameters
> +--------------------------------+  /
> |  RMD-specific Parameter n      | /
> +--------------------------------+/
> | +----------------------------+ |
> | | QSPEC-Type = Generic QSPEC | |
> | +============================+ |
> | |                            | | Encapsulated "external" QSPEC
> | |          ....              | |
> | +----------------------------+ |
> +--------------------------------+
> The important difference to 1 is that QoS-NSLP only sees
> and passes one QSPEC up to the RMF. The type of the
> QSPEC thus may change within a domain. As written in my earlier
> mail:
> 1. QNI signals its QoS requirements with its Generic QSPEC
> 2. The ingress edge QNE will map these to its own QoS model, e.g.
>    choosing the corresponding DiffServ or Y.1541 class.
>    The mapped parameter values may be carried within
>    the above described local RMD-QSPEC.
> 3. The egress QNE will process the RMD-QSPEC and pass only
>     the encapsulated Generic QSPEC to the next domain, that
>     may perform its own mapping again possibly without requiring
>     to transmit its own local QSPEC.
> 
> > 3. Recurse at the NSLP level, have a separate domain-private NSLP
> > session, and that would contain a domain-private set of paramteres
> > instead of the end-to-end QSPEC. I'll note that this approach allows
> > (but not requires) the two signaling state machines (end-end and
> > edge-edge through interior nodes) to be decoupled.
> 
> > We already have (3) for other purposes, like aggregation and tunneling.
> 
> But Georgios point was probably more to avoid having a
> separate signaling and to piggyback additional information instead.
> So this would be an optimization, not possible with approach 3 only.
> 
> My view is the following:
> 1 is more complicated than 2, at least for the QoS NSLP.
> 2 could be used in addition to 3.
> I'm really open to hear cases were approach 2 fails badly
> or is disadvantageous over approach 1.
> 
> Regards,
>  Roland
> 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 10 12:28:20 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GiaAF-0008OE-05; Fri, 10 Nov 2006 12:27:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GiaAC-0008GA-Sa
	for nsis@ietf.org; Fri, 10 Nov 2006 12:27:36 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GiaAB-0002vr-Il
	for nsis@ietf.org; Fri, 10 Nov 2006 12:27:36 -0500
Received: from sbz-30.cs.helsinki.fi (sbz-30.cs.helsinki.fi [128.214.9.98])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Fri, 10 Nov 2006 19:27:35 +0200
	id 000AFD0C.4554B687.000074FD
Date: Fri, 10 Nov 2006 19:27:34 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: nsis@ietf.org
Message-ID: <Pine.LNX.4.64.0611101911080.14922@sbz-30.cs.Helsinki.FI>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e1e48a527f609d1be2bc8d8a70eb76cb
Subject: [NSIS] Summary of forthcoming QoS NSLP changes
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Dear all,

Here is a list of the changes I intend to do to the current QoS NSLP spec. 
After those, we are ready for WGLC. The spec is looking good. If there are 
no arguments, I could do these within a week.

- Make reduced refreshes the default behavior, add a new error code for 
  "full refreshes required".

- Some clarifications needed for proxy mode (and GIST),
  object order (SHOULD use a certain, MUST accept in any order), tear 
  operation, response/notify (RESPONSE is triggered by an incoming 
  message, NOTIFY is fully asynchronous)

- Summary refresh messages: one NSLP message can refresh several 
  sessions (use a new object with SID+RSN pairs)

- Summary tear: use a new object with SIDs only.

- RSN MUST NOT be interpreted when receiving a tear. If two QNEs are not 
  in synch, what can you? Update the reservation, then tear? What if the 
  update fails? There is no reason for QNEs to be in synch, you just 
  remove the resources you had allocated.

- Re-introduce the ack-flag. This is needed in order to do diagnostics, 
  say, you want to make sure the next peer actually processed your message 
  and understood it. Also need to add a new error code that tells the 
  QNI if the "path is truncated", i.e., a peer is responding to the Ack. 
  It is a policy decision on a QNE whether it should send e.g. an 
  Ack-flaged refresh when it receives such a refresh from its upstream 
  peer. 

- [After concensus] Update the draft on "stacking".


Yell now if some of these are not right things to do.

You could also yell, if these are, that would be nice, too. :)

Jukka

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 06:50:30 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjaJ2-0000oS-J2; Mon, 13 Nov 2006 06:48:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjaHa-0008HS-QQ
	for nsis@ietf.org; Mon, 13 Nov 2006 06:47:22 -0500
Received: from mail121.messagelabs.com ([216.82.241.195])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gja5J-0007Cu-7V
	for nsis@ietf.org; Mon, 13 Nov 2006 06:34:44 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-11.tower-121.messagelabs.com!1163417680!12735404!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 14846 invoked from network); 13 Nov 2006 11:34:40 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-11.tower-121.messagelabs.com with SMTP;
	13 Nov 2006 11:34:40 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kADBYegI006038; 
	Mon, 13 Nov 2006 06:34:40 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst12.ugd.att.com [135.38.164.89])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kADBYTQ0005992; Mon, 13 Nov 2006 06:34:29 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 05:34:29 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F76@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <4553A860.40004@tm.uka.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Thread-Index: AccETTBRKwUymAEBRBukYt/ILQ8d+wCxt4KA
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: "Roland Bless" <bless@tm.uka.de>, "David R Oran" <oran@cisco.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2857c5c041d6c02d7181d602c22822c8
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi All,

I agree that we should go with option 2 (encapsulated QSPEC) versus
option 1 (stacked QSPEC) as the easier one to implement (according to
discussions with Roland).  We should also retain option 3 (tunneled
QSPEC) as currently specified.

IMO we should retain the QoS class parameters as optional parameters in
the QSPEC document.  One reason is that support for DiffServ was an
original basic requirement on the QSPEC, and that requirement still must
be met.  Additionally, it would require initiating new QOSM
specification documents if these parameters are moved elsewhere -- it is
far simpler, and appropriate, to retain them as is in the QSPEC
document.

Hopefully we are now ready to revise the QSPEC document according to the
agreements summarized in the presentation at IETF-67
https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=3D6=
7
(subject to the revisions of including option 2/encapsulated QSPEC).

Thanks,
Jerry

-----Original Message-----
From: Roland Bless [mailto:bless@tm.uka.de]=20
Sent: Thursday, November 09, 2006 5:15 PM
To: David R Oran
Cc: NSIS Group
Subject: Re: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

Hi Dave and all,

David R Oran wrote:
> In thinking about what protocol designs could meet the above
> requirements, I discerned three. One is essentially identical to the
> former/existing QSPEC stacking functionality we wanted to eliminate. A
> second is a proposal by Roland Bless. The third is to not handle this
in
> QSPECS as all, but by recursing at the NSLP level.
>=20
> In more detail, the three alternatives are:
>=20
> 1. Do complete QSPEC stacking, and deal with the issues around what
> constraints have to be met by the stacked QSPEC, whether the QSPEC
lower
> in the "stack" is opacified, and what the rules are for pushing,
> swapping, and popping.

> 2. Allow domain-parameters to be added to a QSPEC on domain entry and
> removed on exit. In this option one has to describe whether anything
is
> opacified and which existing QSPEC parameters get "re-interpreted" if
any.

It's actually not the _same_ QSPEC with added parameters,
but it would be a "new" QSPEC (type). So to give an
example in order to make it clear, an RMD-QSPEC would
look like this:
----------------------------------
| QSPEC-Type =3D RMD-QSPEC         |  Common QSPEC header
+=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D+\
|  RMD-specific Parameter 1      | \
+--------------------------------+  \
|             ....               |   local QSPEC parameters
+--------------------------------+  /
|  RMD-specific Parameter n      | /
+--------------------------------+/
| +----------------------------+ |
| | QSPEC-Type =3D Generic QSPEC | |
| =
+=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D+ |
| |                            | | Encapsulated "external" QSPEC
| |          ....              | |
| +----------------------------+ |
+--------------------------------+
The important difference to 1 is that QoS-NSLP only sees
and passes one QSPEC up to the RMF. The type of the
QSPEC thus may change within a domain. As written in my earlier
mail:
1. QNI signals its QoS requirements with its Generic QSPEC
2. The ingress edge QNE will map these to its own QoS model, e.g.
   choosing the corresponding DiffServ or Y.1541 class.
   The mapped parameter values may be carried within
   the above described local RMD-QSPEC.
3. The egress QNE will process the RMD-QSPEC and pass only
    the encapsulated Generic QSPEC to the next domain, that
    may perform its own mapping again possibly without requiring
    to transmit its own local QSPEC.

> 3. Recurse at the NSLP level, have a separate domain-private NSLP
> session, and that would contain a domain-private set of paramteres
> instead of the end-to-end QSPEC. I'll note that this approach allows
> (but not requires) the two signaling state machines (end-end and
> edge-edge through interior nodes) to be decoupled.

> We already have (3) for other purposes, like aggregation and
tunneling.

But Georgios point was probably more to avoid having a
separate signaling and to piggyback additional information instead.
So this would be an optimization, not possible with approach 3 only.

My view is the following:
1 is more complicated than 2, at least for the QoS NSLP.
2 could be used in addition to 3.
I'm really open to hear cases were approach 2 fails badly
or is disadvantageous over approach 1.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 06:57:47 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjaQy-0006tS-76; Mon, 13 Nov 2006 06:57:04 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjaQw-0006sy-WD
	for nsis@ietf.org; Mon, 13 Nov 2006 06:57:03 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GjaQu-0003fU-2b
	for nsis@ietf.org; Mon, 13 Nov 2006 06:57:02 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kADBuhUA000572; Mon, 13 Nov 2006 12:56:46 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>,
	"'Roland Bless'" <bless@tm.uka.de>, "'David R Oran'" <oran@cisco.com>
Subject: RE: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 12:56:46 +0100
Message-ID: <001b01c7071a$ce332170$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F76@KCCLUST06EVS1.ugd.att.com>
Thread-Index: AccETTBRKwUymAEBRBukYt/ILQ8d+wCxt4KAAAGe//A=
X-Spam-Score: 0.3 () FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Mon, 13 Nov 2006 12:56:47 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8f374d0786b25a451ef87d82c076f593
Cc: 'NSIS Group' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

I agree with you!
I think that we now have reached a consensus on all open issues and we can 
proceed on finalizing all NSIS base drafts!

Best Regards,
Georgios 

> -----Original Message-----
> From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com] 
> Sent: maandag 13 november 2006 12:34
> To: Roland Bless; David R Oran
> Cc: Ash, Gerald R (Jerry), ALABS; NSIS Group
> Subject: RE: [NSIS] Alternatives for supporting domain-local 
> admissioncontroloperation in NSLP/QSPEC
> 
> Hi All,
> 
> I agree that we should go with option 2 (encapsulated QSPEC) 
> versus option 1 (stacked QSPEC) as the easier one to 
> implement (according to discussions with Roland).  We should 
> also retain option 3 (tunneled
> QSPEC) as currently specified.
> 
> IMO we should retain the QoS class parameters as optional 
> parameters in the QSPEC document.  One reason is that support 
> for DiffServ was an original basic requirement on the QSPEC, 
> and that requirement still must be met.  Additionally, it 
> would require initiating new QOSM specification documents if 
> these parameters are moved elsewhere -- it is far simpler, 
> and appropriate, to retain them as is in the QSPEC document.
> 
> Hopefully we are now ready to revise the QSPEC document 
> according to the agreements summarized in the presentation at IETF-67
> https://datatracker.ietf.org/public/meeting_materials.cgi?meet
> ing_num=67
> (subject to the revisions of including option 2/encapsulated QSPEC).
> 
> Thanks,
> Jerry
> 
> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de]
> Sent: Thursday, November 09, 2006 5:15 PM
> To: David R Oran
> Cc: NSIS Group
> Subject: Re: [NSIS] Alternatives for supporting domain-local 
> admission controloperation in NSLP/QSPEC
> 
> Hi Dave and all,
> 
> David R Oran wrote:
> > In thinking about what protocol designs could meet the above 
> > requirements, I discerned three. One is essentially 
> identical to the 
> > former/existing QSPEC stacking functionality we wanted to 
> eliminate. A 
> > second is a proposal by Roland Bless. The third is to not 
> handle this
> in
> > QSPECS as all, but by recursing at the NSLP level.
> > 
> > In more detail, the three alternatives are:
> > 
> > 1. Do complete QSPEC stacking, and deal with the issues around what 
> > constraints have to be met by the stacked QSPEC, whether the QSPEC
> lower
> > in the "stack" is opacified, and what the rules are for pushing, 
> > swapping, and popping.
> 
> > 2. Allow domain-parameters to be added to a QSPEC on domain 
> entry and 
> > removed on exit. In this option one has to describe whether anything
> is
> > opacified and which existing QSPEC parameters get 
> "re-interpreted" if
> any.
> 
> It's actually not the _same_ QSPEC with added parameters, but 
> it would be a "new" QSPEC (type). So to give an example in 
> order to make it clear, an RMD-QSPEC would look like this:
> ----------------------------------
> | QSPEC-Type = RMD-QSPEC         |  Common QSPEC header
> +================================+\
> |  RMD-specific Parameter 1      | \
> +--------------------------------+  \
> |             ....               |   local QSPEC parameters
> +--------------------------------+  /
> |  RMD-specific Parameter n      | /
> +--------------------------------+/
> | +----------------------------+ |
> | | QSPEC-Type = Generic QSPEC | |
> | +============================+ |
> | |                            | | Encapsulated "external" QSPEC
> | |          ....              | |
> | +----------------------------+ |
> +--------------------------------+
> The important difference to 1 is that QoS-NSLP only sees and 
> passes one QSPEC up to the RMF. The type of the QSPEC thus 
> may change within a domain. As written in my earlier
> mail:
> 1. QNI signals its QoS requirements with its Generic QSPEC 2. 
> The ingress edge QNE will map these to its own QoS model, e.g.
>    choosing the corresponding DiffServ or Y.1541 class.
>    The mapped parameter values may be carried within
>    the above described local RMD-QSPEC.
> 3. The egress QNE will process the RMD-QSPEC and pass only
>     the encapsulated Generic QSPEC to the next domain, that
>     may perform its own mapping again possibly without requiring
>     to transmit its own local QSPEC.
> 
> > 3. Recurse at the NSLP level, have a separate domain-private NSLP 
> > session, and that would contain a domain-private set of paramteres 
> > instead of the end-to-end QSPEC. I'll note that this 
> approach allows 
> > (but not requires) the two signaling state machines (end-end and 
> > edge-edge through interior nodes) to be decoupled.
> 
> > We already have (3) for other purposes, like aggregation and
> tunneling.
> 
> But Georgios point was probably more to avoid having a 
> separate signaling and to piggyback additional information instead.
> So this would be an optimization, not possible with approach 3 only.
> 
> My view is the following:
> 1 is more complicated than 2, at least for the QoS NSLP.
> 2 could be used in addition to 3.
> I'm really open to hear cases were approach 2 fails badly or 
> is disadvantageous over approach 1.
> 
> Regards,
>  Roland
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 10:40:18 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjduG-0007sk-Bq; Mon, 13 Nov 2006 10:39:32 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjduF-0007sZ-0w
	for nsis@ietf.org; Mon, 13 Nov 2006 10:39:31 -0500
Received: from mailgw4.ericsson.se ([193.180.251.62])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gjdu8-00038S-BL
	for nsis@ietf.org; Mon, 13 Nov 2006 10:39:30 -0500
Received: from esealmw126.eemea.ericsson.se (unknown [153.88.254.123])
	by mailgw4.ericsson.se (Symantec Mail Security) with ESMTP id 7C6DCB4C; 
	Mon, 13 Nov 2006 16:39:17 +0100 (CET)
Received: from esealmw116.eemea.ericsson.se ([153.88.200.7]) by
	esealmw126.eemea.ericsson.se with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 13 Nov 2006 16:39:17 +0100
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 16:39:16 +0100
Message-ID: <05F34B6959F19F43BBD78FFCD4C15D055E0009@esealmw116.eemea.ericsson.se>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F76@KCCLUST06EVS1.ugd.att.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Thread-Index: AccETTBRKwUymAEBRBukYt/ILQ8d+wCxt4KAAAk3kMA=
From: =?iso-8859-1?Q?Attila_B=E1der_=28IJ/ETH=29?= <attila.bader@ericsson.com>
To: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>,
	"Roland Bless" <bless@tm.uka.de>, "David R Oran" <oran@cisco.com>
X-OriginalArrivalTime: 13 Nov 2006 15:39:17.0231 (UTC)
	FILETIME=[E0F93FF0:01C70739]
X-Brightmail-Tracker: AAAAAA==
X-Spam-Score: 0.0 (/)
X-Scan-Signature: dbb8771284c7a36189745aa720dc20ab
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi All, I also think it is good compromise. Option 3 is important =
feature, so we should keep it. Option 2 is OK to replace stacking.

Best regards, Attila=20

>-----Original Message-----
>From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com]=20
>Sent: Monday, November 13, 2006 12:34 PM
>To: Roland Bless; David R Oran
>Cc: Ash, Gerald R (Jerry), ALABS; NSIS Group
>Subject: RE: [NSIS] Alternatives for supporting domain-local=20
>admissioncontroloperation in NSLP/QSPEC
>
>Hi All,
>
>I agree that we should go with option 2 (encapsulated QSPEC)=20
>versus option 1 (stacked QSPEC) as the easier one to implement=20
>(according to discussions with Roland).  We should also retain=20
>option 3 (tunneled
>QSPEC) as currently specified.
>
>IMO we should retain the QoS class parameters as optional=20
>parameters in the QSPEC document.  One reason is that support=20
>for DiffServ was an original basic requirement on the QSPEC,=20
>and that requirement still must be met.  Additionally, it=20
>would require initiating new QOSM specification documents if=20
>these parameters are moved elsewhere -- it is far simpler, and=20
>appropriate, to retain them as is in the QSPEC document.
>
>Hopefully we are now ready to revise the QSPEC document=20
>according to the agreements summarized in the presentation at IETF-67
>https://datatracker.ietf.org/public/meeting_materials.cgi?meeti
>ng_num=3D67
>(subject to the revisions of including option 2/encapsulated QSPEC).
>
>Thanks,
>Jerry
>
>-----Original Message-----
>From: Roland Bless [mailto:bless@tm.uka.de]
>Sent: Thursday, November 09, 2006 5:15 PM
>To: David R Oran
>Cc: NSIS Group
>Subject: Re: [NSIS] Alternatives for supporting domain-local=20
>admission controloperation in NSLP/QSPEC
>
>Hi Dave and all,
>
>David R Oran wrote:
>> In thinking about what protocol designs could meet the above=20
>> requirements, I discerned three. One is essentially identical to the=20
>> former/existing QSPEC stacking functionality we wanted to=20
>eliminate. A=20
>> second is a proposal by Roland Bless. The third is to not handle this
>in
>> QSPECS as all, but by recursing at the NSLP level.
>>=20
>> In more detail, the three alternatives are:
>>=20
>> 1. Do complete QSPEC stacking, and deal with the issues around what=20
>> constraints have to be met by the stacked QSPEC, whether the QSPEC
>lower
>> in the "stack" is opacified, and what the rules are for pushing,=20
>> swapping, and popping.
>
>> 2. Allow domain-parameters to be added to a QSPEC on domain=20
>entry and=20
>> removed on exit. In this option one has to describe whether anything
>is
>> opacified and which existing QSPEC parameters get "re-interpreted" if
>any.
>
>It's actually not the _same_ QSPEC with added parameters, but=20
>it would be a "new" QSPEC (type). So to give an example in=20
>order to make it clear, an RMD-QSPEC would look like this:
>----------------------------------
>| QSPEC-Type =3D RMD-QSPEC         |  Common QSPEC header
>+=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D+\
>|  RMD-specific Parameter 1      | \
>+--------------------------------+  \
>|             ....               |   local QSPEC parameters
>+--------------------------------+  /
>|  RMD-specific Parameter n      | /
>+--------------------------------+/
>| +----------------------------+ |
>| | QSPEC-Type =3D Generic QSPEC | |
>| =
+=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D+ |
>| |                            | | Encapsulated "external" QSPEC
>| |          ....              | |
>| +----------------------------+ |
>+--------------------------------+
>The important difference to 1 is that QoS-NSLP only sees and=20
>passes one QSPEC up to the RMF. The type of the QSPEC thus may=20
>change within a domain. As written in my earlier
>mail:
>1. QNI signals its QoS requirements with its Generic QSPEC 2.=20
>The ingress edge QNE will map these to its own QoS model, e.g.
>   choosing the corresponding DiffServ or Y.1541 class.
>   The mapped parameter values may be carried within
>   the above described local RMD-QSPEC.
>3. The egress QNE will process the RMD-QSPEC and pass only
>    the encapsulated Generic QSPEC to the next domain, that
>    may perform its own mapping again possibly without requiring
>    to transmit its own local QSPEC.
>
>> 3. Recurse at the NSLP level, have a separate domain-private NSLP=20
>> session, and that would contain a domain-private set of paramteres=20
>> instead of the end-to-end QSPEC. I'll note that this approach allows=20
>> (but not requires) the two signaling state machines (end-end and=20
>> edge-edge through interior nodes) to be decoupled.
>
>> We already have (3) for other purposes, like aggregation and
>tunneling.
>
>But Georgios point was probably more to avoid having a=20
>separate signaling and to piggyback additional information instead.
>So this would be an optimization, not possible with approach 3 only.
>
>My view is the following:
>1 is more complicated than 2, at least for the QoS NSLP.
>2 could be used in addition to 3.
>I'm really open to hear cases were approach 2 fails badly or=20
>is disadvantageous over approach 1.
>
>Regards,
> Roland
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 11:13:01 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjePS-0008CJ-Cp; Mon, 13 Nov 2006 11:11:46 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjePQ-0008CB-VF
	for nsis@ietf.org; Mon, 13 Nov 2006 11:11:44 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GjePO-0007iM-L4
	for nsis@ietf.org; Mon, 13 Nov 2006 11:11:44 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Mon, 13 Nov 2006 11:07:45 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006111311074519238 ; Mon, 13 Nov 2006 11:07:45 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <WZT2N7V8>; Mon, 13 Nov 2006 11:07:45 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EE7@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: Georgios Karagiannis <karagian@cs.utwente.nl>,
	"Ash, Gerald R (Jerry), ALABS" <gash@att.com>,
	Roland Bless <bless@tm.uka.de>, David R Oran <oran@cisco.com>
Subject: RE: [NSIS] Alternatives for supporting domain-localadmissioncontr
	oloperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 11:07:34 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 13 Nov 2006 16:07:37.0782 (UTC)
	FILETIME=[D694BD60:01C7073D]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d6aff76b15f3f56fcb94490e1052e4
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios and all:

The fact is that all QOS model-specific extensions, explanations, and
references MUST be DROPPED per consensus of this IETF meeting.  

In this way, the problems of all NON-interoperability among the
standard-based IETF QOS models (RSVP, IntServ, and DiffServ) and
so-called "private" non-IETF, non-standard QOS models (Y.1541 and
3GPP/Y.1541) will be avoided in this standard-track "generic" QSPEC RFC.

If anyone wants to avoid this consensus of the IETF meeting, they MUST
come up with the technical proposal how to solve the
NON-interoperability problems among the standard-based IETF QOS models
(RSVP, IntServ, and DiffServ) and so-called "private" non-IETF,
non-standard QOS models (Y.1541 and 3GPP/Y.1541). If brute forces are
used to add any references of QOS model-specific things as the present
draft stands, the consensus of the IETF will be broken and the draft
will NOT move for IESG approval as we will be again going to the circle
of interoperability issues.

Best regards,
Radhika

-----Original Message-----
From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On Behalf Of
Georgios Karagiannis
Sent: Monday, November 13, 2006 6:57 AM
To: 'Ash, Gerald R (Jerry), ALABS'; 'Roland Bless'; 'David R Oran'
Cc: 'NSIS Group'
Subject: RE: [NSIS] Alternatives for supporting
domain-localadmissioncontroloperation in NSLP/QSPEC

Hi Jerry

I agree with you!
I think that we now have reached a consensus on all open issues and we
can 
proceed on finalizing all NSIS base drafts!

Best Regards,
Georgios 


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 11:34:17 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjekD-0005VE-Hc; Mon, 13 Nov 2006 11:33:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjekB-0005Ux-Ux
	for nsis@ietf.org; Mon, 13 Nov 2006 11:33:11 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gjek9-0002cl-Hb
	for nsis@ietf.org; Mon, 13 Nov 2006 11:33:11 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kADGWxKv012440; Mon, 13 Nov 2006 17:33:05 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <RADHIKA.R.ROY@saic.com>,
	"'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>,
	"'Roland Bless'" <bless@tm.uka.de>, "'David R Oran'" <oran@cisco.com>
Subject: RE: [NSIS] Alternatives for supporting
	domain-localadmissioncontroloperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 17:32:54 +0100
Message-ID: <002f01c70741$61c2e620$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <62D92A9A02BCC845B202323D49A48426230EE7@0591-ITS-EXMP02.us.saic.com>
Thread-Index: AccHPmlW7hn2S8ITSSGqC9NCzTPWowAAquPw
X-Spam-Score: 0.3 () FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Mon, 13 Nov 2006 17:33:05 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: c0bedb65cce30976f0bf60a0a39edea4
Cc: 'NSIS Group' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roy

There are nomore interoperability issues since the only parameter 
that can be considered as mandatory is the Token Bucket parameter, 
which is used as the traffic source description.

Best regards,
Georgios



> -----Original Message-----
> From: RADHIKA.R.ROY@saic.com [mailto:RADHIKA.R.ROY@saic.com] 
> Sent: maandag 13 november 2006 17:08
> To: Georgios Karagiannis; Ash, Gerald R (Jerry), ALABS; 
> Roland Bless; David R Oran
> Cc: NSIS Group
> Subject: RE: [NSIS] Alternatives for supporting 
> domain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Georgios and all:
> 
> The fact is that all QOS model-specific extensions, 
> explanations, and references MUST be DROPPED per consensus of 
> this IETF meeting.  
> 
> In this way, the problems of all NON-interoperability among 
> the standard-based IETF QOS models (RSVP, IntServ, and 
> DiffServ) and so-called "private" non-IETF, non-standard QOS 
> models (Y.1541 and
> 3GPP/Y.1541) will be avoided in this standard-track "generic" 
> QSPEC RFC.
> 
> If anyone wants to avoid this consensus of the IETF meeting, 
> they MUST come up with the technical proposal how to solve 
> the NON-interoperability problems among the standard-based 
> IETF QOS models (RSVP, IntServ, and DiffServ) and so-called 
> "private" non-IETF, non-standard QOS models (Y.1541 and 
> 3GPP/Y.1541). If brute forces are used to add any references 
> of QOS model-specific things as the present draft stands, the 
> consensus of the IETF will be broken and the draft will NOT 
> move for IESG approval as we will be again going to the 
> circle of interoperability issues.
> 
> Best regards,
> Radhika
> 
> -----Original Message-----
> From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On 
> Behalf Of Georgios Karagiannis
> Sent: Monday, November 13, 2006 6:57 AM
> To: 'Ash, Gerald R (Jerry), ALABS'; 'Roland Bless'; 'David R Oran'
> Cc: 'NSIS Group'
> Subject: RE: [NSIS] Alternatives for supporting 
> domain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Hi Jerry
> 
> I agree with you!
> I think that we now have reached a consensus on all open 
> issues and we can proceed on finalizing all NSIS base drafts!
> 
> Best Regards,
> Georgios 
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 12:27:04 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjfZU-0007Xb-Kr; Mon, 13 Nov 2006 12:26:12 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjfZT-0007XF-V0
	for nsis@ietf.org; Mon, 13 Nov 2006 12:26:11 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GjfZR-0001il-Jm
	for nsis@ietf.org; Mon, 13 Nov 2006 12:26:11 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Mon, 13 Nov 2006 12:12:37 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006111312123721259 ; Mon, 13 Nov 2006 12:12:37 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <WZT2318V>; Mon, 13 Nov 2006 12:12:37 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EE9@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: Georgios Karagiannis <karagian@cs.utwente.nl>, RADHIKA.R.ROY@saic.com,
	"Ash, Gerald R (Jerry), ALABS" <gash@att.com>,
	Roland Bless <bless@tm.uka.de>, David R Oran <oran@cisco.com>
Subject: RE: [NSIS] Alternatives for supportingdomain-localadmissioncontro
	loperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 12:12:33 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 13 Nov 2006 17:12:34.0335 (UTC)
	FILETIME=[E91B8AF0:01C70746]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 22bbb45ef41b733eb2d03ee71ece8243
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios:

This is only true for the standard-based IETF QOS Models (RSVP, IntServ,
and DiffServ). 

All other so-called QOS models (non-standards DSTE, Y.1541, and
3GPP/Y.1541) have no concept of Token Bucket as they do not have QOS
signaling messages defined of their own other than some values of
delays, jitter, and packet losses. It happens to be, like all SLAs,
these implementation-specific value-based QOS models can also be mapped
into DiffServ QOS messages.

The value-based QOS models are ONLY applicable for implementation
agreements (e.g., OSP). For example, some specific values of some
parameters of DiffServ need to be agreed among some implementers or
ISPs, and they will develop SLAs for this. This does not mean we will
include all SLAs of the whole world as a part of this "generic" standard
QSPEC RFC.

In this respect, specific value-based QOS models like DSTE, Y.1541, and
3GPP/Y.1541 are nothing but some value-based QOS parameters of "service
level agreements (SLAs).

If needed, any reference to DiffServ QOS model also needs be taken off
from this standard-track RFC. The token bucket is good enough to be used
by DiffServ, RSVP, and IntServ as they are the QOS models per standards
and have their own QOS signaling messages that will provide
"interoperability."

This standard track RFC even does NOT need to include any references to
DiffServ QOS model as well. Again, if all SLAs are mapped into DiffServ,
that does not mean all implementation-specific value-based SLAs are QOS
models and then we should start including one SLA after another into
this standard track RFC. The inclusion of DiffServ QOS model seems to us
as a "ploy" for inclusion of other SLA-type QOS things.

What does "force" the NSIS WG to "mess" the whole standard-track RFC
after all these hard works "forcing" the same non-interoperability
issues mixing between apples and oranges in the same draft?

Why does the NSIS WG not write separate individual draft for each QOS
model (RSVP, IntServ, DiffServ, DSTE, Y.1541, 3GPP/Y.1541) how to use
the QSPEC using NSIS QOS signaling messages?

Best regards,
Radhika

-----Original Message-----
From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On Behalf Of
Georgios Karagiannis
Sent: Monday, November 13, 2006 11:33 AM
To: RADHIKA.R.ROY@saic.com; 'Ash, Gerald R (Jerry), ALABS'; 'Roland
Bless'; 'David R Oran'
Cc: 'NSIS Group'
Subject: RE: [NSIS] Alternatives for
supportingdomain-localadmissioncontroloperation in NSLP/QSPEC

Hi Roy

There are nomore interoperability issues since the only parameter 
that can be considered as mandatory is the Token Bucket parameter, 
which is used as the traffic source description.

Best regards,
Georgios



> -----Original Message-----
> From: RADHIKA.R.ROY@saic.com [mailto:RADHIKA.R.ROY@saic.com] 
> Sent: maandag 13 november 2006 17:08
> To: Georgios Karagiannis; Ash, Gerald R (Jerry), ALABS; 
> Roland Bless; David R Oran
> Cc: NSIS Group
> Subject: RE: [NSIS] Alternatives for supporting 
> domain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Georgios and all:
> 
> The fact is that all QOS model-specific extensions, 
> explanations, and references MUST be DROPPED per consensus of 
> this IETF meeting.  
> 
> In this way, the problems of all NON-interoperability among 
> the standard-based IETF QOS models (RSVP, IntServ, and 
> DiffServ) and so-called "private" non-IETF, non-standard QOS 
> models (Y.1541 and
> 3GPP/Y.1541) will be avoided in this standard-track "generic" 
> QSPEC RFC.
> 
> If anyone wants to avoid this consensus of the IETF meeting, 
> they MUST come up with the technical proposal how to solve 
> the NON-interoperability problems among the standard-based 
> IETF QOS models (RSVP, IntServ, and DiffServ) and so-called 
> "private" non-IETF, non-standard QOS models (Y.1541 and 
> 3GPP/Y.1541). If brute forces are used to add any references 
> of QOS model-specific things as the present draft stands, the 
> consensus of the IETF will be broken and the draft will NOT 
> move for IESG approval as we will be again going to the 
> circle of interoperability issues.
> 
> Best regards,
> Radhika
> 
> -----Original Message-----
> From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On 
> Behalf Of Georgios Karagiannis
> Sent: Monday, November 13, 2006 6:57 AM
> To: 'Ash, Gerald R (Jerry), ALABS'; 'Roland Bless'; 'David R Oran'
> Cc: 'NSIS Group'
> Subject: RE: [NSIS] Alternatives for supporting 
> domain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Hi Jerry
> 
> I agree with you!
> I think that we now have reached a consensus on all open 
> issues and we can proceed on finalizing all NSIS base drafts!
> 
> Best Regards,
> Georgios 
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 12:36:49 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gjfil-0004RY-Kb; Mon, 13 Nov 2006 12:35:47 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gjfik-0004Og-WB
	for nsis@ietf.org; Mon, 13 Nov 2006 12:35:47 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gjfij-0002xx-FR
	for nsis@ietf.org; Mon, 13 Nov 2006 12:35:46 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kADHZaC3014556; Mon, 13 Nov 2006 18:35:40 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: <RADHIKA.R.ROY@saic.com>,
	"'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>,
	"'Roland Bless'" <bless@tm.uka.de>, "'David R Oran'" <oran@cisco.com>
Subject: RE: [NSIS] Alternatives for
	supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 18:35:31 +0100
Message-ID: <003f01c7074a$20c9d120$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <62D92A9A02BCC845B202323D49A48426230EE9@0591-ITS-EXMP02.us.saic.com>
Thread-Index: AccHSM1+BTZUVTP8SiK2Ajsg9hVRwQAAGZ3w
X-Spam-Score: 0.3 () FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Mon, 13 Nov 2006 18:35:41 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 156eddb66af16eef49a76ae923b15b92
Cc: 'NSIS Group' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Radhika

The optional parameters are just included to simplify the 
parameter encoding that can be used by the QoS models supported by NSIS.
Suppose that a QoS model needs to use the PHB class as parameter, then
instead of 
reinventing its own encoding for this parameter, the encoding 
provided in the QSPEC template draft can be used.

Thus, the QSPEC template provides guidelines on how the encoding of such
parameters 
should be accomplished. I think that this is something usefull!

Best Regards,
Georgios


 

> -----Original Message-----
> From: RADHIKA.R.ROY@saic.com [mailto:RADHIKA.R.ROY@saic.com] 
> Sent: maandag 13 november 2006 18:13
> To: Georgios Karagiannis; RADHIKA.R.ROY@saic.com; Ash, Gerald 
> R (Jerry), ALABS; Roland Bless; David R Oran
> Cc: NSIS Group
> Subject: RE: [NSIS] Alternatives for 
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Georgios:
> 
> This is only true for the standard-based IETF QOS Models 
> (RSVP, IntServ, and DiffServ). 
> 
> All other so-called QOS models (non-standards DSTE, Y.1541, and
> 3GPP/Y.1541) have no concept of Token Bucket as they do not 
> have QOS signaling messages defined of their own other than 
> some values of delays, jitter, and packet losses. It happens 
> to be, like all SLAs, these implementation-specific 
> value-based QOS models can also be mapped into DiffServ QOS messages.
> 
> The value-based QOS models are ONLY applicable for 
> implementation agreements (e.g., OSP). For example, some 
> specific values of some parameters of DiffServ need to be 
> agreed among some implementers or ISPs, and they will develop 
> SLAs for this. This does not mean we will include all SLAs of 
> the whole world as a part of this "generic" standard QSPEC RFC.
> 
> In this respect, specific value-based QOS models like DSTE, 
> Y.1541, and
> 3GPP/Y.1541 are nothing but some value-based QOS parameters 
> of "service level agreements (SLAs).
> 
> If needed, any reference to DiffServ QOS model also needs be 
> taken off from this standard-track RFC. The token bucket is 
> good enough to be used by DiffServ, RSVP, and IntServ as they 
> are the QOS models per standards and have their own QOS 
> signaling messages that will provide "interoperability."
> 
> This standard track RFC even does NOT need to include any 
> references to DiffServ QOS model as well. Again, if all SLAs 
> are mapped into DiffServ, that does not mean all 
> implementation-specific value-based SLAs are QOS models and 
> then we should start including one SLA after another into 
> this standard track RFC. The inclusion of DiffServ QOS model 
> seems to us as a "ploy" for inclusion of other SLA-type QOS things.
> 
> What does "force" the NSIS WG to "mess" the whole 
> standard-track RFC after all these hard works "forcing" the 
> same non-interoperability issues mixing between apples and 
> oranges in the same draft?
> 
> Why does the NSIS WG not write separate individual draft for 
> each QOS model (RSVP, IntServ, DiffServ, DSTE, Y.1541, 
> 3GPP/Y.1541) how to use the QSPEC using NSIS QOS signaling messages?
> 
> Best regards,
> Radhika
> 
> -----Original Message-----
> From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On 
> Behalf Of Georgios Karagiannis
> Sent: Monday, November 13, 2006 11:33 AM
> To: RADHIKA.R.ROY@saic.com; 'Ash, Gerald R (Jerry), ALABS'; 
> 'Roland Bless'; 'David R Oran'
> Cc: 'NSIS Group'
> Subject: RE: [NSIS] Alternatives for
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Hi Roy
> 
> There are nomore interoperability issues since the only 
> parameter that can be considered as mandatory is the Token 
> Bucket parameter, which is used as the traffic source description.
> 
> Best regards,
> Georgios
> 
> 
> 
> > -----Original Message-----
> > From: RADHIKA.R.ROY@saic.com [mailto:RADHIKA.R.ROY@saic.com]
> > Sent: maandag 13 november 2006 17:08
> > To: Georgios Karagiannis; Ash, Gerald R (Jerry), ALABS; 
> Roland Bless; 
> > David R Oran
> > Cc: NSIS Group
> > Subject: RE: [NSIS] Alternatives for supporting 
> > domain-localadmissioncontroloperation in NSLP/QSPEC
> > 
> > Georgios and all:
> > 
> > The fact is that all QOS model-specific extensions, 
> explanations, and 
> > references MUST be DROPPED per consensus of this IETF meeting.
> > 
> > In this way, the problems of all NON-interoperability among the 
> > standard-based IETF QOS models (RSVP, IntServ, and
> > DiffServ) and so-called "private" non-IETF, non-standard QOS models 
> > (Y.1541 and
> > 3GPP/Y.1541) will be avoided in this standard-track "generic" 
> > QSPEC RFC.
> > 
> > If anyone wants to avoid this consensus of the IETF 
> meeting, they MUST 
> > come up with the technical proposal how to solve the 
> > NON-interoperability problems among the standard-based IETF 
> QOS models 
> > (RSVP, IntServ, and DiffServ) and so-called "private" non-IETF, 
> > non-standard QOS models (Y.1541 and 3GPP/Y.1541). If brute 
> forces are 
> > used to add any references of QOS model-specific things as 
> the present 
> > draft stands, the consensus of the IETF will be broken and 
> the draft 
> > will NOT move for IESG approval as we will be again going to the 
> > circle of interoperability issues.
> > 
> > Best regards,
> > Radhika
> > 
> > -----Original Message-----
> > From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] 
> On Behalf 
> > Of Georgios Karagiannis
> > Sent: Monday, November 13, 2006 6:57 AM
> > To: 'Ash, Gerald R (Jerry), ALABS'; 'Roland Bless'; 'David R Oran'
> > Cc: 'NSIS Group'
> > Subject: RE: [NSIS] Alternatives for supporting 
> > domain-localadmissioncontroloperation in NSLP/QSPEC
> > 
> > Hi Jerry
> > 
> > I agree with you!
> > I think that we now have reached a consensus on all open 
> issues and we 
> > can proceed on finalizing all NSIS base drafts!
> > 
> > Best Regards,
> > Georgios
> > 
> 
> 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 13:21:40 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjgPy-0003YK-N2; Mon, 13 Nov 2006 13:20:26 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjgPw-0003Jv-DV
	for nsis@ietf.org; Mon, 13 Nov 2006 13:20:24 -0500
Received: from mclmx2.mail.saic.com ([149.8.64.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GjgPu-0000W8-V9
	for nsis@ietf.org; Mon, 13 Nov 2006 13:20:24 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx2.mail.saic.com; Mon, 13 Nov 2006 13:18:55 -0500
Received: from 0015-ITS-EXBH01.us.saic.com ([10.43.229.18])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006111313185408328 ; Mon, 13 Nov 2006 13:18:55 -0500
Received: from 0591-ITS-EXMP02.us.saic.com ([10.42.81.248]) by
	0015-ITS-EXBH01.us.saic.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 13 Nov 2006 13:18:52 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for
	supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
Date: Mon, 13 Nov 2006 13:18:51 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EEA@0591-ITS-EXMP02.us.saic.com>
In-Reply-To: <003f01c7074a$20c9d120$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for
	supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
Thread-Index: AccHSM1+BTZUVTP8SiK2Ajsg9hVRwQAAGZ3wAABYu1A=
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: <karagian@cs.utwente.nl>, <RADHIKA.R.ROY@saic.com>,
	"Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>,
	"Roland Bless" <bless@tm.uka.de>, "David R Oran" <oran@cisco.com>
X-OriginalArrivalTime: 13 Nov 2006 18:18:52.0993 (UTC)
	FILETIME=[2C929310:01C70750]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 325b777e1a3a618c889460b612a65510
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios:

I think that we understand each other clearly and have agreement as
well. What we differ is where and how the works need to be done.

Kindly see Inline [RRR]

Best regards,
Radhika

-----Original Message-----
From: karagian@cs.utwente.nl [mailto:karagian@cs.utwente.nl]=20
Sent: Monday, November 13, 2006 12:36 PM
To: RADHIKA.R.ROY@saic.com; 'Ash, Gerald R (Jerry), ALABS'; 'Roland
Bless'; 'David R Oran'
Cc: 'NSIS Group'
Subject: RE: [NSIS] Alternatives for
supportingdomain-localadmissioncontroloperation in NSLP/QSPEC

Hi Radhika

The optional parameters are just included to simplify the=20
parameter encoding that can be used by the QoS models supported by NSIS.

[RRR] The "generic" texts without referring any particular QOS model as
stated above are always OK for this "generic" RFC.

Suppose that a QoS model needs to use the PHB class as parameter, then
instead of=20
reinventing its own encoding for this parameter, the encoding=20
provided in the QSPEC template draft can be used.

[RRR] Again, it is also OK per principles stated above. However, it is
NOT OK, to provide some extensions of some specific QOS models as a part
of this generic standard track RFC. It is out-of-scope for this RFC.

[RRR] Moreover, the inflexible tightly integrated DSTE, Y.1541, and
3GPP/Y1415 value-based QOS classes are so problematic that it will
almost non-usable for any practical purposes unless some extensions in
QOS signaling messages are created for them as they do not have any
signaling messages of their own. In fact, they do not need one because
all they want to send some values. What it means that any references of
these QOS models with some extensions here and here are also incomplete.
In all cases, a separate draft is needed for each of these Value-based
QOS models.

[RRR] The question is this: Why do then these editors use brute forces
for inclusion of DSTE, Y.1541, and 3GPP/Y1415 in this draft? The answer
is this: To obtain some quasi-legitimacy of being standards - to be
accepted by the whole world. As a result, the non-interoperability has
become mandatory. Why should NSIS WG pay the price of
non-interoperability?

Thus, the QSPEC template provides guidelines on how the encoding of such
parameters=20
should be accomplished. I think that this is something useful!

[RRR] Encoding of QSPEC template for each QOS model needs to be done in
separate individual drafts as described above. Definitely it is the most
useful part for creating standards. There are a lot of works involved:
End-to-End negotiations and detail call flows. If you want deal with all
those things in this RFC, let us include these things in this generic
draft!!! Indirectly, interoperability issues will also come. Do you like
to include this as well?

[RRR] The way end-to-end negotiations to be done in RSVP, IntServ, and
DiffServ will be fundamentally different from those of the value-based
DSTE, Y.1541, and 3GPP/Y.1541. On the top of this, you can use SIP as an
example application. You can how the complex the QOS signaling message
flows is. However, we have no problems with RSVP, IntServ, and DiffServ.

[RRR] Conclusion: It is not that we cannot include some encoding related
to some QOS-specific models, but these need to done separately with due
examples and detail call flows. That is why, any specific references to
any QOS models will be "incomplete" and will create more questions than
answering them in this generic QSPEC RFC.

Best Regards,
Georgios


=20

> -----Original Message-----
> From: RADHIKA.R.ROY@saic.com [mailto:RADHIKA.R.ROY@saic.com]=20
> Sent: maandag 13 november 2006 18:13
> To: Georgios Karagiannis; RADHIKA.R.ROY@saic.com; Ash, Gerald=20
> R (Jerry), ALABS; Roland Bless; David R Oran
> Cc: NSIS Group
> Subject: RE: [NSIS] Alternatives for=20
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>=20
> Georgios:
>=20
> This is only true for the standard-based IETF QOS Models=20
> (RSVP, IntServ, and DiffServ).=20
>=20
> All other so-called QOS models (non-standards DSTE, Y.1541, and
> 3GPP/Y.1541) have no concept of Token Bucket as they do not=20
> have QOS signaling messages defined of their own other than=20
> some values of delays, jitter, and packet losses. It happens=20
> to be, like all SLAs, these implementation-specific=20
> value-based QOS models can also be mapped into DiffServ QOS messages.
>=20
> The value-based QOS models are ONLY applicable for=20
> implementation agreements (e.g., OSP). For example, some=20
> specific values of some parameters of DiffServ need to be=20
> agreed among some implementers or ISPs, and they will develop=20
> SLAs for this. This does not mean we will include all SLAs of=20
> the whole world as a part of this "generic" standard QSPEC RFC.
>=20
> In this respect, specific value-based QOS models like DSTE,=20
> Y.1541, and
> 3GPP/Y.1541 are nothing but some value-based QOS parameters=20
> of "service level agreements (SLAs).
>=20
> If needed, any reference to DiffServ QOS model also needs be=20
> taken off from this standard-track RFC. The token bucket is=20
> good enough to be used by DiffServ, RSVP, and IntServ as they=20
> are the QOS models per standards and have their own QOS=20
> signaling messages that will provide "interoperability."
>=20
> This standard track RFC even does NOT need to include any=20
> references to DiffServ QOS model as well. Again, if all SLAs=20
> are mapped into DiffServ, that does not mean all=20
> implementation-specific value-based SLAs are QOS models and=20
> then we should start including one SLA after another into=20
> this standard track RFC. The inclusion of DiffServ QOS model=20
> seems to us as a "ploy" for inclusion of other SLA-type QOS things.
>=20
> What does "force" the NSIS WG to "mess" the whole=20
> standard-track RFC after all these hard works "forcing" the=20
> same non-interoperability issues mixing between apples and=20
> oranges in the same draft?
>=20
> Why does the NSIS WG not write separate individual draft for=20
> each QOS model (RSVP, IntServ, DiffServ, DSTE, Y.1541,=20
> 3GPP/Y.1541) how to use the QSPEC using NSIS QOS signaling messages?
>=20
> Best regards,
> Radhika
>=20
> -----Original Message-----
> From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On=20
> Behalf Of Georgios Karagiannis
> Sent: Monday, November 13, 2006 11:33 AM
> To: RADHIKA.R.ROY@saic.com; 'Ash, Gerald R (Jerry), ALABS';=20
> 'Roland Bless'; 'David R Oran'
> Cc: 'NSIS Group'
> Subject: RE: [NSIS] Alternatives for
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>=20
> Hi Roy
>=20
> There are nomore interoperability issues since the only=20
> parameter that can be considered as mandatory is the Token=20
> Bucket parameter, which is used as the traffic source description.
>=20
> Best regards,
> Georgios
>=20
>=20
>=20
> > -----Original Message-----
> > From: RADHIKA.R.ROY@saic.com [mailto:RADHIKA.R.ROY@saic.com]
> > Sent: maandag 13 november 2006 17:08
> > To: Georgios Karagiannis; Ash, Gerald R (Jerry), ALABS;=20
> Roland Bless;=20
> > David R Oran
> > Cc: NSIS Group
> > Subject: RE: [NSIS] Alternatives for supporting=20
> > domain-localadmissioncontroloperation in NSLP/QSPEC
> >=20
> > Georgios and all:
> >=20
> > The fact is that all QOS model-specific extensions,=20
> explanations, and=20
> > references MUST be DROPPED per consensus of this IETF meeting.
> >=20
> > In this way, the problems of all NON-interoperability among the=20
> > standard-based IETF QOS models (RSVP, IntServ, and
> > DiffServ) and so-called "private" non-IETF, non-standard QOS models=20
> > (Y.1541 and
> > 3GPP/Y.1541) will be avoided in this standard-track "generic"=20
> > QSPEC RFC.
> >=20
> > If anyone wants to avoid this consensus of the IETF=20
> meeting, they MUST=20
> > come up with the technical proposal how to solve the=20
> > NON-interoperability problems among the standard-based IETF=20
> QOS models=20
> > (RSVP, IntServ, and DiffServ) and so-called "private" non-IETF,=20
> > non-standard QOS models (Y.1541 and 3GPP/Y.1541). If brute=20
> forces are=20
> > used to add any references of QOS model-specific things as=20
> the present=20
> > draft stands, the consensus of the IETF will be broken and=20
> the draft=20
> > will NOT move for IESG approval as we will be again going to the=20
> > circle of interoperability issues.
> >=20
> > Best regards,
> > Radhika
> >=20
> > -----Original Message-----
> > From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org]=20
> On Behalf=20
> > Of Georgios Karagiannis
> > Sent: Monday, November 13, 2006 6:57 AM
> > To: 'Ash, Gerald R (Jerry), ALABS'; 'Roland Bless'; 'David R Oran'
> > Cc: 'NSIS Group'
> > Subject: RE: [NSIS] Alternatives for supporting=20
> > domain-localadmissioncontroloperation in NSLP/QSPEC
> >=20
> > Hi Jerry
> >=20
> > I agree with you!
> > I think that we now have reached a consensus on all open=20
> issues and we=20
> > can proceed on finalizing all NSIS base drafts!
> >=20
> > Best Regards,
> > Georgios
> >=20
>=20
>=20
>=20
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>=20


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 13 20:33:21 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjnA7-0004DJ-Mx; Mon, 13 Nov 2006 20:32:31 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gjn6t-0001y2-P1
	for nsis@ietf.org; Mon, 13 Nov 2006 20:29:11 -0500
Received: from difra-computer.de ([81.169.157.103] helo=www.difra-computer.de)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gjn1Q-0002HL-8Y
	for nsis@ietf.org; Mon, 13 Nov 2006 20:23:33 -0500
Received: from win2003.dickmann2003.home (p548A7218.dip.t-dialin.net
	[84.138.114.24])
	by www.difra-computer.de (Postfix) with ESMTP id 7B4A33E402D
	for <nsis@ietf.org>; Tue, 14 Nov 2006 02:21:05 +0100 (CET)
Received: from [127.0.0.1] ([192.168.28.2]) by win2003.dickmann2003.home with
	Microsoft SMTPSVC(6.0.3790.1830); Tue, 14 Nov 2006 02:09:08 +0100
Message-ID: <45591984.2010904@christian-dickmann.de>
Date: Mon, 13 Nov 2006 20:19:00 -0500
From: Christian Dickmann <mail@christian-dickmann.de>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: nsis@ietf.org
Content-Type: text/plain; charset=ISO-8859-15; format=flowed
Content-Transfer-Encoding: 7bit
X-OriginalArrivalTime: 14 Nov 2006 01:09:08.0439 (UTC)
	FILETIME=[7C858E70:01C70789]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 68c8cc8a64a9d0402e43b8eee9fc4199
Subject: [NSIS] GIST API primitive SetStateLifetime
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi,

I have a question regarding the SetStateLifetime ( NSLPID, MRI, 
State-Lifetime ) primitive of the GIST API.
Why is there no session ID present here? The key for GIST routing state 
is the 3-tupel (NSLPID, MRI, SID),
thus the SetStateLifetime primitive potentially effects multiple routing 
states at a time.

Regards,
Christian Dickmann


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 08:48:47 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gjydn-00070x-Eo; Tue, 14 Nov 2006 08:47:55 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gjydm-00070s-Ak
	for nsis@ietf.org; Tue, 14 Nov 2006 08:47:54 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gjydk-0005En-1a
	for nsis@ietf.org; Tue, 14 Nov 2006 08:47:54 -0500
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gjyda-0007wz-5Z; Tue, 14 Nov 2006 14:47:43 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id F3058B534;
	Tue, 14 Nov 2006 14:47:41 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GjydZ-00046u-En; Tue, 14 Nov 2006 14:47:41 +0100
Message-ID: <4559C8FC.5000103@tm.uka.de>
Date: Tue, 14 Nov 2006 14:47:40 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] Alternatives for
	supporting	domain-localadmissioncontroloperation in NSLP/QSPEC
References: <002f01c70741$61c2e620$4c0d5982@dynamic.ewi.utwente.nl>
In-Reply-To: <002f01c70741$61c2e620$4c0d5982@dynamic.ewi.utwente.nl>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -2.9 (--)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	1.5 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 1ac7cc0a4cd376402b85bc1961a86ac2
Cc: "'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>, RADHIKA.R.ROY@saic.com,
	'David R Oran' <oran@cisco.com>, 'NSIS Group' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

Georgios Karagiannis wrote:
> There are nomore interoperability issues since the only parameter 
> that can be considered as mandatory is the Token Bucket parameter, 
> which is used as the traffic source description.

Yep. But as also already discussed, specifying a TMOD/TSPEC only is
pretty useless, i.e. you'll need to specify constraints. Otherwise,
best effort is sufficient and you can save your QoS signaling
completely.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 10:12:46 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GjzxC-0004UN-QD; Tue, 14 Nov 2006 10:12:02 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GjzxB-0004UI-LJ
	for nsis@ietf.org; Tue, 14 Nov 2006 10:12:01 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GjzxA-0000Bc-9h
	for nsis@ietf.org; Tue, 14 Nov 2006 10:12:01 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kAEFBDtO026083; Tue, 14 Nov 2006 16:11:31 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Roland Bless'" <bless@tm.uka.de>
Subject: RE: [NSIS] Alternatives for
	supporting	domain-localadmissioncontroloperation in NSLP/QSPEC
Date: Tue, 14 Nov 2006 16:11:08 +0100
Message-ID: <003001c707ff$22a78f90$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <4559C8FC.5000103@tm.uka.de>
Thread-Index: AccH83eypQqqXd6RR9+uR+gajucafQAC5v3g
X-Spam-Score: 0.3 () FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Tue, 14 Nov 2006 16:11:31 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: bb8f917bb6b8da28fc948aeffb74aa17
Cc: "'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>, RADHIKA.R.ROY@saic.com,
	'David R Oran' <oran@cisco.com>, 'NSIS Group' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland

Agreed!

Best Regards,
Georgios 

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: dinsdag 14 november 2006 14:48
> To: Georgios Karagiannis
> Cc: RADHIKA.R.ROY@saic.com; 'Ash, Gerald R (Jerry), ALABS'; 
> 'David R Oran'; 'NSIS Group'
> Subject: Re: [NSIS] Alternatives for supporting 
> domain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Hi Georgios,
> 
> Georgios Karagiannis wrote:
> > There are nomore interoperability issues since the only 
> parameter that 
> > can be considered as mandatory is the Token Bucket 
> parameter, which is 
> > used as the traffic source description.
> 
> Yep. But as also already discussed, specifying a TMOD/TSPEC 
> only is pretty useless, i.e. you'll need to specify 
> constraints. Otherwise, best effort is sufficient and you can 
> save your QoS signaling completely.
> 
> Regards,
>  Roland
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 11:13:54 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk0uP-0005VS-Cd; Tue, 14 Nov 2006 11:13:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gk0uN-0005Ub-W4
	for nsis@ietf.org; Tue, 14 Nov 2006 11:13:11 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gk0uM-0002AX-2S
	for nsis@ietf.org; Tue, 14 Nov 2006 11:13:11 -0500
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gk0uJ-0002xg-Q3; Tue, 14 Nov 2006 17:13:09 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 588D4B584;
	Tue, 14 Nov 2006 17:13:07 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gk0uI-0004xk-VH; Tue, 14 Nov 2006 17:13:07 +0100
Message-ID: <4559EB12.3040900@tm.uka.de>
Date: Tue, 14 Nov 2006 17:13:06 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
Subject: Re: [NSIS] Summary of forthcoming QoS NSLP changes
References: <Pine.LNX.4.64.0611101911080.14922@sbz-30.cs.Helsinki.FI>
In-Reply-To: <Pine.LNX.4.64.0611101911080.14922@sbz-30.cs.Helsinki.FI>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -3.0 (---)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	1.4 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 10ba05e7e8a9aa6adb025f426bef3a30
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi Jukka,

Jukka MJ Manner wrote:
> Here is a list of the changes I intend to do to the current QoS NSLP spec. 
> After those, we are ready for WGLC. The spec is looking good. If there are 
> no arguments, I could do these within a week.

Some additions for clarification. :-)

> - Make reduced refreshes the default behavior, add a new error code for 
>   "full refreshes required".

We already have an error message saying "full QSPEC required" that
is used when a new node on the path was hit. "full refreshes required"
should be used when a node doesn't support reduced refreshes (and
the sending node should remember that for subsequent refreshes), so
the above description may be still a bit too similar to the already
existing one.

> - Some clarifications needed for proxy mode (and GIST),
>   object order (SHOULD use a certain, MUST accept in any order), tear 
>   operation, response/notify (RESPONSE is triggered by an incoming 
>   message, NOTIFY is fully asynchronous)
> 
> - Summary refresh messages: one NSLP message can refresh several 
>   sessions (use a new object with SID+RSN pairs)

This would require that GIST implementations follow the SHOULD
in:
                                 Note that GIST if possible SHOULD use
   the same SII-Handle for multiple sessions to the same peer, since
   this then allows signaling applications to aggregate some signaling,
   such as summary refreshes or bulk teardowns.

if the implementation does not, summary refreshes are not possible
(but I guess we don't need an explicit hint from GIST about that
feature, because the NSLP will not find any identical SII-handles?

> - Summary tear: use a new object with SIDs only.
> 
> - RSN MUST NOT be interpreted when receiving a tear. If two QNEs are not 
>   in synch, what can you? Update the reservation, then tear? What if the 
>   update fails? There is no reason for QNEs to be in synch, you just 
>   remove the resources you had allocated.

Yes, we discussed these issues in the hallway...maybe someone has
another view?

> - Re-introduce the ack-flag. This is needed in order to do diagnostics, 
>   say, you want to make sure the next peer actually processed your message 
>   and understood it. Also need to add a new error code that tells the 
^^^^^^^^^^^^^^^^^^^^^^
Sorry to jump in, but that very text was already in there as a rationale
and I didn't get the point in the past. The essential point is that you
want to detect whether the QoS NSLP peer application is still alive
or not while GIST continues to work. It's not about reliable
transmission, it's about a faster dead peer diagnosis?
A further point was how and when
to detect the missing response. Currently retransmissions (sec. 5.2.4.)
are only defined on an end-to-end basis, not between QNEs.
So we must define approriate mechanisms, probably only
for refreshes w/o RII set (these usually go only between
adjacent QNEs).

>   QNI if the "path is truncated", i.e., a peer is responding to the Ack. 

More specifically: a QNE should notify its upstream QNE if its
downstream QNE is probably dead (not responding to that ACK)

>   It is a policy decision on a QNE whether it should send e.g. an 
>   Ack-flaged refresh when it receives such a refresh from its upstream 
>   peer. 

Hm... Usually an incoming refresh (w/o RII set) doesn't trigger
an new one downstream. Or do you mean only the ACK flag setting?

> - [After concensus] Update the draft on "stacking".

Yep.

> Yell now if some of these are not right things to do.

Yelling...

> You could also yell, if these are, that would be nice, too. :)

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 11:14:34 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk0v9-0005bs-PF; Tue, 14 Nov 2006 11:13:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gk0v7-0005bW-9d
	for nsis@ietf.org; Tue, 14 Nov 2006 11:13:57 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gk0v5-0002Gv-QO
	for nsis@ietf.org; Tue, 14 Nov 2006 11:13:57 -0500
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gk0us-0002zV-6u; Tue, 14 Nov 2006 17:13:47 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id AF54BB584;
	Tue, 14 Nov 2006 17:13:41 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gk0ur-0004xo-Dz; Tue, 14 Nov 2006 17:13:41 +0100
Message-ID: <4559EB34.9080808@tm.uka.de>
Date: Tue, 14 Nov 2006 17:13:40 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Subject: Re: [NSIS] Alternatives for supporting domain-local
	admission	controloperation in NSLP/QSPEC
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F76@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F76@KCCLUST06EVS1.ugd.att.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -3.0 (---)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	1.4 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f4c2cf0bccc868e4cc88dace71fb3f44
Cc: David R Oran <oran@cisco.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry,

Ash, Gerald R (Jerry), ALABS wrote:
> I agree that we should go with option 2 (encapsulated QSPEC) versus
> option 1 (stacked QSPEC) as the easier one to implement (according to
> discussions with Roland).  We should also retain option 3 (tunneled
> QSPEC) as currently specified.

Ok, great. However, I'm still interested if someone comes up
with a need for real stacking in QoS NSLP.

> IMO we should retain the QoS class parameters as optional parameters in
> the QSPEC document.  One reason is that support for DiffServ was an
> original basic requirement on the QSPEC, and that requirement still must
> be met.  Additionally, it would require initiating new QOSM
> specification documents if these parameters are moved elsewhere -- it is
> far simpler, and appropriate, to retain them as is in the QSPEC
> document.

I don't agree here, as already expressed in the meeting. I'd really
like to see different QSPEC definitions for internal signaling if
required. The QoSM/QSPEC documents don't need to be long and they
could give guidelines on the mapping from the "generic" QSPEC to it's
specific classes. Furthermore, we already have drafts for This would be
IMHO a cleaner approach. I really
don't see a benefit in signaling these parameters from end to end.

There are several reasons for this (but I may have only a very
limited view, so please other people like operators: comment
on that):
As a provider I would not trust the end-system to choose the correct
Diffserv class (PHB). Furthermore, this will vary depending on the
providers network, e.g., although two different domains use DiffServ,
they may use a totally different service provisioning, e.g. one uses
6 classes, the other 12 classes. I don't really think that the
end-system should try to get this mapping correct. The end-system
should stick to the generic QSPEC (TMOD+constraints) that can be
mapped by the edge router to the according DiffServ classes, which
are supported in the particular DiffServ domain.
Consequently, if a domain attaches such a class parameter for
internal use, it is of limited use for other domains. So my guess
is that leaving these QoSM-specific parameters in the spec doesn't
help interoperability that much.

If there is, however, consensus for keeping the QoS class parameters
in the generic QSPEC as optional, I'd like to see something like a
"scope" parameter flag like Jukka proposed.

> Hopefully we are now ready to revise the QSPEC document according to the
> agreements summarized in the presentation at IETF-67
> https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=67
> (subject to the revisions of including option 2/encapsulated QSPEC).

As expressed in the meeting and above, I agree with the
presentation for the most part of it.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 11:25:04 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk14t-0002ys-QU; Tue, 14 Nov 2006 11:24:03 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gk14s-0002yO-GJ
	for nsis@ietf.org; Tue, 14 Nov 2006 11:24:02 -0500
Received: from mail.gmx.net ([213.165.64.20])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gk14r-0004B7-1m
	for nsis@ietf.org; Tue, 14 Nov 2006 11:24:02 -0500
Received: (qmail invoked by alias); 14 Nov 2006 16:23:59 -0000
Received: from p549849D0.dip.t-dialin.net (EHLO [192.168.2.33]) [84.152.73.208]
	by mail.gmx.net (mp027) with SMTP; 14 Nov 2006 17:23:59 +0100
X-Authenticated: #29516787
Message-ID: <4559EDA4.1030903@gmx.net>
Date: Tue, 14 Nov 2006 17:24:04 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: Roland Bless <bless@tm.uka.de>
Subject: Re: [NSIS] Alternatives for supporting
	domain-local	admission	controloperation in NSLP/QSPEC
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F76@KCCLUST06EVS1.ugd.att.com>
	<4559EB34.9080808@tm.uka.de>
In-Reply-To: <4559EB34.9080808@tm.uka.de>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cd26b070c2577ac175cd3a6d878c6248
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>,
	David R Oran <oran@cisco.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland,

I agree with your observation regarding the QoS class parameters. I 
think it would be much cleaner to remove them from the QSPEC document as 
I commented already in the past. I don't know how complicated the 
"scope" flag is but I would appreciate not to introduce additional 
complexity into protocols.

I also agree with your comment about stacking of QSPECs. An analysis 
would be appropriate.

Ciao
Hannes

Roland Bless wrote:
> Hi Jerry,
> 
> Ash, Gerald R (Jerry), ALABS wrote:
>> I agree that we should go with option 2 (encapsulated QSPEC) versus
>> option 1 (stacked QSPEC) as the easier one to implement (according to
>> discussions with Roland).  We should also retain option 3 (tunneled
>> QSPEC) as currently specified.
> 
> Ok, great. However, I'm still interested if someone comes up
> with a need for real stacking in QoS NSLP.
> 
>> IMO we should retain the QoS class parameters as optional parameters in
>> the QSPEC document.  One reason is that support for DiffServ was an
>> original basic requirement on the QSPEC, and that requirement still must
>> be met.  Additionally, it would require initiating new QOSM
>> specification documents if these parameters are moved elsewhere -- it is
>> far simpler, and appropriate, to retain them as is in the QSPEC
>> document.
> 
> I don't agree here, as already expressed in the meeting. I'd really
> like to see different QSPEC definitions for internal signaling if
> required. The QoSM/QSPEC documents don't need to be long and they
> could give guidelines on the mapping from the "generic" QSPEC to it's
> specific classes. Furthermore, we already have drafts for This would be
> IMHO a cleaner approach. I really
> don't see a benefit in signaling these parameters from end to end.
> 
> There are several reasons for this (but I may have only a very
> limited view, so please other people like operators: comment
> on that):
> As a provider I would not trust the end-system to choose the correct
> Diffserv class (PHB). Furthermore, this will vary depending on the
> providers network, e.g., although two different domains use DiffServ,
> they may use a totally different service provisioning, e.g. one uses
> 6 classes, the other 12 classes. I don't really think that the
> end-system should try to get this mapping correct. The end-system
> should stick to the generic QSPEC (TMOD+constraints) that can be
> mapped by the edge router to the according DiffServ classes, which
> are supported in the particular DiffServ domain.
> Consequently, if a domain attaches such a class parameter for
> internal use, it is of limited use for other domains. So my guess
> is that leaving these QoSM-specific parameters in the spec doesn't
> help interoperability that much.
> 
> If there is, however, consensus for keeping the QoS class parameters
> in the generic QSPEC as optional, I'd like to see something like a
> "scope" parameter flag like Jukka proposed.
> 
>> Hopefully we are now ready to revise the QSPEC document according to the
>> agreements summarized in the presentation at IETF-67
>> https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=67
>> (subject to the revisions of including option 2/encapsulated QSPEC).
> 
> As expressed in the meeting and above, I agree with the
> presentation for the most part of it.
> 
> Regards,
>  Roland
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 11:27:14 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk17P-0005lS-O6; Tue, 14 Nov 2006 11:26:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gk17P-0005kr-0H
	for nsis@ietf.org; Tue, 14 Nov 2006 11:26:39 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gk17M-0004i2-4D
	for nsis@ietf.org; Tue, 14 Nov 2006 11:26:38 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Tue, 14 Nov 2006 18:26:31 +0200
	id 000AFD49.4559EE37.000076E5
Date: Tue, 14 Nov 2006 18:26:27 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: Roland Bless <bless@tm.uka.de>
Subject: Re: [NSIS] Summary of forthcoming QoS NSLP changes
In-Reply-To: <4559EB12.3040900@tm.uka.de>
Message-ID: <Pine.LNX.4.64.0611141820530.21037@sbz-31.cs.Helsinki.FI>
References: <Pine.LNX.4.64.0611101911080.14922@sbz-30.cs.Helsinki.FI>
	<4559EB12.3040900@tm.uka.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 825e642946eda55cd9bc654a36dab8c2
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi,

comments (acks) below. Thanks for the feedback and clarifications to my 
list.

Jukka

On Tue, 14 Nov 2006, Roland Bless wrote:

> 
> Hi Jukka,
> 
> Jukka MJ Manner wrote:
> > Here is a list of the changes I intend to do to the current QoS NSLP spec. 
> > After those, we are ready for WGLC. The spec is looking good. If there are 
> > no arguments, I could do these within a week.
> 
> Some additions for clarification. :-)
> 
> > - Make reduced refreshes the default behavior, add a new error code for 
> >   "full refreshes required".
> 
> We already have an error message saying "full QSPEC required" that
> is used when a new node on the path was hit. "full refreshes required"
> should be used when a node doesn't support reduced refreshes (and
> the sending node should remember that for subsequent refreshes), so
> the above description may be still a bit too similar to the already
> existing one.

That is what I meant, was just too lazy to write a long description. :)

> 
> > - Some clarifications needed for proxy mode (and GIST),
> >   object order (SHOULD use a certain, MUST accept in any order), tear 
> >   operation, response/notify (RESPONSE is triggered by an incoming 
> >   message, NOTIFY is fully asynchronous)
> > 
> > - Summary refresh messages: one NSLP message can refresh several 
> >   sessions (use a new object with SID+RSN pairs)
> 
> This would require that GIST implementations follow the SHOULD
> in:
>                                  Note that GIST if possible SHOULD use
>    the same SII-Handle for multiple sessions to the same peer, since
>    this then allows signaling applications to aggregate some signaling,
>    such as summary refreshes or bulk teardowns.
> 
> if the implementation does not, summary refreshes are not possible
> (but I guess we don't need an explicit hint from GIST about that
> feature, because the NSLP will not find any identical SII-handles?
> 

Yes.

> > - Summary tear: use a new object with SIDs only.
> > 
> > - RSN MUST NOT be interpreted when receiving a tear. If two QNEs are not 
> >   in synch, what can you? Update the reservation, then tear? What if the 
> >   update fails? There is no reason for QNEs to be in synch, you just 
> >   remove the resources you had allocated.
> 
> Yes, we discussed these issues in the hallway...maybe someone has
> another view?
> 
> > - Re-introduce the ack-flag. This is needed in order to do diagnostics, 
> >   say, you want to make sure the next peer actually processed your message 
> >   and understood it. Also need to add a new error code that tells the 
> ^^^^^^^^^^^^^^^^^^^^^^
> Sorry to jump in, but that very text was already in there as a rationale
> and I didn't get the point in the past. The essential point is that you
> want to detect whether the QoS NSLP peer application is still alive
> or not while GIST continues to work. It's not about reliable
> transmission, it's about a faster dead peer diagnosis?
> A further point was how and when
> to detect the missing response. Currently retransmissions (sec. 5.2.4.)
> are only defined on an end-to-end basis, not between QNEs.
> So we must define approriate mechanisms, probably only
> for refreshes w/o RII set (these usually go only between
> adjacent QNEs).

Yes. The retransmission procedure should be the same for e2e and p2p 
messages.

> 
> >   QNI if the "path is truncated", i.e., a peer is responding to the Ack. 
> 
> More specifically: a QNE should notify its upstream QNE if its
> downstream QNE is probably dead (not responding to that ACK)

And the upstream QNE should then send this notify further to the QNI.

> 
> >   It is a policy decision on a QNE whether it should send e.g. an 
> >   Ack-flaged refresh when it receives such a refresh from its upstream 
> >   peer. 
> 
> Hm... Usually an incoming refresh (w/o RII set) doesn't trigger
> an new one downstream. Or do you mean only the ACK flag setting?

I meant, a downstream QNE does not need to send a refresh using Ack, when 
its upstream QNE did that.

> 
> > - [After concensus] Update the draft on "stacking".
> 
> Yep.
> 
> > Yell now if some of these are not right things to do.
> 
> Yelling...
> 
> > You could also yell, if these are, that would be nice, too. :)
> 
> Regards,
>  Roland
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 15:02:25 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk4TB-0006lf-LD; Tue, 14 Nov 2006 15:01:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gk4TB-0006lZ-3M
	for nsis@ietf.org; Tue, 14 Nov 2006 15:01:21 -0500
Received: from mail120.messagelabs.com ([216.82.250.83])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gk4T8-0002fn-6u
	for nsis@ietf.org; Tue, 14 Nov 2006 15:01:21 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-14.tower-120.messagelabs.com!1163534474!14396995!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 9322 invoked from network); 14 Nov 2006 20:01:14 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-14.tower-120.messagelabs.com with SMTP;
	14 Nov 2006 20:01:14 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAEK1CBD023606; 
	Tue, 14 Nov 2006 15:01:14 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst12.ugd.att.com [135.38.164.89])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kAEK0fpK023381; Tue, 14 Nov 2006 15:01:05 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
Date: Tue, 14 Nov 2006 14:00:59 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F7C@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <4559EB34.9080808@tm.uka.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local admission
	control operation in NSLP/QSPEC
Thread-Index: AccIB+jyVb8i2/l9Qua18BAAdd7EsgAEQdBQ
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: "Roland Bless" <bless@tm.uka.de>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b280b4db656c3ca28dd62e5e0b03daa8
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>, NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland,

> > IMO we should retain the QoS class parameters as optional=20
> > parameters in
> > the QSPEC document.  One reason is that support for DiffServ was an
> > original basic requirement on the QSPEC, and that=20
> > requirement still must
> > be met.  Additionally, it would require initiating new QOSM
> > specification documents if these parameters are moved elsewhere
> > -- it is
> > far simpler, and appropriate, to retain them as is in the QSPEC
> > document.
=20
> I don't agree here, as already expressed in the meeting. I'd really
> like to see different QSPEC definitions for internal signaling if
> required. The QoSM/QSPEC documents don't need to be long and they
> could give guidelines on the mapping from the "generic" QSPEC to it's
> specific classes. Furthermore, we already have drafts for=20
> This would be IMHO a cleaner approach. I really
> don't see a benefit in signaling these parameters from end to end.
>=20
> There are several reasons for this (but I may have only a very
> limited view, so please other people like operators: comment
> on that):
> As a provider I would not trust the end-system to choose the correct
> Diffserv class (PHB). Furthermore, this will vary depending on the
> providers network, e.g., although two different domains use DiffServ,
> they may use a totally different service provisioning, e.g. one uses
> 6 classes, the other 12 classes. I don't really think that the
> end-system should try to get this mapping correct. The end-system
> should stick to the generic QSPEC (TMOD+constraints) that can be
> mapped by the edge router to the according DiffServ classes, which
> are supported in the particular DiffServ domain.
> Consequently, if a domain attaches such a class parameter for
> internal use, it is of limited use for other domains. So my guess
> is that leaving these QoSM-specific parameters in the spec doesn't
> help interoperability that much.

You have a point regarding signaling PHB class across domains using
inconsistent classes.  However IMO it should still be possible for a QNI
to signal a PHB class if called for by the QNI's QOSM, and in which
end-to-end interoperability/consistency may be achieved. =20

In addition, the Y.1541 parameter can be regarded as a 'summary'
constraint parameter encompassing the latency, jitter, PLR, and PER
constraints (as stated in the summary slides at IETF-67).  These
parameters are intended to be signaled end-to-end across multiple
domains.

Therefore the QNI needs to signal these end-to-end parameters in some
cases.

It seems that the proposal is to introduce a QoS NSLP specific 'generic
QOSM', and everybody must map to this new QOSM.  We have tried
oftentimes to standardize a 'generic QOSM' and have not succeeded.  OTOH
we have for a long time specified these parameters in a single document.
I don't see the advantage in having multiple documents specify these
QSPEC parameters. =20

Thanks,
Regards,
Jerry
=20
> If there is, however, consensus for keeping the QoS class parameters
> in the generic QSPEC as optional, I'd like to see something like a
> "scope" parameter flag like Jukka proposed.
>=20
> > Hopefully we are now ready to revise the QSPEC document=20
> > according to the
> > agreements summarized in the presentation at IETF-67
> >=20
> >
https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=3D6=
7
> > (subject to the revisions of including option 2/encapsulated QSPEC).

> As expressed in the meeting and above, I agree with the
> presentation for the most part of it.
>=20
> Regards,
>  Roland
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 16:02:28 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk5PI-0006ue-ML; Tue, 14 Nov 2006 16:01:24 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gk5PH-0006sE-Gg
	for nsis@ietf.org; Tue, 14 Nov 2006 16:01:23 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gk5PD-0004JX-TZ
	for nsis@ietf.org; Tue, 14 Nov 2006 16:01:23 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Tue, 14 Nov 2006 16:01:11 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006111416011025577 ; Tue, 14 Nov 2006 16:01:10 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <WZT2R22Y>; Tue, 14 Nov 2006 16:01:11 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EF1@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>,
	Roland Bless <bless@tm.uka.de>
Subject: RE: [NSIS] Alternatives for supporting domain-local admission con
	troloperation in NSLP/QSPEC
Date: Tue, 14 Nov 2006 16:01:07 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 14 Nov 2006 21:01:07.0756 (UTC)
	FILETIME=[015B46C0:01C70830]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6cca30437e2d04f45110f2ff8dc1b1d5
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Inline [RRR].

-----Original Message-----
From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On Behalf Of
Ash, Gerald R (Jerry), ALABS
Sent: Tuesday, November 14, 2006 3:01 PM
To: Roland Bless
Cc: Ash, Gerald R (Jerry), ALABS; NSIS Group
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

Hi Roland,
...

I don't see the advantage in having multiple documents specify these
QSPEC parameters.  
...

[RRR] Multiple QOSM documents are already there. Each QOSM has its
unique needs to be satisfied. This generic QSPEC MUST not be
overburdened with because value-based SLAs like Y.1451, DSTE, and
3GPP/Y.1451 are NON-interoperable among themselves. They are also
NON-interoperable with other QOS models (RSVP, IntServ, and DiffServ).

[RRR] Even one document is not enough for Y.1451 because it does not
have its own signaling messages. First, Y.1451 has to create its own
extensions in NSIS QOS signaling messages to transfer its own set of
tightly coupled value-specific SLA QOS parameters. It will then need to
show how Y.1451 parameters will be used on e2e basis for call flows and
QOS negotiations in its own SLA-specific domain (please note that Y.1451
and DiffServ domain are not the same or similar. They are apples and
oranges).

[RRR] Example call flows using SIP application (or even network layer
NSIS signaling messages) need to be shown for Y.1451 for making people
believe that Y.1451 QOS negotiations work on e2e basis. The key question
is this: What kinds of QOS signaling messages one needs in NSIS QOS
signaling messages for Y.1451? Will those signaling messages work for
Y.1451 with any applications like SIP for QOS negotiations? For example,
SIP may use many codec types in a given INVITE message and each codec
will have its own QOS requirements. How will QOS signaling messages for
Y.1541 be used during codec negotiations?

[RRR] The Y.1451 QOS signaling message call flows need to be shown in
the draft to test whether those QOS signaling messages are sufficient.
Unless the draft itself shows how Y.1451 works using the NSIS QOS
signaling messages, it is almost "useless" to PRESS the entire NSIS WG
and IETF as a whole by making a un-substantiated statement that Y.1451
will work if some extensions are used in QSPEC draft for Y.1451. The
"generic" QSPEC draft is not the right place to do so.

[RRR] By the way, it is NOT the case for RSVP, IntServ, and DiffServ. If
so, what is missing in Y.1451?


Thanks,
Regards,
Jerry
 
> If there is, however, consensus for keeping the QoS class parameters
> in the generic QSPEC as optional, I'd like to see something like a
> "scope" parameter flag like Jukka proposed.
> 
> > Hopefully we are now ready to revise the QSPEC document 
> > according to the
> > agreements summarized in the presentation at IETF-67
> > 
> >
https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=67
> > (subject to the revisions of including option 2/encapsulated QSPEC).

> As expressed in the meeting and above, I agree with the
> presentation for the most part of it.
> 
> Regards,
>  Roland
> 


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 14 19:24:44 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk8ZX-0002bm-2e; Tue, 14 Nov 2006 19:24:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gk8ZV-0002XJ-Jp; Tue, 14 Nov 2006 19:24:09 -0500
Received: from stsc1260-eth-s1-s1p1-vip.va.neustar.com ([156.154.16.129]
	helo=chiedprmail1.ietf.org)
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1Gjzgf-0005gs-FP; Tue, 14 Nov 2006 09:54:57 -0500
Received: from smtp.dei.uc.pt ([193.137.203.228] helo=smtp2.dei.uc.pt)
	by chiedprmail1.ietf.org with esmtp (Exim 4.43)
	id 1GjzVq-0004q1-9w; Tue, 14 Nov 2006 09:43:49 -0500
Received: from zoing ([193.136.207.116]) (authenticated bits=0)
	by smtp2.dei.uc.pt (8.13.8/8.13.8) with ESMTP id kAEEgNYc025279
	(version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO);
	Tue, 14 Nov 2006 14:42:23 GMT
Message-Id: <200611141442.kAEEgNYc025279@smtp2.dei.uc.pt>
From: =?iso-8859-1?Q?Lu=EDs_Cordeiro?= <cordeiro@dei.uc.pt>
To: "'Roland Bless'" <bless@tm.uka.de>, "'NSIS Group'" <nsis@ietf.org>,
	<nsis-imp@ietf.org>
Date: Tue, 14 Nov 2006 14:42:21 -0000
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
X-Mailer: Microsoft Office Outlook, Build 11.0.5510
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: AccEbzklsUZK52kISYGjughYWbOgwwDinOyQ
In-Reply-To: <4553E20A.4070304@tm.uka.de>
X-FCTUC-DEI-SIC-MailScanner-Information: Please contact helpdesk@dei.uc.pt for
	more information
X-FCTUC-DEI-SIC-MailScanner: Found to be clean
X-FCTUC-DEI-SIC-MailScanner-SpamCheck: not spam (whitelisted),
	SpamAssassin (score=-1.8, required 3, autolearn=not spam,
	ALL_TRUSTED -1.80)
X-FCTUC-DEI-SIC-MailScanner-From: cordeiro@dei.uc.pt
X-Spam-Status: No
X-Spam-Score: 0.0 (/)
X-Scan-Signature: b19722fc8d3865b147c75ae2495625f2
Cc: 
Subject: [NSIS] RE: [Nsis-imp] Next Interop
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland,

Thanks for the invite.

I talked with the University of Coimbra implementation team and the
middle/end of February would be better.

What do you think?

Lu=EDs Cordeiro

-----Original Message-----
From: Roland Bless [mailto:bless@tm.uka.de]=20
Sent: sexta-feira, 10 de Novembro de 2006 2:21
To: NSIS Group; nsis-imp@ietf.org
Subject: [Nsis-imp] Next Interop

Hi all,

I just wanted to suggest a date for the next NSIS Interop.
Next IETF is from 18th-23rd March 2007. So having it at
the beginning of March would be good (5.3.-9.3.2007).
Maybe the week before is also good.
Other suggestions?

Regards,
 Roland

_______________________________________________
NSIS-imp mailing list
NSIS-imp@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/nsis-imp


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 15 04:53:02 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GkHR8-000350-5o; Wed, 15 Nov 2006 04:52:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GkHR6-00034t-GV
	for nsis@ietf.org; Wed, 15 Nov 2006 04:52:04 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GkHR5-0003xd-1v
	for nsis@ietf.org; Wed, 15 Nov 2006 04:52:04 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GkHQx-00012k-PT; Wed, 15 Nov 2006 10:51:57 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GkHQw-0001qt-4p; Wed, 15 Nov 2006 10:51:54 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 08BDBB58C;
	Wed, 15 Nov 2006 10:51:53 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GkHQu-0008HX-J4; Wed, 15 Nov 2006 10:51:52 +0100
Message-ID: <455AE338.6000509@tm.uka.de>
Date: Wed, 15 Nov 2006 10:51:52 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Subject: Re: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F7C@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F7C@KCCLUST06EVS1.ugd.att.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -2.3 (--)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-1.1 BAYES_05 BODY: Bayesian spam probability is 1 to 5%
	[score: 0.0350]
	0.6 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 02ec665d00de228c50c93ed6b5e4fc1a
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry,

Ash, Gerald R (Jerry), ALABS wrote:
> You have a point regarding signaling PHB class across domains using
> inconsistent classes.  However IMO it should still be possible for a QNI
> to signal a PHB class if called for by the QNI's QOSM, and in which
> end-to-end interoperability/consistency may be achieved.  

As far as I know from old DiffServ times, each domain is free to
use its own DSCPs and PHBs (i.e. it may even remap PHBs on each other).
Furthermore, the "service" usually doesn't only depend on the PHB,
but also on the PDB, i.e. PHBs + traffic conditioning mechanisms.
I guess that it makes quite a difference what traffic conditioning
a domain is using. So signaling a PHB-ID alone isn't very meaningful
between different DiffServ domains (unless they are a DS region under
the same administration/provider) IMHO. So if end-to-end is within
a single domain or consistent DS region this may work, but between
different ASes?

> In addition, the Y.1541 parameter can be regarded as a 'summary'
> constraint parameter encompassing the latency, jitter, PLR, and PER
> constraints (as stated in the summary slides at IETF-67).  These
> parameters are intended to be signaled end-to-end across multiple
> domains.

Hmm. so it's only an abbreviation for the generic QSPEC parameters?
I see it as a profile description how to set the generic QSPEC
parameter values for different QoS classes defined in Y.1541.
But why not simply use the generic QSPEC?
I'm not against having a separate Y1541-QSPEC.
If this is going to get wide support, great!

> Therefore the QNI needs to signal these end-to-end parameters in some
> cases.
> 
> It seems that the proposal is to introduce a QoS NSLP specific 'generic
> QOSM', and everybody must map to this new QOSM.  We have tried

Oh no, please, big misunderstanging here!
I'm NOT talking about a generic QoSM, I'm talking about
a generic _QSPEC_. This is a really significant difference. While the
generic QSPEC may be mapped to any QoSM, mapping between QoSMs is not
going to work. I believe that it is really essential to allow different
QoSMs to be used. However, we need a common understanding of what
service is required between domains and I see the generic QSPEC as
a viable means to express that. In my understanding a QOSM is a means
to provide services, e.g. as proposed in Appendix VI of Y.1541 (using
a mapping from Y.1541 classes to Y.1221/DiffServ PHBs). This must be
completed, however, by specifying resource admission control algorithms
and traffic conditioning mechanisms.

> oftentimes to standardize a 'generic QOSM' and have not succeeded.  OTOH
> we have for a long time specified these parameters in a single document.
> I don't see the advantage in having multiple documents specify these
> QSPEC parameters.

So we must check for consensus to leave Y.1541 in the generic QSPEC.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 15 05:18:26 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GkHpu-0003MJ-86; Wed, 15 Nov 2006 05:17:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GkHpt-0003ME-7C
	for nsis@ietf.org; Wed, 15 Nov 2006 05:17:41 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GkHpq-00087H-Mv
	for nsis@ietf.org; Wed, 15 Nov 2006 05:17:41 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kAFAHUfq008682; Wed, 15 Nov 2006 11:17:34 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Roland Bless'" <bless@tm.uka.de>,
	"'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Date: Wed, 15 Nov 2006 11:17:24 +0100
Message-ID: <001a01c7089f$41e9eba0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
In-Reply-To: <455AE338.6000509@tm.uka.de>
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: AccInBom9fDnkjbtTz+rJpQcCR1aAQAAfoUw
X-Spam-Score: 0.3 () FVGT_s_OBFU_Q1
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Wed, 15 Nov 2006 11:17:35 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 5ebbf074524e58e662bc8209a6235027
Cc: 'NSIS Group' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland

But the QSPEC template draft is not a generic QSPEC draft.
The Token Bucket and the constraints can be seen as something 
that can be common and can be used for the interoperability and the rest of
the encoded 
parameters can just be used as 
guidelines on how the QoS models can encode their parameters.

Thus maybe Jerry could separate the QSEOPC template draft into:
A section/part that describes the Token Bucket + constraints. 
This part could then emphasize that the described parameters can 
be used for interoperability purposes.

Another part of the QSPEC template draft will describe the other parameters.

The goal of this part is to provide guidelines to QoS models on how to
encode the different 
parameters, with the currently available parameters used as examples.

Best Regards,
Georgios



 

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de] 
> Sent: woensdag 15 november 2006 10:52
> To: Ash, Gerald R (Jerry), ALABS
> Cc: NSIS Group
> Subject: Re: [NSIS] Alternatives for supporting domain-local 
> admission controloperation in NSLP/QSPEC
> 
> Hi Jerry,
> 
> Ash, Gerald R (Jerry), ALABS wrote:
> > You have a point regarding signaling PHB class across domains using 
> > inconsistent classes.  However IMO it should still be 
> possible for a 
> > QNI to signal a PHB class if called for by the QNI's QOSM, and in 
> > which end-to-end interoperability/consistency may be achieved.
> 
> As far as I know from old DiffServ times, each domain is free 
> to use its own DSCPs and PHBs (i.e. it may even remap PHBs on 
> each other).
> Furthermore, the "service" usually doesn't only depend on the 
> PHB, but also on the PDB, i.e. PHBs + traffic conditioning mechanisms.
> I guess that it makes quite a difference what traffic 
> conditioning a domain is using. So signaling a PHB-ID alone 
> isn't very meaningful between different DiffServ domains 
> (unless they are a DS region under the same 
> administration/provider) IMHO. So if end-to-end is within a 
> single domain or consistent DS region this may work, but 
> between different ASes?
> 
> > In addition, the Y.1541 parameter can be regarded as a 'summary'
> > constraint parameter encompassing the latency, jitter, PLR, and PER 
> > constraints (as stated in the summary slides at IETF-67).  These 
> > parameters are intended to be signaled end-to-end across multiple 
> > domains.
> 
> Hmm. so it's only an abbreviation for the generic QSPEC parameters?
> I see it as a profile description how to set the generic 
> QSPEC parameter values for different QoS classes defined in Y.1541.
> But why not simply use the generic QSPEC?
> I'm not against having a separate Y1541-QSPEC.
> If this is going to get wide support, great!
> 
> > Therefore the QNI needs to signal these end-to-end 
> parameters in some 
> > cases.
> > 
> > It seems that the proposal is to introduce a QoS NSLP specific 
> > 'generic QOSM', and everybody must map to this new QOSM.  We have 
> > tried
> 
> Oh no, please, big misunderstanging here!
> I'm NOT talking about a generic QoSM, I'm talking about a 
> generic _QSPEC_. This is a really significant difference. 
> While the generic QSPEC may be mapped to any QoSM, mapping 
> between QoSMs is not going to work. I believe that it is 
> really essential to allow different QoSMs to be used. 
> However, we need a common understanding of what service is 
> required between domains and I see the generic QSPEC as a 
> viable means to express that. In my understanding a QOSM is a 
> means to provide services, e.g. as proposed in Appendix VI of 
> Y.1541 (using a mapping from Y.1541 classes to 
> Y.1221/DiffServ PHBs). This must be completed, however, by 
> specifying resource admission control algorithms and traffic 
> conditioning mechanisms.
> 
> > oftentimes to standardize a 'generic QOSM' and have not succeeded.  
> > OTOH we have for a long time specified these parameters in 
> a single document.
> > I don't see the advantage in having multiple documents 
> specify these 
> > QSPEC parameters.
> 
> So we must check for consensus to leave Y.1541 in the generic QSPEC.
> 
> Regards,
>  Roland
> 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 15 07:20:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GkJjd-0000Ie-23; Wed, 15 Nov 2006 07:19:21 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GkJjb-0000Hp-6q; Wed, 15 Nov 2006 07:19:19 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1GkJjX-00016z-T3; Wed, 15 Nov 2006 07:19:19 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GkJjM-0007eL-O8; Wed, 15 Nov 2006 13:19:08 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GkJjL-0001ce-Qn; Wed, 15 Nov 2006 13:19:03 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 770B0B4F7;
	Wed, 15 Nov 2006 13:19:01 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GkJjI-0008Pf-Vo; Wed, 15 Nov 2006 13:19:01 +0100
Message-ID: <455B05B4.2090806@tm.uka.de>
Date: Wed, 15 Nov 2006 13:19:00 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: =?ISO-8859-1?Q?Lu=EDs_Cordeiro?= <cordeiro@dei.uc.pt>
Subject: Re: [NSIS] RE: [Nsis-imp] Next Interop
References: <200611141442.kAEEgNYc025279@smtp2.dei.uc.pt>
In-Reply-To: <200611141442.kAEEgNYc025279@smtp2.dei.uc.pt>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Spam-Score: -3.1 (---)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	1.3 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cf4fa59384e76e63313391b70cd0dd25
Cc: nsis-imp@ietf.org, 'NSIS Group' <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Luís,

Luís Cordeiro wrote:
> Thanks for the invite.
> 
> I talked with the University of Coimbra implementation team and the
> middle/end of February would be better.
> 
> What do you think?

I just saw that the week from 19.2.-23.2.2007 would
be indeed a good fit, since there is a conference
in the week from 26.2. on.

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 15 08:52:23 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GkLAV-0003YY-TD; Wed, 15 Nov 2006 08:51:11 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GkLAU-0003YS-B1
	for nsis@ietf.org; Wed, 15 Nov 2006 08:51:10 -0500
Received: from mclmx2.mail.saic.com ([149.8.64.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GkLAS-0006vh-VX
	for nsis@ietf.org; Wed, 15 Nov 2006 08:51:10 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx2.mail.saic.com; Wed, 15 Nov 2006 08:50:55 -0500
Received: from 0015-ITS-EXBH01.us.saic.com ([10.43.229.18])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006111508505126008 ; Wed, 15 Nov 2006 08:50:53 -0500
Received: from 0591-ITS-EXMP02.us.saic.com ([10.42.81.248]) by
	0015-ITS-EXBH01.us.saic.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Wed, 15 Nov 2006 08:50:49 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Date: Wed, 15 Nov 2006 08:50:48 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EF2@0591-ITS-EXMP02.us.saic.com>
In-Reply-To: <001a01c7089f$41e9eba0$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Thread-Index: AccInBom9fDnkjbtTz+rJpQcCR1aAQAAfoUwAAbihGA=
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: "Georgios Karagiannis" <karagian@cs.utwente.nl>,
	"Roland Bless" <bless@tm.uka.de>,
	"Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
X-OriginalArrivalTime: 15 Nov 2006 13:50:49.0980 (UTC)
	FILETIME=[0F2D0BC0:01C708BD]
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 848ed35f2a4fc0638fa89629cb640f48
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios and Ronald:

Please see inline [RRR].

BR/Radhika

-----Original Message-----
From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On Behalf Of
Georgios Karagiannis
Sent: Wednesday, November 15, 2006 5:17 AM
To: 'Roland Bless'; 'Ash, Gerald R (Jerry), ALABS'
Cc: 'NSIS Group'
Subject: RE: [NSIS] Alternatives for supporting domain-local
admissioncontroloperation in NSLP/QSPEC

Hi Roland

But the QSPEC template draft is not a generic QSPEC draft.
The Token Bucket and the constraints can be seen as something=20
that can be common and can be used for the interoperability and the rest
of
the encoded parameters can just be used as guidelines on how the QoS
models can encode their parameters.

[RRR] It depends. Of course, this QSPEC draft is a "generic" one. If it
is a generic guideline without referring to any particular QOS model, it
is OK. However, to go through for all QOS models individually, it is
out-of-scope because it is a kind of implementation how to use this
QSPEC. Moreover, we have problems to accept Y.1451, DSTE, and
3GPP/Y.1451 as QOS models because they do not qualify as ones, but do
qualify as QOS SLAs.

[RRR] The use of token bucket with required constraints by any
individual QOS model is also QOS model implementation-specific criteria.
It MUST be done right, not saying some rudimentary extensions here or
there in this "generic" draft. There has to be separate drafts for all
QOS models. Moreover, there are a lot of other things need to be
addressed for each QOS model in connection with the encoding.

Thus maybe Jerry could separate the QSEOPC template draft into:
A section/part that describes the Token Bucket + constraints.=20
This part could then emphasize that the described parameters can=20
be used for interoperability purposes.

[RRR] It is permissible as long as it is generic one without referring
to any particular QOS models.

Another part of the QSPEC template draft will describe the other
parameters.
The goal of this part is to provide guidelines to QoS models on how to
encode the different  parameters, with the currently available
parameters used as examples.

[RRR] Again, as explained above, it has to be "generic one" without
referring to any particular QOS model. To mention any particular QOS
model is known as "use case," and any use case is out-of-scope of this
draft. If DiffServ QOS model is taken as the baseline for
interoperability, any mention of this model should also be avoided
unless it is unavoidable.

[RRR] The reason is simple: RSVP, IntServ, and DiffServ are qualified as
QOS models. They also interwork among themselves. Even for these, there
are separate QOSM drafts and more individual drafts will come out in
NSIS along this line. So, let us do it right separately without messing
up this "generic" draft.

[RRR] Y.1451, DSTE, and 3GPP/Y.1451 are QOS SLAs, not QOS models. These
QOS SLAs MUST not be included for any "standard" track RFCs. Period.
They are not interoperable either. They need a lot of works just to
prove how they will work using NSIS QOS signaling protocols and need
extensions here and there as they do not have any signaling messages of
their own. Is this clear?

Best Regards,
Georgios



=20

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de]=20
> Sent: woensdag 15 november 2006 10:52
> To: Ash, Gerald R (Jerry), ALABS
> Cc: NSIS Group
> Subject: Re: [NSIS] Alternatives for supporting domain-local=20
> admission controloperation in NSLP/QSPEC
>=20
> Hi Jerry,
>=20
> Ash, Gerald R (Jerry), ALABS wrote:
> > You have a point regarding signaling PHB class across domains using=20
> > inconsistent classes.  However IMO it should still be=20
> possible for a=20
> > QNI to signal a PHB class if called for by the QNI's QOSM, and in=20
> > which end-to-end interoperability/consistency may be achieved.
>=20
> As far as I know from old DiffServ times, each domain is free=20
> to use its own DSCPs and PHBs (i.e. it may even remap PHBs on=20
> each other).
> Furthermore, the "service" usually doesn't only depend on the=20
> PHB, but also on the PDB, i.e. PHBs + traffic conditioning mechanisms.
> I guess that it makes quite a difference what traffic=20
> conditioning a domain is using. So signaling a PHB-ID alone=20
> isn't very meaningful between different DiffServ domains=20
> (unless they are a DS region under the same=20
> administration/provider) IMHO. So if end-to-end is within a=20
> single domain or consistent DS region this may work, but=20
> between different ASes?
>=20
> > In addition, the Y.1541 parameter can be regarded as a 'summary'
> > constraint parameter encompassing the latency, jitter, PLR, and PER=20
> > constraints (as stated in the summary slides at IETF-67).  These=20
> > parameters are intended to be signaled end-to-end across multiple=20
> > domains.
>=20
> Hmm. so it's only an abbreviation for the generic QSPEC parameters?
> I see it as a profile description how to set the generic=20
> QSPEC parameter values for different QoS classes defined in Y.1541.
> But why not simply use the generic QSPEC?
> I'm not against having a separate Y1541-QSPEC.
> If this is going to get wide support, great!
>=20
> > Therefore the QNI needs to signal these end-to-end=20
> parameters in some=20
> > cases.
> >=20
> > It seems that the proposal is to introduce a QoS NSLP specific=20
> > 'generic QOSM', and everybody must map to this new QOSM.  We have=20
> > tried
>=20
> Oh no, please, big misunderstanging here!
> I'm NOT talking about a generic QoSM, I'm talking about a=20
> generic _QSPEC_. This is a really significant difference.=20
> While the generic QSPEC may be mapped to any QoSM, mapping=20
> between QoSMs is not going to work. I believe that it is=20
> really essential to allow different QoSMs to be used.=20
> However, we need a common understanding of what service is=20
> required between domains and I see the generic QSPEC as a=20
> viable means to express that. In my understanding a QOSM is a=20
> means to provide services, e.g. as proposed in Appendix VI of=20
> Y.1541 (using a mapping from Y.1541 classes to=20
> Y.1221/DiffServ PHBs). This must be completed, however, by=20
> specifying resource admission control algorithms and traffic=20
> conditioning mechanisms.
>=20
> > oftentimes to standardize a 'generic QOSM' and have not succeeded. =20
> > OTOH we have for a long time specified these parameters in=20
> a single document.
> > I don't see the advantage in having multiple documents=20
> specify these=20
> > QSPEC parameters.
>=20
> So we must check for consensus to leave Y.1541 in the generic QSPEC.
>=20
> Regards,
>  Roland
>=20
>=20
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
>=20



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Wed Nov 15 15:52:07 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GkRiz-0005Gs-An; Wed, 15 Nov 2006 15:51:13 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GkRio-00052P-Pb; Wed, 15 Nov 2006 15:51:02 -0500
Received: from stsc1260-eth-s1-s1p1-vip.va.neustar.com ([156.154.16.129]
	helo=chiedprmail1.ietf.org)
	by ietf-mx.ietf.org with esmtp (Exim 4.43)
	id 1GkRio-0002AT-Nz; Wed, 15 Nov 2006 15:51:02 -0500
Received: from ns4.neustar.com ([156.154.24.139])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43)
	id 1GkRio-0006LM-DU; Wed, 15 Nov 2006 15:51:02 -0500
Received: from stiedprstage1.ietf.org (stiedprstage1.va.neustar.com
	[10.31.47.10]) by ns4.neustar.com (Postfix) with ESMTP id 6F8972AC67;
	Wed, 15 Nov 2006 20:50:02 +0000 (GMT)
Received: from ietf by stiedprstage1.ietf.org with local (Exim 4.43)
	id 1GkRhq-0008GS-5V; Wed, 15 Nov 2006 15:50:02 -0500
Content-Type: Multipart/Mixed; Boundary="NextPart"
Mime-Version: 1.0
To: i-d-announce@ietf.org
From: Internet-Drafts@ietf.org
Message-Id: <E1GkRhq-0008GS-5V@stiedprstage1.ietf.org>
Date: Wed, 15 Nov 2006 15:50:02 -0500
X-Spam-Score: -2.5 (--)
X-Scan-Signature: b280b4db656c3ca28dd62e5e0b03daa8
Cc: nsis@ietf.org
Subject: [NSIS] I-D ACTION:draft-ietf-nsis-y1541-qosm-03.txt 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

--NextPart

A New Internet-Draft is available from the on-line Internet-Drafts 
directories.
This draft is a work item of the Next Steps in Signaling Working Group of the IETF.

	Title		: Y.1541-QOSM -- Y.1541 QoS Model for Networks Using Y.1541 QoS Classes
	Author(s)	: J. Ash, et al.
	Filename	: draft-ietf-nsis-y1541-qosm-03.txt
	Pages		: 12
	Date		: 2006-11-15
	
This draft describes a QoS-NSLP QoS model (QOSM) based on ITU-T
   Recommendation Y.1541 QoS signaling requirements.  Y.1541 specifies 6
   standard QoS classes, and the Y.1541-QOSM extensions include

A URL for this Internet-Draft is:
http://www.ietf.org/internet-drafts/draft-ietf-nsis-y1541-qosm-03.txt

To remove yourself from the I-D Announcement list, send a message to 
i-d-announce-request@ietf.org with the word unsubscribe in the body of 
the message. 
You can also visit https://www1.ietf.org/mailman/listinfo/I-D-announce 
to change your subscription settings.

Internet-Drafts are also available by anonymous FTP. Login with the 
username "anonymous" and a password of your e-mail address. After 
logging in, type "cd internet-drafts" and then 
"get draft-ietf-nsis-y1541-qosm-03.txt".

A list of Internet-Drafts directories can be found in
http://www.ietf.org/shadow.html 
or ftp://ftp.ietf.org/ietf/1shadow-sites.txt

Internet-Drafts can also be obtained by e-mail.

Send a message to:
	mailserv@ietf.org.
In the body type:
	"FILE /internet-drafts/draft-ietf-nsis-y1541-qosm-03.txt".
	
NOTE:	The mail server at ietf.org can return the document in
	MIME-encoded form by using the "mpack" utility.  To use this
	feature, insert the command "ENCODING mime" before the "FILE"
	command.  To decode the response(s), you will need "munpack" or
	a MIME-compliant mail reader.  Different MIME-compliant mail readers
	exhibit different behavior, especially when dealing with
	"multipart" MIME messages (i.e. documents which have been split
	up into multiple messages), so check your local documentation on
	how to manipulate these messages.

Below is the data which will enable a MIME compliant mail reader
implementation to automatically retrieve the ASCII version of the
Internet-Draft.

--NextPart
Content-Type: Multipart/Alternative; Boundary="OtherAccess"

--OtherAccess
Content-Type: Message/External-body; access-type="mail-server";
	server="mailserv@ietf.org"

Content-Type: text/plain
Content-ID: <2006-11-15133331.I-D@ietf.org>

ENCODING mime
FILE /internet-drafts/draft-ietf-nsis-y1541-qosm-03.txt

--OtherAccess
Content-Type: Message/External-body; name="draft-ietf-nsis-y1541-qosm-03.txt";
	site="ftp.ietf.org"; access-type="anon-ftp";
	directory="internet-drafts"

Content-Type: text/plain
Content-ID: <2006-11-15133331.I-D@ietf.org>


--OtherAccess--

--NextPart
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--NextPart--




From nsis-bounces@ietf.org Thu Nov 16 10:12:39 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gkiu8-0004nd-6R; Thu, 16 Nov 2006 10:11:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gkiu7-0004nT-M3
	for nsis@ietf.org; Thu, 16 Nov 2006 10:11:51 -0500
Received: from stsc1260-eth-s1-s1p1-vip.va.neustar.com ([156.154.16.129]
	helo=chiedprmail1.ietf.org)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GkLdJ-0001xc-DW
	for nsis@ietf.org; Wed, 15 Nov 2006 09:20:57 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by chiedprmail1.ietf.org with esmtp (Exim 4.43) id 1GkLPv-0003wg-I4
	for nsis@ietf.org; Wed, 15 Nov 2006 09:07:08 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Wed, 15 Nov 2006 09:06:43 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006111509064307305 ; Wed, 15 Nov 2006 09:06:43 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <WZT2S778>; Wed, 15 Nov 2006 09:06:43 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EF3@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: Roland Bless <bless@tm.uka.de>,
	"Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Subject: RE: [NSIS] Alternatives for supporting domain-local admission con
	troloperation in NSLP/QSPEC
Date: Wed, 15 Nov 2006 09:06:35 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 15 Nov 2006 14:06:36.0356 (UTC)
	FILETIME=[43429040:01C708BF]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 6e922792024732fb1bb6f346e63517e4
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Roland:

I fully agree with you as we are implementing DiffServ domains in the
worldwide network as you have mentioned. The implementation of each
DiffServ domain may differ although they use the same or similar DSCPs
and PHBs.

In addition, the allocation of bandwidth in each DiffServ domain is also
a matter of implementations. Some use static means with over
provisioning while others use dynamic bandwidth allocation using some
proprietary schemes on the top of the DiffServ domain.

You are also right in pointing out that there is also specific QSOM
draft for each QOS model. The QOSM is NOT a generic one.

You are right in clarifying the things that each QOS model will use the
"generic QSPEC" draft for "encoding," "Extensions in NSIS QOS signaling
messages if any," "specific guidelines how use constraints in
conjunction with Token Bucket," and many other specific things.

Why do people want to "mess" the same "generic QSPEC" standard track RFC
using some implementation-specific guidelines that are sure to be
"incomplete" for standard QOS models and some non-standard QOS SLAs?

Best regards,
Radhika

-----Original Message-----
From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org] On Behalf Of
Roland Bless
Sent: Wednesday, November 15, 2006 4:52 AM
To: Ash, Gerald R (Jerry), ALABS
Cc: NSIS Group
Subject: Re: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

Hi Jerry,

Ash, Gerald R (Jerry), ALABS wrote:
> You have a point regarding signaling PHB class across domains using
> inconsistent classes.  However IMO it should still be possible for a
QNI
> to signal a PHB class if called for by the QNI's QOSM, and in which
> end-to-end interoperability/consistency may be achieved.  

As far as I know from old DiffServ times, each domain is free to
use its own DSCPs and PHBs (i.e. it may even remap PHBs on each other).
Furthermore, the "service" usually doesn't only depend on the PHB,
but also on the PDB, i.e. PHBs + traffic conditioning mechanisms.
I guess that it makes quite a difference what traffic conditioning
a domain is using. So signaling a PHB-ID alone isn't very meaningful
between different DiffServ domains (unless they are a DS region under
the same administration/provider) IMHO. So if end-to-end is within
a single domain or consistent DS region this may work, but between
different ASes?

> In addition, the Y.1541 parameter can be regarded as a 'summary'
> constraint parameter encompassing the latency, jitter, PLR, and PER
> constraints (as stated in the summary slides at IETF-67).  These
> parameters are intended to be signaled end-to-end across multiple
> domains.

Hmm. so it's only an abbreviation for the generic QSPEC parameters?
I see it as a profile description how to set the generic QSPEC
parameter values for different QoS classes defined in Y.1541.
But why not simply use the generic QSPEC?
I'm not against having a separate Y1541-QSPEC.
If this is going to get wide support, great!

> Therefore the QNI needs to signal these end-to-end parameters in some
> cases.
> 
> It seems that the proposal is to introduce a QoS NSLP specific
'generic
> QOSM', and everybody must map to this new QOSM.  We have tried

Oh no, please, big misunderstanging here!
I'm NOT talking about a generic QoSM, I'm talking about
a generic _QSPEC_. This is a really significant difference. While the
generic QSPEC may be mapped to any QoSM, mapping between QoSMs is not
going to work. I believe that it is really essential to allow different
QoSMs to be used. However, we need a common understanding of what
service is required between domains and I see the generic QSPEC as
a viable means to express that. In my understanding a QOSM is a means
to provide services, e.g. as proposed in Appendix VI of Y.1541 (using
a mapping from Y.1541 classes to Y.1221/DiffServ PHBs). This must be
completed, however, by specifying resource admission control algorithms
and traffic conditioning mechanisms.

> oftentimes to standardize a 'generic QOSM' and have not succeeded.
OTOH
> we have for a long time specified these parameters in a single
document.
> I don't see the advantage in having multiple documents specify these
> QSPEC parameters.

So we must check for consensus to leave Y.1541 in the generic QSPEC.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 17 12:33:11 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gl7Z3-00066M-GM; Fri, 17 Nov 2006 12:31:45 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gl7Z2-00066H-KL
	for nsis@ietf.org; Fri, 17 Nov 2006 12:31:44 -0500
Received: from mail.gmx.net ([213.165.64.20])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gl7YC-0008IV-8v
	for nsis@ietf.org; Fri, 17 Nov 2006 12:31:43 -0500
Received: (qmail invoked by alias); 17 Nov 2006 17:30:51 -0000
Received: from socks1.netz.sbs.de (EHLO [192.35.17.26]) [192.35.17.26]
	by mail.gmx.net (mp029) with SMTP; 17 Nov 2006 18:30:51 +0100
X-Authenticated: #29516787
Message-ID: <455DF1CE.5040004@gmx.net>
Date: Fri, 17 Nov 2006 18:30:54 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: nsis <nsis@ietf.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 08e48e05374109708c00c6208b534009
Subject: [NSIS] Meeting Minutes
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all,

here is a first proposal for the meeting minutes.

http://www3.ietf.org/proceedings/06nov/minutes/nsis.txt

Ciao
Hannes

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 17 13:02:51 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gl82B-0004jy-Mb; Fri, 17 Nov 2006 13:01:51 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gl82A-0004jc-Hl
	for nsis@ietf.org; Fri, 17 Nov 2006 13:01:50 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gl81x-0003lW-4O
	for nsis@ietf.org; Fri, 17 Nov 2006 13:01:50 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Fri, 17 Nov 2006 13:01:27 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006111713012727079 ; Fri, 17 Nov 2006 13:01:27 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <WZT2XVVV>; Fri, 17 Nov 2006 13:01:27 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF2F@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: Roland Bless <bless@tm.uka.de>,
	"Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Subject: RE: [NSIS] Alternatives for supporting domain-local admission con
	troloperation in NSLP/QSPEC
Date: Fri, 17 Nov 2006 13:01:21 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 17 Nov 2006 18:01:21.0705 (UTC)
	FILETIME=[639BC190:01C70A72]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 21be852dc93f0971708678c18d38c096
Cc: NSIS Group <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Roland:
 
A side conclusion of your email is as follows:

1.	

	DiffServ's DSCPs and PHBs provide a overall "generic" frame work how
a DiffServ QOS domain needs to work.
2.	

	The actual implementations of a given set of DSCPs and PHBs may vary
how QOS services are defined because QOS services are defined in the end
systems, and no one standardize any " end users' services" per se. For
example, a group of ISPs may be the end-users of a DiffServ domain that may
be agreed among them using a SLA. They may implement some services in this
DiffServ domain SLA that include real-time services (e.g., conversational
audio/video),  near-real-time services (e.g., streaming audio/video), and
non-real-time data services (e.g., interactive data, file transfer). They
may choose a set of DSCP and PHB for each service. Please note that these
ISPs still have not defined any specific values of any QOS performance
parameters. It just tells what kind of DiffServ services (DSCP Type, PHB
Type) will be used, Accordingly queuing disciplines, priority, dropping, or
other services in DiffServ nodes (i.e., routers) will be used for real-time,
near-real-time, and non-real-time application services
flows/packets/messages. This may be enough for these ISPs to satisfy their
DiffServ-based SLA needs.
3.	

	Another group of ISPs may use the same as described in item 2 for
their DiffServ-based SLA, but they may go further. For example, they may
also define specific values of delay, jitter, and packet loss for each
service such as  real-time (e.g., Y.1451 QOS Class I), near-real-time (e.g.,
Y.1451 QOS Class II), and non-real-time (e.g., Y.1451 QOS Class III)
application service. This may be the most "inflexible" SLA implementation.
4.	

	There can be numerous examples how different SLAs can be implemented
based on the "generic" DSCPs and PHBs standards.

In fact, the "generic" DHCP and PHB-based DiffServ standard is one of the
"most successful" QOS standards that have even been created in the history
of QOS. It shows the "beauty" of human creativity for its "enormous
flexibility," "usefulness" and "scalability" to address the QOS problems for
a large-scale vast complex IP network.

Let us also learn the above lesson in creation of the "generic" QSPEC
standard-track RFC removing any value-based implementation-specific QOS SLAs
like Y.1451, DSTE, and 3GPP/Y1451.

Best regards,

Radhika


  _____  

From: nsis-bounces@ietf.org on behalf of Roy, Radhika R.
Sent: Wed 11/15/2006 9:06 AM
To: Roland Bless; Ash, Gerald R (Jerry), ALABS
Cc: NSIS Group
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC



Roland:

I fully agree with you as we are implementing DiffServ domains in the
worldwide network as you have mentioned. The implementation of each
DiffServ domain may differ although they use the same or similar DSCPs
and PHBs.

In addition, the allocation of bandwidth in each DiffServ domain is also
a matter of implementations. Some use static means with over
provisioning while others use dynamic bandwidth allocation using some
proprietary schemes on the top of the DiffServ domain.

You are also right in pointing out that there is also specific QSOM
draft for each QOS model. The QOSM is NOT a generic one.

You are right in clarifying the things that each QOS model will use the
"generic QSPEC" draft for "encoding," "Extensions in NSIS QOS signaling
messages if any," "specific guidelines how use constraints in
conjunction with Token Bucket," and many other specific things.

Why do people want to "mess" the same "generic QSPEC" standard track RFC
using some implementation-specific guidelines that are sure to be
"incomplete" for standard QOS models and some non-standard QOS SLAs?

Best regards,
Radhika

-----Original Message-----
From: nsis-bounces@ietf.org [mailto:nsis-bounces@ietf.org
<mailto:nsis-bounces@ietf.org> ] On Behalf Of
Roland Bless
Sent: Wednesday, November 15, 2006 4:52 AM
To: Ash, Gerald R (Jerry), ALABS
Cc: NSIS Group
Subject: Re: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

Hi Jerry,

Ash, Gerald R (Jerry), ALABS wrote:
> You have a point regarding signaling PHB class across domains using
> inconsistent classes.  However IMO it should still be possible for a
QNI
> to signal a PHB class if called for by the QNI's QOSM, and in which
> end-to-end interoperability/consistency may be achieved. 

As far as I know from old DiffServ times, each domain is free to
use its own DSCPs and PHBs (i.e. it may even remap PHBs on each other).
Furthermore, the "service" usually doesn't only depend on the PHB,
but also on the PDB, i.e. PHBs + traffic conditioning mechanisms.
I guess that it makes quite a difference what traffic conditioning
a domain is using. So signaling a PHB-ID alone isn't very meaningful
between different DiffServ domains (unless they are a DS region under
the same administration/provider) IMHO. So if end-to-end is within
a single domain or consistent DS region this may work, but between
different ASes?

> In addition, the Y.1541 parameter can be regarded as a 'summary'
> constraint parameter encompassing the latency, jitter, PLR, and PER
> constraints (as stated in the summary slides at IETF-67).  These
> parameters are intended to be signaled end-to-end across multiple
> domains.

Hmm. so it's only an abbreviation for the generic QSPEC parameters?
I see it as a profile description how to set the generic QSPEC
parameter values for different QoS classes defined in Y.1541.
But why not simply use the generic QSPEC?
I'm not against having a separate Y1541-QSPEC.
If this is going to get wide support, great!

> Therefore the QNI needs to signal these end-to-end parameters in some
> cases.
>
> It seems that the proposal is to introduce a QoS NSLP specific
'generic
> QOSM', and everybody must map to this new QOSM.  We have tried

Oh no, please, big misunderstanging here!
I'm NOT talking about a generic QoSM, I'm talking about
a generic _QSPEC_. This is a really significant difference. While the
generic QSPEC may be mapped to any QoSM, mapping between QoSMs is not
going to work. I believe that it is really essential to allow different
QoSMs to be used. However, we need a common understanding of what
service is required between domains and I see the generic QSPEC as
a viable means to express that. In my understanding a QOSM is a means
to provide services, e.g. as proposed in Appendix VI of Y.1541 (using
a mapping from Y.1541 classes to Y.1221/DiffServ PHBs). This must be
completed, however, by specifying resource admission control algorithms
and traffic conditioning mechanisms.

> oftentimes to standardize a 'generic QOSM' and have not succeeded.
OTOH
> we have for a long time specified these parameters in a single
document.
> I don't see the advantage in having multiple documents specify these
> QSPEC parameters.

So we must check for consensus to leave Y.1541 in the generic QSPEC.

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 20 12:07:34 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GmCbb-0007tC-Br; Mon, 20 Nov 2006 12:06:51 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GmCba-0007sw-2U
	for nsis@ietf.org; Mon, 20 Nov 2006 12:06:50 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GmCbU-0007Rs-Oq
	for nsis@ietf.org; Mon, 20 Nov 2006 12:06:50 -0500
X-VirusChecked: Checked
X-Env-Sender: acmorton@att.com
X-Msg-Ref: server-11.tower-146.messagelabs.com!1164042403!5748532!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 4268 invoked from network); 20 Nov 2006 17:06:43 -0000
Received: from unknown (HELO attrh0i.attrh.att.com) (134.24.146.4)
	by server-11.tower-146.messagelabs.com with SMTP;
	20 Nov 2006 17:06:43 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh0i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAKH6hxn000935
	for <nsis@ietf.org>; Mon, 20 Nov 2006 12:06:43 -0500 (EST)
Received: from maillennium.att.com (dns.maillennium.att.com [135.25.114.99])
	by attrh0i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAKH6bO3000876
	for <nsis@ietf.org>; Mon, 20 Nov 2006 12:06:37 -0500 (EST)
Received: from acmt.att.com (acmt.mt.att.com[135.16.251.40](misconfigured
	sender)) by maillennium.att.com (mailgw1) with SMTP
	id <20061120170637gw100r41fie>; Mon, 20 Nov 2006 17:06:37 +0000
Message-Id: <7.0.1.0.0.20061120115123.01300370@att.com>
X-Mailer: QUALCOMM Windows Eudora Version 7.0.1.0
Date: Mon, 20 Nov 2006 12:06:36 -0500
To: nsis@ietf.org
From: Al Morton <acmorton@att.com>
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
X-Spam-Score: 0.0 (/)
X-Scan-Signature: cf4fa59384e76e63313391b70cd0dd25
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

It seems to me there is benefit in keeping the Y.1541 class
in the QSPEC, because the Available class number could be modified along
the reservation path. Even if the individual parameters change
in the "QoS Available", it would be important recognize that a
(for example, delay) threshold has been crossed and that a lesser class
is now all that is available. This seems consistent with:

>    QoS Available: QSPEC object containing parameters describing the
>    available resources.  They are used to collect information along a
>    reservation path.

(I had wanted to make this comment during the meeting, but
was busy taking notes for my friend, Hannes)

Al


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 20 13:21:42 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GmDlb-00056B-Ma; Mon, 20 Nov 2006 13:21:15 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GmDlb-00054p-1N
	for nsis@ietf.org; Mon, 20 Nov 2006 13:21:15 -0500
Received: from mclmx2.mail.saic.com ([149.8.64.32])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GmDlY-0006ta-MI
	for nsis@ietf.org; Mon, 20 Nov 2006 13:21:15 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx2.mail.saic.com for nsis@ietf.org;
	Mon, 20 Nov 2006 13:20:57 -0500
Received: from 0015-ITS-EXBH01.us.saic.com ([10.43.229.18])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006112013205718345
	for <nsis@ietf.org>; Mon, 20 Nov 2006 13:20:57 -0500
Received: from 0591-ITS-EXMP02.us.saic.com ([10.42.81.248]) by
	0015-ITS-EXBH01.us.saic.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 20 Nov 2006 13:20:56 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Date: Mon, 20 Nov 2006 13:20:56 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF36@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Thread-Index: AccMxrul+Uow4nigSH+kKKFPEmj9CAACEJ4O
References: <7.0.1.0.0.20061120115123.01300370@att.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: <nsis@ietf.org>
X-OriginalArrivalTime: 20 Nov 2006 18:20:56.0479 (UTC)
	FILETIME=[9F114AF0:01C70CD0]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 057ebe9b96adec30a7efb2aeda4c26a4
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1288393642=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1288393642==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C70CD0.9ED51DAC"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C70CD0.9ED51DAC
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

All:
=20
Again, it is the same problems to do any extensions that relates to =
value-based implementation-specific "non-standard" QOS SLAs like Y.1541, =
DSTE, and 3GPP/Y1541 in this "generic" QSPEC RFC.
=20
However, if any parameter in "QOS Available" template is "generic" =
enough and does NOT mention anything-specific to Y.1541, DSTE, and =
3GPP/Y1541 and is useful for the IETF standard QOS models like RSVP, =
DiffServ, and IntServ, it may then be considered.
=20
All QOSM draft are there. These drafts and any implemetation-sepcific =
drafts for each QOS model that show some examples or even extensions of =
the "generic" QSPEC RFC can be used need to be produced separately.
=20
Best regards,
Radhika

________________________________

From: nsis-bounces@ietf.org on behalf of Al Morton
Sent: Mon 11/20/2006 12:06 PM
To: nsis@ietf.org
Subject: RE: [NSIS] Alternatives for supporting domain-local =
admissioncontroloperation in NSLP/QSPEC



It seems to me there is benefit in keeping the Y.1541 class
in the QSPEC, because the Available class number could be modified along
the reservation path. Even if the individual parameters change
in the "QoS Available", it would be important recognize that a
(for example, delay) threshold has been crossed and that a lesser class
is now all that is available. This seems consistent with:

>    QoS Available: QSPEC object containing parameters describing the
>    available resources.  They are used to collect information along a
>    reservation path.

(I had wanted to make this comment during the meeting, but
was busy taking notes for my friend, Hannes)

Al



------_=_NextPart_001_01C70CD0.9ED51DAC
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<HTML dir=3Dltr><HEAD><TITLE>RE: [NSIS] Alternatives for supporting =
domain-local admissioncontroloperation in NSLP/QSPEC</TITLE>=0A=
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dunicode">=0A=
<META content=3D"MSHTML 6.00.2900.2995" name=3DGENERATOR></HEAD>=0A=
<BODY>=0A=
<DIV id=3DidOWAReplyText52317 dir=3Dltr>=0A=
<DIV dir=3Dltr><FONT face=3DArial color=3D#000000 =
size=3D2>All:</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr>Again, it is the same problems to do any =
extensions&nbsp;that relates to value-based implementation-specific =
"non-standard" QOS SLAs like Y.1541, DSTE, and 3GPP/Y1541&nbsp;in this =
"generic" QSPEC RFC.</DIV>=0A=
<DIV dir=3Dltr>&nbsp;</DIV>=0A=
<DIV dir=3Dltr>However, if any parameter in "QOS Available" template is =
"generic" enough and does&nbsp;NOT mention anything-specific to Y.1541, =
DSTE, and 3GPP/Y1541&nbsp;and is useful for the IETF standard QOS models =
like RSVP, DiffServ, and IntServ, it may then be considered.</DIV>=0A=
<DIV dir=3Dltr>&nbsp;</DIV>=0A=
<DIV dir=3Dltr>All QOSM draft are there. These drafts and any =
implemetation-sepcific drafts for each QOS model that show some examples =
or even extensions of the "generic" QSPEC RFC can be used need to be =
produced separately.</DIV>=0A=
<DIV dir=3Dltr>&nbsp;</DIV>=0A=
<DIV dir=3Dltr>Best regards,</DIV>=0A=
<DIV dir=3Dltr>Radhika</DIV></DIV>=0A=
<DIV dir=3Dltr><BR>=0A=
<HR tabIndex=3D-1>=0A=
<FONT face=3DTahoma size=3D2><B>From:</B> nsis-bounces@ietf.org on =
behalf of Al Morton<BR><B>Sent:</B> Mon 11/20/2006 12:06 =
PM<BR><B>To:</B> nsis@ietf.org<BR><B>Subject:</B> RE: [NSIS] =
Alternatives for supporting domain-local admissioncontroloperation in =
NSLP/QSPEC<BR></FONT><BR></DIV>=0A=
<DIV>=0A=
<P><FONT size=3D2>It seems to me there is benefit in keeping the Y.1541 =
class<BR>in the QSPEC, because the Available class number could be =
modified along<BR>the reservation path. Even if the individual =
parameters change<BR>in the "QoS Available", it would be important =
recognize that a<BR>(for example, delay) threshold has been crossed and =
that a lesser class<BR>is now all that is available. This seems =
consistent with:<BR><BR>&gt;&nbsp;&nbsp;&nbsp; QoS Available: QSPEC =
object containing parameters describing the<BR>&gt;&nbsp;&nbsp;&nbsp; =
available resources.&nbsp; They are used to collect information along =
a<BR>&gt;&nbsp;&nbsp;&nbsp; reservation path.<BR><BR>(I had wanted to =
make this comment during the meeting, but<BR>was busy taking notes for =
my friend, Hannes)<BR><BR>Al<BR></FONT></P></DIV></BODY></HTML>
------_=_NextPart_001_01C70CD0.9ED51DAC--


--===============1288393642==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1288393642==--




From nsis-bounces@ietf.org Mon Nov 20 17:06:39 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GmHH2-0007hZ-59; Mon, 20 Nov 2006 17:05:56 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GmHH0-0007cu-Vg
	for nsis@ietf.org; Mon, 20 Nov 2006 17:05:54 -0500
Received: from mail.gmx.de ([213.165.64.20] helo=mail.gmx.net)
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GmHGz-0002g9-Ns
	for nsis@ietf.org; Mon, 20 Nov 2006 17:05:54 -0500
Received: (qmail invoked by alias); 20 Nov 2006 22:05:52 -0000
Received: from p54987D38.dip.t-dialin.net (EHLO [192.168.2.32]) [84.152.125.56]
	by mail.gmx.net (mp028) with SMTP; 20 Nov 2006 23:05:52 +0100
X-Authenticated: #29516787
Message-ID: <456226C4.7030005@gmx.net>
Date: Mon, 20 Nov 2006 23:05:56 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: NSIS <nsis@ietf.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 7bac9cb154eb5790ae3b2913587a40de
Subject: [NSIS] QSPEC Next Draft Version?
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi,

in the DIME working group we are working on a Diameter QoS document and 
some of us believe that the QSPEC draft would provide good input with 
respect to the QoS parameter description.

When can we expect an updated draft to have some details to look at?

Ciao
Hannes

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 21 05:40:08 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GmT1u-0000Yr-GM; Tue, 21 Nov 2006 05:39:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GmT1t-0000Yh-8T
	for nsis@ietf.org; Tue, 21 Nov 2006 05:39:05 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GmT1o-0008Sa-P5
	for nsis@ietf.org; Tue, 21 Nov 2006 05:39:05 -0500
Received: from sbz-31.cs.helsinki.fi (sbz-31.cs.helsinki.fi [128.214.9.99])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Tue, 21 Nov 2006 12:38:55 +0200
	id 000B3AF5.4562D73F.00002F66
Date: Tue, 21 Nov 2006 12:38:55 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: nsis@ietf.org
Message-ID: <Pine.LNX.4.64.0611211228220.5808@sbz-31.cs.Helsinki.FI>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 0bc60ec82efc80c84b8d02f4b0e4de22
Subject: [NSIS] QoS NSLP v. "13-pre1"
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi all,

I have done all the edits I had in my list to the QoS NSLP spec, the main 
points are in recent thread with Roland. From the url below you can find 
the current draft v. "13-pre1" and a diff to v. 12.

http://www.cs.helsinki.fi/u/jmanner/qos-nslp/

A direct link to the diff:

http://www.cs.helsinki.fi/u/jmanner/qos-nslp/draft-ietf-nsis-qos-nslp-12-from-3-pre1.diff.html


I would appreciate comments on the changes I did (what is wrong, what I 
missed, etc.).


I still have one problem with the spec, namely the way the NSLP suggests 
to handle mobility and path changes. I think we need to review that 
statement, because it seems like a "hack" (playing with the RSN value), 
and not really useful. And the interactions with GIST look partly wrong. 
The folks working on the mobility draft (me included) could comment. The 
section I'm talking about is 3.2.10, especially 3.2.10.2, in the v.12 
draft.

BR,
Jukka

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 21 07:43:00 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GmUws-0004s0-Ru; Tue, 21 Nov 2006 07:42:02 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GmUwq-0004rJ-Vx
	for nsis@ietf.org; Tue, 21 Nov 2006 07:42:00 -0500
Received: from mail120.messagelabs.com ([216.82.250.83])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GmUwp-0000Yl-Jo
	for nsis@ietf.org; Tue, 21 Nov 2006 07:42:00 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-15.tower-120.messagelabs.com!1164112916!14320302!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 17550 invoked from network); 21 Nov 2006 12:41:56 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-15.tower-120.messagelabs.com with SMTP;
	21 Nov 2006 12:41:56 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kALCftNe011048; 
	Tue, 21 Nov 2006 07:41:55 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kALCfqXx011040; Tue, 21 Nov 2006 07:41:53 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] QSPEC Next Draft Version?
Date: Tue, 21 Nov 2006 06:41:52 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7F93@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <456226C4.7030005@gmx.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] QSPEC Next Draft Version?
Thread-Index: AccM8E5+E56ZjRhCS2iThNl2DB1e7wAduAAg
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: "Hannes Tschofenig" <Hannes.Tschofenig@gmx.net>, "NSIS" <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b431ad66d60be2d47c7bfeb879db82c
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Hannes,

The update is in progress, and is being done according to the agreements
we reached at IETF-67 and subsequently on the list. =20

One unresolved issue remains on the suggestion from Roland to move QoS
class parameters to QOSM-specific specifications, RE the thread starting
at http://www1.ietf.org/mail-archive/web/nsis/current/msg07158.html.=20
Discussion followed at
http://www1.ietf.org/mail-archive/web/nsis/current/msg07194.html
http://www1.ietf.org/mail-archive/web/nsis/current/msg07195.html
http://www1.ietf.org/mail-archive/web/nsis/current/msg07206.html
http://www1.ietf.org/mail-archive/web/nsis/current/msg07207.html
http://www1.ietf.org/mail-archive/web/nsis/current/msg07209.html
http://www1.ietf.org/mail-archive/web/nsis/current/msg07212.html
http://www1.ietf.org/mail-archive/web/nsis/current/msg07213.html
http://www1.ietf.org/mail-archive/web/nsis/current/msg07220.html

There are arguments both ways, and for me the last post from Al is an
important point that I agree with.  Barring a further 'consensus' call
on this issue from the chairs, I'll proceed for now with QoS class
parameters remaining in the draft.  This can always be changed later of
course if we agree to do so.

I hope to publish the next version by next week.

Thanks,
Regards,
Jerry=20

-----Original Message-----
From: Hannes Tschofenig [mailto:Hannes.Tschofenig@gmx.net]=20
Sent: Monday, November 20, 2006 5:06 PM
To: NSIS
Subject: [NSIS] QSPEC Next Draft Version?

Hi,

in the DIME working group we are working on a Diameter QoS document and=20
some of us believe that the QSPEC draft would provide good input with=20
respect to the QoS parameter description.

When can we expect an updated draft to have some details to look at?

Ciao
Hannes

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Tue Nov 21 07:46:52 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GmV0z-00066p-EQ; Tue, 21 Nov 2006 07:46:17 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GmV0y-00066T-T3
	for nsis@ietf.org; Tue, 21 Nov 2006 07:46:16 -0500
Received: from mail.gmx.de ([213.165.64.20] helo=mail.gmx.net)
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GmV0x-0001Rf-Eq
	for nsis@ietf.org; Tue, 21 Nov 2006 07:46:16 -0500
Received: (qmail invoked by alias); 21 Nov 2006 12:46:14 -0000
Received: from socks1.netz.sbs.de (EHLO [192.35.17.26]) [192.35.17.26]
	by mail.gmx.net (mp001) with SMTP; 21 Nov 2006 13:46:14 +0100
X-Authenticated: #29516787
Message-ID: <4562F51A.1010506@gmx.net>
Date: Tue, 21 Nov 2006 13:46:18 +0100
From: Hannes Tschofenig <Hannes.Tschofenig@gmx.net>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
Subject: Re: [NSIS] QSPEC Next Draft Version?
References: <9473683187ADC049A855ED2DA739ABCA0DDC7F93@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7F93@KCCLUST06EVS1.ugd.att.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Y-GMX-Trusted: 0
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 02ec665d00de228c50c93ed6b5e4fc1a
Cc: NSIS <nsis@ietf.org>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry,

thanks for the quick response and the detailed feedback.
I am looking forward to review the next version of the document.

Ciao
Hannes

Ash, Gerald R (Jerry), ALABS wrote:
> Hi Hannes,
> 
> The update is in progress, and is being done according to the agreements
> we reached at IETF-67 and subsequently on the list.  
> 
> One unresolved issue remains on the suggestion from Roland to move QoS
> class parameters to QOSM-specific specifications, RE the thread starting
> at http://www1.ietf.org/mail-archive/web/nsis/current/msg07158.html. 
> Discussion followed at
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07194.html
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07195.html
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07206.html
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07207.html
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07209.html
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07212.html
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07213.html
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07220.html
> 
> There are arguments both ways, and for me the last post from Al is an
> important point that I agree with.  Barring a further 'consensus' call
> on this issue from the chairs, I'll proceed for now with QoS class
> parameters remaining in the draft.  This can always be changed later of
> course if we agree to do so.
> 
> I hope to publish the next version by next week.
> 
> Thanks,
> Regards,
> Jerry 
> 
> -----Original Message-----
> From: Hannes Tschofenig [mailto:Hannes.Tschofenig@gmx.net] 
> Sent: Monday, November 20, 2006 5:06 PM
> To: NSIS
> Subject: [NSIS] QSPEC Next Draft Version?
> 
> Hi,
> 
> in the DIME working group we are working on a Diameter QoS document and 
> some of us believe that the QSPEC draft would provide good input with 
> respect to the QoS parameter description.
> 
> When can we expect an updated draft to have some details to look at?
> 
> Ciao
> Hannes
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 23 02:57:28 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gn9Rh-0004og-AT; Thu, 23 Nov 2006 02:56:33 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gn9Rg-0004oV-8W
	for nsis@ietf.org; Thu, 23 Nov 2006 02:56:32 -0500
Received: from mgw-ext12.nokia.com ([131.228.20.171])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gn9Re-0006uW-08
	for nsis@ietf.org; Thu, 23 Nov 2006 02:56:32 -0500
Received: from esebh107.NOE.Nokia.com (esebh107.ntc.nokia.com [172.21.143.143])
	by mgw-ext12.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kAN7tXIQ020236; Thu, 23 Nov 2006 09:56:18 +0200
Received: from esebh103.NOE.Nokia.com ([172.21.143.33]) by
	esebh107.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 23 Nov 2006 09:55:11 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh103.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 23 Nov 2006 09:55:13 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Date: Thu, 23 Nov 2006 09:55:10 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCCED0780@esebe199.NOE.Nokia.com>
In-Reply-To: <7.0.1.0.0.20061120115123.01300370@att.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local
	admissioncontroloperation in NSLP/QSPEC
Thread-Index: AccMxuozJqpzhPVjQ/C3ME5ggmNEPgCDXdeA
From: <john.loughney@nokia.com>
To: <acmorton@att.com>, <nsis@ietf.org>
X-OriginalArrivalTime: 23 Nov 2006 07:55:13.0085 (UTC)
	FILETIME=[B4B33ED0:01C70ED4]
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 0bc60ec82efc80c84b8d02f4b0e4de22
Cc: 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all,

>It seems to me there is benefit in keeping the Y.1541 class in=20
>the QSPEC, because the Available class number could be=20
>modified along the reservation path. Even if the individual=20
>parameters change in the "QoS Available", it would be=20
>important recognize that a (for example, delay) threshold has=20
>been crossed and that a lesser class is now all that is=20
>available. This seems consistent with:
>
>>    QoS Available: QSPEC object containing parameters describing the
>>    available resources.  They are used to collect information along a
>>    reservation path.
>
>(I had wanted to make this comment during the meeting, but was=20
>busy taking notes for my friend, Hannes)

At this point, I don't think I have seen any strong consensus for
keeping
the Y.1541 class in the QSPEC. I think it would be useful when Y.1541
QoS
models are involved in the e2e signaling, but its use as a general QoS
parameter is debatable.  Additionally, I don't think there has been
support
for this from other QoS models, nor many people speaking up in favor for
keeping this as a general parameter.

John=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 23 08:39:11 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnEmN-0007uO-7E; Thu, 23 Nov 2006 08:38:15 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnEmM-0007qQ-2F
	for nsis@ietf.org; Thu, 23 Nov 2006 08:38:14 -0500
Received: from mo-p00-ob.rzone.de ([81.169.146.162])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GnEmJ-000469-U5
	for nsis@ietf.org; Thu, 23 Nov 2006 08:38:14 -0500
Received: from [192.168.178.33] (p5484AFD9.dip0.t-ipconnect.de
	[84.132.175.217]) by post.webmailer.de (mrclete mo48) (RZmta
	3.4) with ESMTP id iANB41J905410I;
	Thu, 23 Nov 2006 14:37:26 +0100 (MET)
Date: Thu, 23 Nov 2006 14:37:26 +0100 (MET)
Message-ID: <4565A3AA.7060306@cs.uni-goettingen.de>
From: Bernd Schloer <bschloer@cs.uni-goettingen.de>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.12) Gecko/20050920
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 7bac9cb154eb5790ae3b2913587a40de
Cc: nsis@ietf.org
Subject: [NSIS] QSPEC Procedures Recv.-Initiated Reservation
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry,

in section 6.4.2 the text says that the RESERVE includes a QoS Available object if
the sender included a Minimum QoS in the Query. That is the difference between case
1 and 2. Then in case 2 the RESERVE includes a Minimum QoS if it was included in the
QUERY. Why is that Min.QoS in brackets? Is this optional? And is this the QNI which
sends the RESERVE?

Thanks,
Bernd

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 23 09:00:49 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnF7X-0002au-LM; Thu, 23 Nov 2006 09:00:07 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnF7U-0002a9-IA
	for nsis@ietf.org; Thu, 23 Nov 2006 09:00:04 -0500
Received: from mail120.messagelabs.com ([216.82.250.83])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GnF7P-0008GZ-Om
	for nsis@ietf.org; Thu, 23 Nov 2006 09:00:04 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-3.tower-120.messagelabs.com!1164290398!17018845!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 10038 invoked from network); 23 Nov 2006 13:59:58 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-3.tower-120.messagelabs.com with SMTP;
	23 Nov 2006 13:59:58 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kANDxw7c003846; 
	Thu, 23 Nov 2006 08:59:58 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kANDxibE003802; Thu, 23 Nov 2006 08:59:54 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting
	domain-localadmissioncontroloperation in NSLP/QSPEC
Date: Thu, 23 Nov 2006 07:59:44 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7FA8@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <BAA65A575825454CBB0103267553FCCCED0780@esebe199.NOE.Nokia.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting
	domain-localadmissioncontroloperation in NSLP/QSPEC
Thread-Index: AccMxuozJqpzhPVjQ/C3ME5ggmNEPgCDXdeAAAwg5QA=
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: <john.loughney@nokia.com>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 8b431ad66d60be2d47c7bfeb879db82c
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>, "Morton, Alfred C,
	JR \(Al\), ALABS" <acmorton@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi John,

Happy Thanksgiving (for U.S. people anyway)!=20

> -----Original Message-----
> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]=20
> Sent: Thursday, November 23, 2006 2:55 AM
> To: Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
> Subject: RE: [NSIS] Alternatives for supporting=20
> domain-localadmissioncontroloperation in NSLP/QSPEC
>=20
> Hi all,
>=20
> >It seems to me there is benefit in keeping the Y.1541 class in=20
> >the QSPEC, because the Available class number could be=20
> >modified along the reservation path. Even if the individual=20
> >parameters change in the "QoS Available", it would be=20
> >important recognize that a (for example, delay) threshold has=20
> >been crossed and that a lesser class is now all that is=20
> >available. This seems consistent with:
> >
> >>    QoS Available: QSPEC object containing parameters describing the
> >>    available resources.  They are used to collect information along
> >>    a reservation path.
> >
> >(I had wanted to make this comment during the meeting, but was=20
> >busy taking notes for my friend, Hannes)

> At this point, I don't think I have seen any strong consensus for
> keeping the Y.1541 class in the QSPEC. I think it would be useful when
> Y.1541 QoS models are involved in the e2e signaling, but its use as a
> general QoS parameter is debatable.  Additionally, I don't think there
> has been support for this from other QoS models, nor many people
> speaking up in favor for keeping this as a general parameter.

I'll of course abide by your call here, and also remove <PHB Class> and
<DSTE Class Type> as per Roland's suggestion.  I'll note however

1. These parameters have been included in the QSPEC document from the
beginning, that is, from the -00 version over 2 years ago.
2. The document passed WGLC last February with these parameters in; IMO
that is consensus to leave in.
3. We just recently had the suggestion to remove these parameters, and
maybe 2 votes of support; it doesn't seem like a strong consensus to
remove.
4. Removing these parameters seems counter to support for DiffServ, as
originally mandated for the QSPEC, especially removing <PHB Class>.

Thanks,
Regards,
Jerry

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 23 09:01:26 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnF8D-0002nX-JW; Thu, 23 Nov 2006 09:00:49 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnF8B-0002mn-Gj
	for nsis@ietf.org; Thu, 23 Nov 2006 09:00:47 -0500
Received: from kyoto.netlab.nec.de ([195.37.70.21])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GnF88-0008Lf-7X
	for nsis@ietf.org; Thu, 23 Nov 2006 09:00:47 -0500
Received: from [10.1.1.109] (mito.netlab.nec.de [195.37.70.39])
	by kyoto.netlab.nec.de (Postfix) with ESMTP id BDC2913CF86
	for <nsis@ietf.org>; Thu, 23 Nov 2006 15:04:08 +0100 (CET)
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Transfer-Encoding: 7bit
Message-Id: <457BBB50-375F-4656-9064-4BA4A3D87A5B@netlab.nec.de>
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
To: nsis <nsis@ietf.org>
From: Martin Stiemerling <stiemerling@netlab.nec.de>
Date: Thu, 23 Nov 2006 15:00:40 +0100
X-Mailer: Apple Mail (2.752.2)
X-Spam-Score: 1.1 (+)
X-Scan-Signature: 30ac594df0e66ffa5a93eb4c48bcb014
Subject: [NSIS] New document editor for
	draft-ietf-nsis-applicability-mobility-signaling
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi all,

Takako Sanda (sanda.takako at jp.panasonic.com) is taking over the  
editor's role for draft-ietf-nsis-applicability-mobility-signaling.  
This is effective by today.

Thanks to Takako for taking the editorship.

   Martin

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 23 09:09:34 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnFG2-00060N-Ra; Thu, 23 Nov 2006 09:08:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnFG2-000604-5u
	for nsis@ietf.org; Thu, 23 Nov 2006 09:08:54 -0500
Received: from mail121.messagelabs.com ([216.82.241.195])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GnFFz-0001XY-Up
	for nsis@ietf.org; Thu, 23 Nov 2006 09:08:54 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-8.tower-121.messagelabs.com!1164290931!12077544!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 1055 invoked from network); 23 Nov 2006 14:08:51 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-8.tower-121.messagelabs.com with SMTP;
	23 Nov 2006 14:08:51 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kANE8oO2006550; 
	Thu, 23 Nov 2006 09:08:50 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kANE8lEF006526; Thu, 23 Nov 2006 09:08:47 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Date: Thu, 23 Nov 2006 08:08:47 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7FA9@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <4565A3AA.7060306@cs.uni-goettingen.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: QSPEC Procedures Recv.-Initiated Reservation
Thread-Index: AccPBKLgw/MD8HjfQ7m8sUppawl0VwAAzmPQ
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: "Bernd Schloer" <bschloer@cs.uni-goettingen.de>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e5ba305d0e64821bf3d8bc5d3bb07228
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>,
	Cornelia Kappler <cornelia.kappler@gmx.net>, nsis@ietf.org,
	"Kappler, Cornelia" <cornelia.kappler@siemens.com>
Subject: [NSIS] RE: QSPEC Procedures Recv.-Initiated Reservation
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Bernd,

> -----Original Message-----
> From: Bernd Schloer [mailto:bschloer@cs.uni-goettingen.de]=20
> Sent: Thursday, November 23, 2006 8:37 AM
> To: Ash, Gerald R (Jerry), ALABS
> Cc: nsis@ietf.org
> Subject: QSPEC Procedures Recv.-Initiated Reservation
>=20
> Hi Jerry,
>=20
> in section 6.4.2 the text says that the RESERVE includes a=20
> QoS Available object if
> the sender included a Minimum QoS in the Query. That is the=20
> difference between case
> 1 and 2. Then in case 2 the RESERVE includes a Minimum QoS if=20
> it was included in the
> QUERY. Why is that Min.QoS in brackets? Is this optional? And=20
> is this the QNI which sends the RESERVE?

I think the brackets should be removed in case 2, I'm not sure why they
are there (Cornelia?).  And yes, it is the QNI in this case that sends
the RESERVE.

Thanks,
Jerry

>=20
> Thanks,
> Bernd
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 23 10:55:07 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnGu1-0007ln-Ff; Thu, 23 Nov 2006 10:54:17 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnGu0-0007kS-6o
	for nsis@ietf.org; Thu, 23 Nov 2006 10:54:16 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GnGty-0002hH-Rd
	for nsis@ietf.org; Thu, 23 Nov 2006 10:54:16 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kANFs2Rx005519; Thu, 23 Nov 2006 16:54:09 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>
Date: Thu, 23 Nov 2006 16:53:59 +0100
Message-ID: <002e01c70f17$9c8bf1b0$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: AccPF5bno8SLBfLYRw6xVaUM5nQz7Q==
X-Spam-Score: 0 () 
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Thu, 23 Nov 2006 16:54:10 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: e1e48a527f609d1be2bc8d8a70eb76cb
Cc: 'NSIS Group' <nsis@ietf.org>
Subject: [NSIS] An issue related to NOTIFY
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Hi Jerry

One of the issues that I have posted some weeks before the IETF meeting in
San Diego
was QSPEC-Issue (GK3), which was independent of the mandatory/optional
parameter 
discussion, see below.

In Section 6.4.5 it is mentioned:
"No QSPEC is carried in the NOTIFY message. The NOTIFY message carries only
the 
Session ID and a INFO_SPEC with the error code as described in [QOS-SIG]."

Note that a NOTIFY is allowed to carry a QSPEC, see QoS-NSLP draft.
Therefore, it should be allowed for QOSMs to specify, also in the case of
preemption, 
to send a QSPEC together with the INFO_SPEC object.

Furthermore please note that a NOTIFY does not carry a Session ID.


Requested change:
Modify text in Section 6.4.5 (Preemption):
From:
"No QSPEC is carried in the NOTIFY message. The NOTIFY message carries only
the 
Session ID and a INFO_SPEC with the error code as described in [QOS-SIG]."

INTO:
"The NOTIFY message carries an  INFO_SPEC with the error code as described 
in [QOS-SIG]. A QOSM specification document may specify wheter a NOTIFY also

carries a QSPEC object."

Could you please make these modifications?

Best regards,
Georgios



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 24 04:33:21 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnXPi-0005Wa-PD; Fri, 24 Nov 2006 04:32:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnXPh-0005WU-2L
	for nsis@ietf.org; Fri, 24 Nov 2006 04:32:05 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GnXPf-0001iX-Jx
	for nsis@ietf.org; Fri, 24 Nov 2006 04:32:05 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kAO9VnVd009019; Fri, 24 Nov 2006 10:31:57 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>,
	<john.loughney@nokia.com>
Subject: RE: [NSIS] Alternatives for
	supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
Date: Fri, 24 Nov 2006 10:31:47 +0100
Message-ID: <000901c70fab$62b9da80$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
Thread-Index: AccMxuozJqpzhPVjQ/C3ME5ggmNEPgCDXdeAAAwg5QAAKY7hIA==
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7FA8@KCCLUST06EVS1.ugd.att.com>
X-Spam-Score: 0.9 () FVGT_s_OBFU_Q1,J_CHICKENPOX_12
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Fri, 24 Nov 2006 10:31:58 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 057ebe9b96adec30a7efb2aeda4c26a4
Cc: "'Morton, Alfred C,JR \(Al\), ALABS'" <acmorton@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry

What about my proposal, sent in a previous e-mail!

"But the QSPEC template draft is not a generic QSPEC draft.
The Token Bucket and the constraints can be seen as something that can be 
common and can be used for the interoperability and the rest of the 
encoded parameters can just be used as guidelines on how the QoS 
models can encode their parameters.

Thus maybe Jerry could separate the QSPEC template draft into:
A section/part that describes the Token Bucket + constraints. 
This part could then emphasize that the described parameters 
can be used for interoperability purposes.

Another part of the QSPEC template draft will describe 
the other parameters.

The goal of this part is to provide guidelines to QoS models 
on how to encode the different parameters, with the currently 
available parameters used as examples."

Best Regards,
Georgios

 

> -----Original Message-----
> From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com] 
> Sent: donderdag 23 november 2006 15:00
> To: john.loughney@nokia.com
> Cc: Ash, Gerald R (Jerry), ALABS; Morton, Alfred C,JR (Al), 
> ALABS; nsis@ietf.org
> Subject: RE: [NSIS] Alternatives for 
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
> 
> Hi John,
> 
> Happy Thanksgiving (for U.S. people anyway)! 
> 
> > -----Original Message-----
> > From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
> > Sent: Thursday, November 23, 2006 2:55 AM
> > To: Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
> > Subject: RE: [NSIS] Alternatives for supporting 
> > domain-localadmissioncontroloperation in NSLP/QSPEC
> > 
> > Hi all,
> > 
> > >It seems to me there is benefit in keeping the Y.1541 class in the 
> > >QSPEC, because the Available class number could be 
> modified along the 
> > >reservation path. Even if the individual parameters change in the 
> > >"QoS Available", it would be important recognize that a 
> (for example, 
> > >delay) threshold has been crossed and that a lesser class 
> is now all 
> > >that is available. This seems consistent with:
> > >
> > >>    QoS Available: QSPEC object containing parameters 
> describing the
> > >>    available resources.  They are used to collect 
> information along
> > >>    a reservation path.
> > >
> > >(I had wanted to make this comment during the meeting, but 
> was busy 
> > >taking notes for my friend, Hannes)
> 
> > At this point, I don't think I have seen any strong consensus for 
> > keeping the Y.1541 class in the QSPEC. I think it would be 
> useful when
> > Y.1541 QoS models are involved in the e2e signaling, but 
> its use as a 
> > general QoS parameter is debatable.  Additionally, I don't 
> think there 
> > has been support for this from other QoS models, nor many people 
> > speaking up in favor for keeping this as a general parameter.
> 
> I'll of course abide by your call here, and also remove <PHB 
> Class> and <DSTE Class Type> as per Roland's suggestion.  
> I'll note however
> 
> 1. These parameters have been included in the QSPEC document 
> from the beginning, that is, from the -00 version over 2 years ago.
> 2. The document passed WGLC last February with these 
> parameters in; IMO that is consensus to leave in.
> 3. We just recently had the suggestion to remove these 
> parameters, and maybe 2 votes of support; it doesn't seem 
> like a strong consensus to remove.
> 4. Removing these parameters seems counter to support for 
> DiffServ, as originally mandated for the QSPEC, especially 
> removing <PHB Class>.
> 
> Thanks,
> Regards,
> Jerry
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 24 04:37:31 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnXUK-0006QU-I9; Fri, 24 Nov 2006 04:36:52 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnXUJ-0006Pz-CB
	for nsis@ietf.org; Fri, 24 Nov 2006 04:36:51 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GnXUF-00035n-Rj
	for nsis@ietf.org; Fri, 24 Nov 2006 04:36:51 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GnXU7-0004Bj-8U; Fri, 24 Nov 2006 10:36:40 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GnXU6-0006w5-Bh; Fri, 24 Nov 2006 10:36:38 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id CD33CB587;
	Fri, 24 Nov 2006 10:36:32 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GnXTy-0007ZK-1x; Fri, 24 Nov 2006 10:36:30 +0100
Message-ID: <4566BD1D.4090402@tm.uka.de>
Date: Fri, 24 Nov 2006 10:36:29 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: "Ash, Gerald R (Jerry), ALABS" <gash@att.com>
Subject: Re: [NSIS] Alternatives for
	supporting	domain-localadmissioncontroloperation in NSLP/QSPEC
References: <9473683187ADC049A855ED2DA739ABCA0DDC7FA8@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <9473683187ADC049A855ED2DA739ABCA0DDC7FA8@KCCLUST06EVS1.ugd.att.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -3.6 (---)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.8 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 97adf591118a232206bdb5a27b217034
Cc: john.loughney@nokia.com, "Morton, Alfred C, JR \(Al\),
	ALABS" <acmorton@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Jerry,

> I'll of course abide by your call here, and also remove <PHB Class> and
> <DSTE Class Type> as per Roland's suggestion.  I'll note however
> 
> 1. These parameters have been included in the QSPEC document from the
> beginning, that is, from the -00 version over 2 years ago.
> 2. The document passed WGLC last February with these parameters in; IMO
> that is consensus to leave in.

You are somewhat right, but
absence of comments is not consensus. However, I guess that the
situation changed when Dave Oran kicked in.

> 3. We just recently had the suggestion to remove these parameters, and
> maybe 2 votes of support; it doesn't seem like a strong consensus to
> remove.

I guess that I'm amongst the 2 votes.

> 4. Removing these parameters seems counter to support for DiffServ, as
> originally mandated for the QSPEC, especially removing <PHB Class>.

I don't want to repeat here what I already said in a previous mail,
but signaling a PHB class has usually only real significance within
a DiffServ domain, but not between domains.
Actually, if there is consensus to leave the Y.1541 class in:
what is the procedure if the mapping from the generic (available)
parameters doesn't correspond to the class? Which specification
takes precedence?

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 24 09:17:55 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnbrP-0006tq-2a; Fri, 24 Nov 2006 09:16:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnbrN-0006tl-SM
	for nsis@ietf.org; Fri, 24 Nov 2006 09:16:57 -0500
Received: from mail146.messagelabs.com ([216.82.245.131])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GnbrI-00011F-Il
	for nsis@ietf.org; Fri, 24 Nov 2006 09:16:57 -0500
X-VirusChecked: Checked
X-Env-Sender: acmorton@att.com
X-Msg-Ref: server-5.tower-146.messagelabs.com!1164377811!5658899!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 4300 invoked from network); 24 Nov 2006 14:16:51 -0000
Received: from unknown (HELO attrh0i.attrh.att.com) (134.24.146.4)
	by server-5.tower-146.messagelabs.com with SMTP;
	24 Nov 2006 14:16:51 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh0i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAOEGpOW009178
	for <nsis@ietf.org>; Fri, 24 Nov 2006 09:16:51 -0500 (EST)
Received: from maillennium.att.com (dns.maillennium.att.com [135.25.114.99])
	by attrh0i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAOEGdhc009134
	for <nsis@ietf.org>; Fri, 24 Nov 2006 09:16:44 -0500 (EST)
Received: from acmt.att.com (unknown[135.210.81.153](misconfigured sender))
	by maillennium.att.com (mailgw1) with SMTP
	id <20061124141638gw100r41dce>; Fri, 24 Nov 2006 14:16:38 +0000
Message-Id: <7.0.1.0.0.20061124084324.01ed9ee8@att.com>
X-Mailer: QUALCOMM Windows Eudora Version 7.0.1.0
Date: Fri, 24 Nov 2006 09:16:37 -0500
To: Roland Bless <bless@tm.uka.de>,
	"Ash, Gerald R (Jerry), ALABS" <gash@att.com>
From: Al Morton <acmorton@att.com>
Subject: Re: [NSIS] Alternatives for
	supporting	domain-localadmissioncontroloperation in NSLP/QSPEC
In-Reply-To: <4566BD1D.4090402@tm.uka.de>
References: <9473683187ADC049A855ED2DA739ABCA0DDC7FA8@KCCLUST06EVS1.ugd.att.com>
	<4566BD1D.4090402@tm.uka.de>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 97adf591118a232206bdb5a27b217034
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

At 04:36 AM 11/24/2006, Roland Bless wrote:
>...Actually, if there is consensus to leave the Y.1541 class in:
>what is the procedure if the mapping from the generic (available)
>parameters doesn't correspond to the class? Which specification
>takes precedence?

Hi Roland,

The mapping seems straightforward to me, but perhaps I've
misunderstood your question. Let me give a brief example
to see if we are using the term "mapping" the same way.

A key point first:  The numerical objectives in the Y.1541
classes are upper bounds on performance. So, the mapping is
really a comparison between the current level of each
performance metric and the upper bound.

This example is based on latency alone:
Say that the Desired Y.1541 class = 0, which has a
latency bound of <=100ms. While the available latency meets
this bound, the available Y.1541 class = 0.

At some point along the path, the available latency is
found to be 120ms. The available Y.1541 class must change
to class = 1, because the 100ms latency bound of class 0
has been exceeded. This may be a significant change for
some users.

In general, the available Y.1541 class is the most demanding
class that can be met by the available performance.

Hope this helps, (and if not, I'll be glad to try again)
Al


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 24 11:16:59 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gndif-0003tG-Dj; Fri, 24 Nov 2006 11:16:05 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gndie-0003tA-EE
	for nsis@ietf.org; Fri, 24 Nov 2006 11:16:04 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gndic-0007CF-V4
	for nsis@ietf.org; Fri, 24 Nov 2006 11:16:04 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GndiW-0003B8-Uz; Fri, 24 Nov 2006 17:15:59 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GndiW-0007hf-1s; Fri, 24 Nov 2006 17:15:56 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 70659B5A2;
	Fri, 24 Nov 2006 17:15:55 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GndiV-0007xq-57; Fri, 24 Nov 2006 17:15:55 +0100
Message-ID: <45671ABA.1090707@tm.uka.de>
Date: Fri, 24 Nov 2006 17:15:54 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Al Morton <acmorton@att.com>
Subject: Re: [NSIS] Alternatives for
	supporting	domain-localadmissioncontroloperation in NSLP/QSPEC
References: <9473683187ADC049A855ED2DA739ABCA0DDC7FA8@KCCLUST06EVS1.ugd.att.com>
	<4566BD1D.4090402@tm.uka.de>
	<7.0.1.0.0.20061124084324.01ed9ee8@att.com>
In-Reply-To: <7.0.1.0.0.20061124084324.01ed9ee8@att.com>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -3.6 (---)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.8 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: d185fa790257f526fedfd5d01ed9c976
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Al,

thanks for your answer. My comments inline.

Al Morton wrote:
> At 04:36 AM 11/24/2006, Roland Bless wrote:
>> ...Actually, if there is consensus to leave the Y.1541 class in:
>> what is the procedure if the mapping from the generic (available)
>> parameters doesn't correspond to the class? Which specification
>> takes precedence?

> The mapping seems straightforward to me, but perhaps I've
> misunderstood your question. Let me give a brief example
> to see if we are using the term "mapping" the same way.
> 
> A key point first:  The numerical objectives in the Y.1541
> classes are upper bounds on performance. So, the mapping is
> really a comparison between the current level of each
> performance metric and the upper bound.

Yes, I'm aware of this. However, intermediate systems
must know the exact values in order to reserve an appropriate
amount of resources, e.g., in your example the current actual
admitted latency.

> This example is based on latency alone:
> Say that the Desired Y.1541 class = 0, which has a
> latency bound of <=100ms. While the available latency meets
> this bound, the available Y.1541 class = 0.

Yep.

> At some point along the path, the available latency is
> found to be 120ms. The available Y.1541 class must change
> to class = 1, because the 100ms latency bound of class 0
> has been exceeded. This may be a significant change for
> some users.

Sure and that's why it should be rejected at this point.
But I guess that the available class is not allowed to be
"changed", because the constraints of that class cannot
be met anymore. In this case it only makes sense to
reject the request (it is not useful to continue with
saying available class=1).

The question I had is:
What about a situation where the QSPEC requests class 0,
but the requested latency upper bound is 150ms ("wrong" mapping)?

> In general, the available Y.1541 class is the most demanding
> class that can be met by the available performance.

That doesn't make sense to me. The class specifies the limits.
So if you cannot meet class 0, then you have to reject the
reservation. It makes no sense to continue the reservation with class 1.
If your application is also satisfied with class 1 it should
say so in its original request.

Furthermore, what about a domain that is able to provide you a
delay of 80ms (would map to class 0), though you requested
only class 1? I guess that a QNE should not change the available
class to 0 in this case.
Possibly there are currently only a few users in
that class 1, so that you get a really low latency service
although you didn't ask for one?
However, traffic for class 0 may be handled by a really
different DiffServ DSCP than that for class 1, i.e. resources
are effectively separated (though they both may use the EF PHB)
within a Y.1541 domain. The major point I see is that the
domain must decide in which of its traffic classes it performs
resource-based admission control. So a boundary QNE should not
follow blindly the available class value, it always has to
check the real values...

> Hope this helps, (and if not, I'll be glad to try again)

Thanks for the example, but it generated actually more questions
on my side :-)

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 24 11:21:32 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GndnL-00054M-3K; Fri, 24 Nov 2006 11:20:55 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GndnK-00051a-Gz
	for nsis@ietf.org; Fri, 24 Nov 2006 11:20:54 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GndnJ-0008Lt-76
	for nsis@ietf.org; Fri, 24 Nov 2006 11:20:54 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1GndnD-0003xT-Ln; Fri, 24 Nov 2006 17:20:52 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1GndnD-0000Y0-7M; Fri, 24 Nov 2006 17:20:47 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 2722CB5A2;
	Fri, 24 Nov 2006 17:20:46 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1GndnB-0007yN-Rc; Fri, 24 Nov 2006 17:20:45 +0100
Message-ID: <45671BDD.8090608@tm.uka.de>
Date: Fri, 24 Nov 2006 17:20:45 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: "Hancock, Robert" <robert.hancock@roke.co.uk>,
	"nsis@ietf.org" <nsis@ietf.org>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-15
Content-Transfer-Encoding: 7bit
X-Spam-Score: -3.6 (---)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.8 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ffa9dfbbe7cc58b3fa6b8ae3e57b0aa3
Cc: 
Subject: [NSIS] Retransmitted Queries
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Robert and all,

In section 5.3.3. there is the statement:

  Retransmitted Queries MUST use different Query-Cookie values.

Is there any rationale for this (yes, I read sec. 8.5)?
I cannot see any real benefit in doing
so (except for the case that you want to measure round-trip times).
Replay attacks seem to cause no harm, because they do not
cause additional state to be installed. In the current situation,
we have often seen the case where retransmissions and responses
are crossing each other (or two queries are transmitted back-to-back)
and the responder has to send one response for each query until the
backoff is large enough to allow a query cookie match for the latest
query/response pair. If we require that every Query must be answered
by a Response (in order to recover from lost Responses) I see an
advantage of using the same query cookie for retransmissions:
The context will be established earlier, as soon as
the first response hits the querier, so the latter will stop the
retransmission of queries. Subsequent responses can be dropped
because state was already established at the querier.
I guess that I overlooked something...

Regards,
 Roland

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Fri Nov 24 11:30:22 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gndw1-0003Bq-8C; Fri, 24 Nov 2006 11:29:53 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gndvz-0003BT-Lb
	for nsis@ietf.org; Fri, 24 Nov 2006 11:29:51 -0500
Received: from iramx2.ira.uni-karlsruhe.de ([141.3.10.81])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gndvy-0003In-8X
	for nsis@ietf.org; Fri, 24 Nov 2006 11:29:51 -0500
Received: from irams1.ira.uni-karlsruhe.de ([141.3.10.5])
	by iramx2.ira.uni-karlsruhe.de with esmtps 
	id 1Gndvl-0005Tb-NH; Fri, 24 Nov 2006 17:29:43 +0100
Received: from i72ms2.tm.uni-karlsruhe.de ([141.3.70.17]
	helo=smtp.ipv6.tm.uni-karlsruhe.de)
	by irams1.ira.uni-karlsruhe.de with esmtps 
	id 1Gndvl-0001fH-4x; Fri, 24 Nov 2006 17:29:37 +0100
Received: from vorta.ipv6.tm.uni-karlsruhe.de (vorta.ipv6.tm.uni-karlsruhe.de
	[IPv6:2001:638:204:6:207:e9ff:fe0c:5e44])
	by smtp.ipv6.tm.uni-karlsruhe.de (Postfix) with ESMTP id 3F501B5A2;
	Fri, 24 Nov 2006 17:29:36 +0100 (CET)
Received: from localhost ([127.0.0.1])
	by vorta.ipv6.tm.uni-karlsruhe.de with esmtp (Exim 4.44)
	id 1Gndvj-0007yc-SG; Fri, 24 Nov 2006 17:29:35 +0100
Message-ID: <45671DEF.9030104@tm.uka.de>
Date: Fri, 24 Nov 2006 17:29:35 +0100
From: Roland Bless <bless@tm.uka.de>
Organization: Institute of Telematics, University of Karlsruhe
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
To: Georgios Karagiannis <karagian@cs.utwente.nl>
Subject: Re: [NSIS] Alternatives
	for	supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
References: <000901c70fab$62b9da80$4c0d5982@dynamic.ewi.utwente.nl>
In-Reply-To: <000901c70fab$62b9da80$4c0d5982@dynamic.ewi.utwente.nl>
X-Enigmail-Version: 0.94.1.1
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
X-Spam-Score: -3.6 (---)
X-Spam-Status: No
X-Spam-Report: -1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
	0.8 AWL AWL: From: address is in the auto white-list
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a7d6aff76b15f3f56fcb94490e1052e4
Cc: john.loughney@nokia.com, "'Ash, Gerald R \(Jerry\), ALABS'" <gash@att.com>,
	"'Morton, Alfred C, JR \(Al\), ALABS'" <acmorton@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

see comments inline.

Georgios Karagiannis wrote:
> What about my proposal, sent in a previous e-mail!
> 
> "But the QSPEC template draft is not a generic QSPEC draft.
> The Token Bucket and the constraints can be seen as something that can be 
> common and can be used for the interoperability and the rest of the 
> encoded parameters can just be used as guidelines on how the QoS 
> models can encode their parameters.
> 
> Thus maybe Jerry could separate the QSPEC template draft into:
> A section/part that describes the Token Bucket + constraints. 
> This part could then emphasize that the described parameters 
> can be used for interoperability purposes.
> 
> Another part of the QSPEC template draft will describe 
> the other parameters.

As I said at the meeting:
the current QSPEC template draft actually contains both aspects
since we started with the template idea first and then added the
interoperability aspects. An even cleaner separation would be
a split into two separate documents:
- one (template focus) explaining the general idea and layout of QSPECs,
  which could be used by other QoS models as guidance to create their
  own QSPECs, and
- one (interoperability aspects) proposing a generic QSPEC as
  an example of a QSPEC instance
However, it's also ok for me to describe both aspects in one document
as it is probably now too late and too disturbing to make this
separation.

> The goal of this part is to provide guidelines to QoS models 
> on how to encode the different parameters, with the currently 
> available parameters used as examples."

Regards,
 Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sat Nov 25 07:46:00 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gnwtv-0006tL-JL; Sat, 25 Nov 2006 07:44:59 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gnwtt-0006tF-Uw
	for nsis@ietf.org; Sat, 25 Nov 2006 07:44:57 -0500
Received: from mail120.messagelabs.com ([216.82.250.83])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1Gnwts-0002Nj-Ji
	for nsis@ietf.org; Sat, 25 Nov 2006 07:44:57 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-5.tower-120.messagelabs.com!1164458695!14343522!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 27509 invoked from network); 25 Nov 2006 12:44:55 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-5.tower-120.messagelabs.com with SMTP;
	25 Nov 2006 12:44:55 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAPCisfP009342; 
	Sat, 25 Nov 2006 07:44:54 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kAPCiluN009315; Sat, 25 Nov 2006 07:44:48 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local admission control
	operation in NSLP/QSPEC
Date: Sat, 25 Nov 2006 06:44:47 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7FB0@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <45671DEF.9030104@tm.uka.de>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local admission
	control operation in NSLP/QSPEC
Thread-Index: AccP5ceRsNCbNznGQaCPBDUG9f2U+wAqQFkg
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: "Roland Bless" <bless@tm.uka.de>,
	"Georgios Karagiannis" <karagian@cs.utwente.nl>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 244a2fd369eaf00ce6820a760a3de2e8
Cc: john.loughney@nokia.com, "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>,
	"Morton, Alfred C, JR \(Al\), ALABS" <acmorton@att.com>,
	nsis@ietf.org, cdvorak@research.att.com
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Roland, Georgios,

See below.=20

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de]=20
> Sent: Friday, November 24, 2006 11:30 AM
> To: Georgios Karagiannis
> Cc: Ash, Gerald R (Jerry), ALABS; john.loughney@nokia.com;=20
> Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
> Subject: Re: [NSIS] Alternatives for=20
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>=20
> Hi Georgios,
>=20
> see comments inline.
>=20
> Georgios Karagiannis wrote:
> > What about my proposal, sent in a previous e-mail!
> >=20
> > "But the QSPEC template draft is not a generic QSPEC draft.
> > The Token Bucket and the constraints can be seen as=20
> something that can be=20
> > common and can be used for the interoperability and the rest of the=20
> > encoded parameters can just be used as guidelines on how the QoS=20
> > models can encode their parameters.
> >=20
> > Thus maybe Jerry could separate the QSPEC template draft into:
> > A section/part that describes the Token Bucket + constraints.=20
> > This part could then emphasize that the described parameters=20
> > can be used for interoperability purposes.
> >=20
> > Another part of the QSPEC template draft will describe=20
> > the other parameters.
> > The goal of this part is to provide guidelines to QoS models=20
> > on how to encode the different parameters, with the currently=20
> > available parameters used as examples."
>=20
> As I said at the meeting:
> the current QSPEC template draft actually contains both aspects
> since we started with the template idea first and then added the
> interoperability aspects. An even cleaner separation would be
> a split into two separate documents:
> - one (template focus) explaining the general idea and layout=20
> of QSPECs,
>   which could be used by other QoS models as guidance to create their
>   own QSPECs, and
> - one (interoperability aspects) proposing a generic QSPEC as
>   an example of a QSPEC instance
> However, it's also ok for me to describe both aspects in one document
> as it is probably now too late and too disturbing to make this
> separation.

OK, Georgios' proposal sounds like a good compromise.

Thanks,
Regards,
Jerry
=20
>=20
> Regards,
>  Roland
>=20
>=20

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sat Nov 25 07:50:04 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GnwyF-0001F8-44; Sat, 25 Nov 2006 07:49:27 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GnwyD-0001Dz-R4
	for nsis@ietf.org; Sat, 25 Nov 2006 07:49:25 -0500
Received: from mail120.messagelabs.com ([216.82.250.83])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GnwyC-0003Rv-Gk
	for nsis@ietf.org; Sat, 25 Nov 2006 07:49:25 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-10.tower-120.messagelabs.com!1164458962!9328971!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 30976 invoked from network); 25 Nov 2006 12:49:23 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-10.tower-120.messagelabs.com with SMTP;
	25 Nov 2006 12:49:23 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAPCnMWS010215; 
	Sat, 25 Nov 2006 07:49:22 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kAPCnKuJ010209; Sat, 25 Nov 2006 07:49:20 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Date: Sat, 25 Nov 2006 06:49:20 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7FB1@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <002e01c70f17$9c8bf1b0$4c0d5982@dynamic.ewi.utwente.nl>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: An issue related to NOTIFY
Thread-Index: AccPF5bno8SLBfLYRw6xVaUM5nQz7QBeFtbA
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: "Georgios Karagiannis" <karagian@cs.utwente.nl>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: f607d15ccc2bc4eaf3ade8ffa8af02a0
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>, NSIS Group <nsis@ietf.org>
Subject: [NSIS] RE: An issue related to NOTIFY
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Georgios,

Your proposed change looks OK to me.

Thanks,
Regards,
Jerry=20

-----Original Message-----
From: Georgios Karagiannis [mailto:karagian@cs.utwente.nl]=20
Sent: Thursday, November 23, 2006 10:54 AM
To: Ash, Gerald R (Jerry), ALABS
Cc: 'NSIS Group'
Subject: An issue related to NOTIFY

Hi Jerry

One of the issues that I have posted some weeks before the IETF meeting
in
San Diego
was QSPEC-Issue (GK3), which was independent of the mandatory/optional
parameter=20
discussion, see below.

In Section 6.4.5 it is mentioned:
"No QSPEC is carried in the NOTIFY message. The NOTIFY message carries
only
the=20
Session ID and a INFO_SPEC with the error code as described in
[QOS-SIG]."

Note that a NOTIFY is allowed to carry a QSPEC, see QoS-NSLP draft.
Therefore, it should be allowed for QOSMs to specify, also in the case
of
preemption,=20
to send a QSPEC together with the INFO_SPEC object.

Furthermore please note that a NOTIFY does not carry a Session ID.

Requested change:
Modify text in Section 6.4.5 (Preemption):
From:
"No QSPEC is carried in the NOTIFY message. The NOTIFY message carries
only
the=20
Session ID and a INFO_SPEC with the error code as described in
[QOS-SIG]."

INTO:
"The NOTIFY message carries an  INFO_SPEC with the error code as
described=20
in [QOS-SIG]. A QOSM specification document may specify wheter a NOTIFY
also

carries a QSPEC object."

Could you please make these modifications?

Best regards,
Georgios

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sun Nov 26 16:50:42 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GoRsr-0006kW-O3; Sun, 26 Nov 2006 16:49:57 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GoRsq-0006kR-ME
	for nsis@ietf.org; Sun, 26 Nov 2006 16:49:56 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GoRsp-000419-7G
	for nsis@ietf.org; Sun, 26 Nov 2006 16:49:56 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Sun, 26 Nov 2006 16:49:48 -0500
Received: from 0015-ITS-EXBH01.us.saic.com ([10.43.229.18])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006112616494808867 ; Sun, 26 Nov 2006 16:49:48 -0500
Received: from 0591-ITS-EXMP02.us.saic.com ([10.42.81.248]) by
	0015-ITS-EXBH01.us.saic.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Sun, 26 Nov 2006 16:49:49 -0500
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Date: Sun, 26 Nov 2006 16:49:48 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF3C@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local admissioncontrol
	operation in NSLP/QSPEC
Thread-Index: AccP5ceRsNCbNznGQaCPBDUG9f2U+wAqQFkgAESpuhA=
References: <9473683187ADC049A855ED2DA739ABCA0DDC7FB0@KCCLUST06EVS1.ugd.att.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: "Roland Bless" <bless@tm.uka.de>,
	"Georgios Karagiannis" <karagian@cs.utwente.nl>
X-OriginalArrivalTime: 26 Nov 2006 21:49:49.0232 (UTC)
	FILETIME=[CBA62300:01C711A4]
X-Spam-Score: 0.1 (/)
X-Scan-Signature: 8a85b14f27c9dcbe0719e27d46abc1f8
Cc: john.loughney@nokia.com, "Morton, Alfred C, JR \(Al\),
	ALABS" <acmorton@att.com>, nsis@ietf.org, cdvorak@research.att.com
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1928790704=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1928790704==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C711A4.CB2308AE"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C711A4.CB2308AE
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Ronald and Georgios:
=20
Why is it too late to clean out the document as Ronald has explained?
=20
NSIS WG in particular and IETF in general have waited almost two years =
to root out these "technically" inconsistencies. It is better NOT to =
publish a standard track RFC like this.
=20
If needed, NSIS WG Chair needs to appoint a "single" EDITOR who can =
clean out this document.
=20
Why does the NSIS WG Chair need to appoint 3 (Three) editors?  Is it not =
the problems of appointing 3 editors for the same simple RFC? Is it not =
because of 3 editors who have put those things defying a "simple" common =
sense that to keep the document is the most important thing from the =
IETF point of view?
=20
Is it not that people still trying to force so-called QOS SLAs like =
Y.1541, DSTE, and 3GPP/Y.1541 in this draft as "standard" QOS models?
=20
A question to Georgios: Would you please provide any technical =
justifications why so called "non-standard" QOS SLAs like Y.1541, DSTE, =
and 3GPP/Y.1541 in this "standard" track RFC in the same level of IETF =
"standard-based" RSVP, IntServ, and DiffServ QOS Model? Is it not mixing =
between "apples" and "oranges?"
=20
If Georgios cannot justify the above, references of so called =
"non-standard" QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 as even =
examples for encoding of QSPEC parameters must NOT be included.
=20
If it is done despite objections, we would think that this QSPEC RFC =
must NOT be a "standard" track RFC because a standard track RFC must NOT =
contain "non-standard" QOS models. This will also be brought to the =
notice of the NSIS WG Chairs and IESG.
=20
Best regards,
Radhika

________________________________

From: nsis-bounces@ietf.org on behalf of Ash, Gerald R (Jerry), ALABS
Sent: Sat 11/25/2006 7:44 AM
To: Roland Bless; Georgios Karagiannis
Cc: john.loughney@nokia.com; Ash, Gerald R (Jerry), ALABS; Morton, =
Alfred C, JR (Al), ALABS; nsis@ietf.org; cdvorak@research.att.com
Subject: RE: [NSIS] Alternatives for supporting domain-local admission =
controloperation in NSLP/QSPEC



Hi Roland, Georgios,

See below.

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de]
> Sent: Friday, November 24, 2006 11:30 AM
> To: Georgios Karagiannis
> Cc: Ash, Gerald R (Jerry), ALABS; john.loughney@nokia.com;
> Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
> Subject: Re: [NSIS] Alternatives for
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>
> Hi Georgios,
>
> see comments inline.
>
> Georgios Karagiannis wrote:
> > What about my proposal, sent in a previous e-mail!
> >
> > "But the QSPEC template draft is not a generic QSPEC draft.
> > The Token Bucket and the constraints can be seen as
> something that can be
> > common and can be used for the interoperability and the rest of the
> > encoded parameters can just be used as guidelines on how the QoS
> > models can encode their parameters.
> >
> > Thus maybe Jerry could separate the QSPEC template draft into:
> > A section/part that describes the Token Bucket + constraints.
> > This part could then emphasize that the described parameters
> > can be used for interoperability purposes.
> >
> > Another part of the QSPEC template draft will describe
> > the other parameters.
> > The goal of this part is to provide guidelines to QoS models
> > on how to encode the different parameters, with the currently
> > available parameters used as examples."
>
> As I said at the meeting:
> the current QSPEC template draft actually contains both aspects
> since we started with the template idea first and then added the
> interoperability aspects. An even cleaner separation would be
> a split into two separate documents:
> - one (template focus) explaining the general idea and layout
> of QSPECs,
>   which could be used by other QoS models as guidance to create their
>   own QSPECs, and
> - one (interoperability aspects) proposing a generic QSPEC as
>   an example of a QSPEC instance
> However, it's also ok for me to describe both aspects in one document
> as it is probably now too late and too disturbing to make this
> separation.

OK, Georgios' proposal sounds like a good compromise.

Thanks,
Regards,
Jerry

>
> Regards,
>  Roland



------_=_NextPart_001_01C711A4.CB2308AE
Content-Type: text/html;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<HTML dir=3Dltr><HEAD><TITLE>RE: [NSIS] Alternatives for supporting =
domain-local admission controloperation in NSLP/QSPEC</TITLE>=0A=
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dunicode">=0A=
<META content=3D"MSHTML 6.00.2900.2995" name=3DGENERATOR></HEAD>=0A=
<BODY>=0A=
<DIV id=3DidOWAReplyText23077 dir=3Dltr>=0A=
<DIV dir=3Dltr><FONT face=3DArial color=3D#000000 size=3D2>Ronald and =
Georgios:</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>Why is it too late to clean =
out the document as Ronald has explained?</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>NSIS WG in particular and =
IETF in general have waited almost two years to root out these =
"technically" inconsistencies. It is better NOT to publish a standard =
track RFC like this.</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>If needed, NSIS WG Chair =
needs to appoint a "single" EDITOR who can clean out this =
document.</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>Why does the NSIS WG Chair =
need to appoint 3 (Three) editors?&nbsp;&nbsp;Is it not the problems of =
appointing 3 editors for the same simple RFC? Is it not because of 3 =
editors who have put those things defying a "simple" common sense that =
to keep the document is the most important thing from the IETF point of =
view?</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>Is it not that people still =
trying to force so-called QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 in =
this draft as "standard" QOS models?</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>A question to Georgios: Would =
you please provide&nbsp;any technical justifications why so called =
"non-standard" QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 in this =
"standard" track RFC in the same level of IETF "standard-based" RSVP, =
IntServ, and DiffServ QOS Model? Is it not mixing between "apples" and =
"oranges?"</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>If Georgios cannot justify =
the above, references of so called "non-standard" QOS SLAs like Y.1541, =
DSTE, and 3GPP/Y.1541 as even examples for encoding of QSPEC parameters =
must NOT be included.</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>If it is done despite =
objections, we would think that this QSPEC RFC must NOT be a "standard" =
track RFC because a standard track RFC must NOT contain "non-standard" =
QOS models. This will also be&nbsp;brought to the notice of the NSIS WG =
Chairs and IESG.</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>Best regards,</FONT></DIV>=0A=
<DIV dir=3Dltr><FONT face=3DArial size=3D2>Radhika</FONT></DIV></DIV>=0A=
<DIV dir=3Dltr><BR>=0A=
<HR tabIndex=3D-1>=0A=
<FONT face=3DTahoma size=3D2><B>From:</B> nsis-bounces@ietf.org on =
behalf of Ash, Gerald R (Jerry), ALABS<BR><B>Sent:</B> Sat 11/25/2006 =
7:44 AM<BR><B>To:</B> Roland Bless; Georgios Karagiannis<BR><B>Cc:</B> =
john.loughney@nokia.com; Ash, Gerald R (Jerry), ALABS; Morton, Alfred C, =
JR (Al), ALABS; nsis@ietf.org; =
cdvorak@research.att.com<BR><B>Subject:</B> RE: [NSIS] Alternatives for =
supporting domain-local admission controloperation in =
NSLP/QSPEC<BR></FONT><BR></DIV>=0A=
<DIV>=0A=
<P><FONT size=3D2>Hi Roland, Georgios,<BR><BR>See below.<BR><BR>&gt; =
-----Original Message-----<BR>&gt; From: Roland Bless [<A =
href=3D"mailto:bless@tm.uka.de">mailto:bless@tm.uka.de</A>]<BR>&gt; =
Sent: Friday, November 24, 2006 11:30 AM<BR>&gt; To: Georgios =
Karagiannis<BR>&gt; Cc: Ash, Gerald R (Jerry), ALABS; =
john.loughney@nokia.com;<BR>&gt; Morton, Alfred C, JR (Al), ALABS; =
nsis@ietf.org<BR>&gt; Subject: Re: [NSIS] Alternatives for<BR>&gt; =
supportingdomain-localadmissioncontroloperation in =
NSLP/QSPEC<BR>&gt;<BR>&gt; Hi Georgios,<BR>&gt;<BR>&gt; see comments =
inline.<BR>&gt;<BR>&gt; Georgios Karagiannis wrote:<BR>&gt; &gt; What =
about my proposal, sent in a previous e-mail!<BR>&gt; &gt;<BR>&gt; &gt; =
"But the QSPEC template draft is not a generic QSPEC draft.<BR>&gt; &gt; =
The Token Bucket and the constraints can be seen as<BR>&gt; something =
that can be<BR>&gt; &gt; common and can be used for the interoperability =
and the rest of the<BR>&gt; &gt; encoded parameters can just be used as =
guidelines on how the QoS<BR>&gt; &gt; models can encode their =
parameters.<BR>&gt; &gt;<BR>&gt; &gt; Thus maybe Jerry could separate =
the QSPEC template draft into:<BR>&gt; &gt; A section/part that =
describes the Token Bucket + constraints.<BR>&gt; &gt; This part could =
then emphasize that the described parameters<BR>&gt; &gt; can be used =
for interoperability purposes.<BR>&gt; &gt;<BR>&gt; &gt; Another part of =
the QSPEC template draft will describe<BR>&gt; &gt; the other =
parameters.<BR>&gt; &gt; The goal of this part is to provide guidelines =
to QoS models<BR>&gt; &gt; on how to encode the different parameters, =
with the currently<BR>&gt; &gt; available parameters used as =
examples."<BR>&gt;<BR>&gt; As I said at the meeting:<BR>&gt; the current =
QSPEC template draft actually contains both aspects<BR>&gt; since we =
started with the template idea first and then added the<BR>&gt; =
interoperability aspects. An even cleaner separation would be<BR>&gt; a =
split into two separate documents:<BR>&gt; - one (template focus) =
explaining the general idea and layout<BR>&gt; of =
QSPECs,<BR>&gt;&nbsp;&nbsp; which could be used by other QoS models as =
guidance to create their<BR>&gt;&nbsp;&nbsp; own QSPECs, and<BR>&gt; - =
one (interoperability aspects) proposing a generic QSPEC =
as<BR>&gt;&nbsp;&nbsp; an example of a QSPEC instance<BR>&gt; However, =
it's also ok for me to describe both aspects in one document<BR>&gt; as =
it is probably now too late and too disturbing to make this<BR>&gt; =
separation.<BR><BR>OK, Georgios' proposal sounds like a good =
compromise.<BR><BR>Thanks,<BR>Regards,<BR>Jerry<BR><BR>&gt;<BR>&gt; =
Regards,<BR>&gt;&nbsp; Roland<BR></FONT></P></DIV></BODY></HTML>
------_=_NextPart_001_01C711A4.CB2308AE--


--===============1928790704==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1928790704==--




From nsis-bounces@ietf.org Sun Nov 26 16:54:56 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GoRx0-0000f8-V3; Sun, 26 Nov 2006 16:54:14 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GoRwz-0000dM-Hn
	for nsis@ietf.org; Sun, 26 Nov 2006 16:54:13 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GoRwn-0005NK-UU
	for nsis@ietf.org; Sun, 26 Nov 2006 16:54:13 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Sun, 26 Nov 2006 16:53:45 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006112616534408962 ; Sun, 26 Nov 2006 16:53:44 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <XMWVK25N>; Sun, 26 Nov 2006 16:53:45 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF3D@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: Georgios Karagiannis <karagian@cs.utwente.nl>,
	"Ash, Gerald R (Jerry), ALABS" <gash@att.com>, john.loughney@nokia.com
Subject: RE: [NSIS] Alternatives forsupportingdomain-localadmissioncontrol
	operation in NSLP/QSPEC
Date: Sun, 26 Nov 2006 16:50:37 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 26 Nov 2006 21:53:31.0024 (UTC)
	FILETIME=[4FD8E900:01C711A5]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 87a3f533bb300b99e2a18357f3c1563d
Cc: "Morton, Alfred C,JR \(Al\), ALABS" <acmorton@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Georgios:
 
Would you please provide any technical justifications why so called
"non-standard" QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 in this
"standard" track RFC in the same level of IETF "standard-based" RSVP,
IntServ, and DiffServ QOS Model? Is it not mixing between "apples" and
"oranges?" 
 
If you cannot justify the above, references of so called "non-standard" QOS
SLAs like Y.1541, DSTE, and 3GPP/Y.1541 as even examples for "encoding" of
QSPEC parameters must NOT be included.
 
Also please see inline [RRR].
 
Best regards,
Radhika

  _____  

From: nsis-bounces@ietf.org on behalf of Georgios Karagiannis
Sent: Fri 11/24/2006 4:31 AM
To: 'Ash, Gerald R (Jerry), ALABS'; john.loughney@nokia.com
Cc: 'Morton, Alfred C,JR (Al), ALABS'; nsis@ietf.org
Subject: RE: [NSIS] Alternatives
forsupportingdomain-localadmissioncontroloperation in NSLP/QSPEC



Hi Jerry

What about my proposal, sent in a previous e-mail!

"But the QSPEC template draft is not a generic QSPEC draft.
The Token Bucket and the constraints can be seen as something that can be
common and can be used for the interoperability and the rest of the
encoded parameters can just be used as guidelines on how the QoS
models can encode their parameters.

Thus maybe Jerry could separate the QSPEC template draft into:
A section/part that describes the Token Bucket + constraints.
This part could then emphasize that the described parameters
can be used for interoperability purposes.

Another part of the QSPEC template draft will describe
the other parameters.

The goal of this part is to provide guidelines to QoS models
on how to encode the different parameters, with the currently
available parameters used as examples."

Best Regards,
Georgios



> -----Original Message-----
> From: Ash, Gerald R (Jerry), ALABS [mailto:gash@att.com
<mailto:gash@att.com> ]
> Sent: donderdag 23 november 2006 15:00
> To: john.loughney@nokia.com
> Cc: Ash, Gerald R (Jerry), ALABS; Morton, Alfred C,JR (Al),
> ALABS; nsis@ietf.org
> Subject: RE: [NSIS] Alternatives for
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>
> Hi John,
>
> Happy Thanksgiving (for U.S. people anyway)!
>
> > -----Original Message-----
> > From: john.loughney@nokia.com [mailto:john.loughney@nokia.com
<mailto:john.loughney@nokia.com> ]
> > Sent: Thursday, November 23, 2006 2:55 AM
> > To: Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
> > Subject: RE: [NSIS] Alternatives for supporting
> > domain-localadmissioncontroloperation in NSLP/QSPEC
> >
> > Hi all,
> >
> > >It seems to me there is benefit in keeping the Y.1541 class in the
> > >QSPEC, because the Available class number could be
> modified along the
> > >reservation path. Even if the individual parameters change in the
> > >"QoS Available", it would be important recognize that a
> (for example,
> > >delay) threshold has been crossed and that a lesser class
> is now all
> > >that is available. This seems consistent with:
> > >
> > >>    QoS Available: QSPEC object containing parameters
> describing the
> > >>    available resources.  They are used to collect
> information along
> > >>    a reservation path.
> > >
> > >(I had wanted to make this comment during the meeting, but
> was busy
> > >taking notes for my friend, Hannes)
>
> > At this point, I don't think I have seen any strong consensus for
> > keeping the Y.1541 class in the QSPEC. I think it would be
> useful when
> > Y.1541 QoS models are involved in the e2e signaling, but
> its use as a
> > general QoS parameter is debatable.  Additionally, I don't
> think there
> > has been support for this from other QoS models, nor many people
> > speaking up in favor for keeping this as a general parameter.
>
> I'll of course abide by your call here, and also remove <PHB
> Class> and <DSTE Class Type> as per Roland's suggestion. 
> I'll note however
>
> 1. These parameters have been included in the QSPEC document
> from the beginning, that is, from the -00 version over 2 years ago.
> 2. The document passed WGLC last February with these
> parameters in; IMO that is consensus to leave in.
> 3. We just recently had the suggestion to remove these
> parameters, and maybe 2 votes of support; it doesn't seem
> like a strong consensus to remove.
> 4. Removing these parameters seems counter to support for
> DiffServ, as originally mandated for the QSPEC, especially
> removing <PHB Class>.
>
> Thanks,
> Regards,
> Jerry
>
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
<https://www1.ietf.org/mailman/listinfo/nsis> 
>



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis
<https://www1.ietf.org/mailman/listinfo/nsis> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Sun Nov 26 17:16:15 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GoSHl-0004Tg-TW; Sun, 26 Nov 2006 17:15:41 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GoSHl-0004Ta-EW
	for nsis@ietf.org; Sun, 26 Nov 2006 17:15:41 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GoSHk-0008LK-5B
	for nsis@ietf.org; Sun, 26 Nov 2006 17:15:41 -0500
Received: from 0015-its-ieg02.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Sun, 26 Nov 2006 17:15:28 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg02.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006112617152209072 ; Sun, 26 Nov 2006 17:15:27 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <XMWVK28K>; Sun, 26 Nov 2006 17:15:16 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426E1CF3E@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: john.loughney@nokia.com
Subject: RE: [NSIS] Alternatives for supportingdomain-localadmissioncontro
	loperation in NSLP/QSPEC
Date: Sun, 26 Nov 2006 17:15:09 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 26 Nov 2006 22:15:09.0907 (UTC)
	FILETIME=[560ABA30:01C711A8]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: a8a20a483a84f747e56475e290ee868e
Cc: "Morton, Alfred C,	JR \(Al\), ALABS" <acmorton@att.com>, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

John:
 
You are right to say, "I have seen any strong consensus for keeping the
Y.1541 class in the QSPEC." 
 
I am adding the following as well:

*	

	The so-called Y.1541, DSTE, and 3GPP/Y.1541 are "NON-Standard" QOS
models. They are rather value-based implementation-specific QOS SLAs and
have no properties that belong to a QOS model.
*	

	Per IETF, Y.1541, and 3GPP/Y.1541  are "Non-Standards."
*	

	Per IETF, only RSVP, IntServ, and DiffServ are "standard" QOS
Models.
*	As explained by many, the mixing of QOS models (RSVP, IntServ, and
DiffServ) and QOS SLAs (Y.1541, DSTE, and 3GPP/Y.1541) in the same draft
means mixing between "apples" and "oranges."
*	If PHBs are included for sake of DiffServ, the QSPECT RFC must NOT
contain anything specific that relates to "non-standard" Y.1541, DSTE, and
3GPP/Y.1541 QOS SLAs.
*	Moreover, a standard track RFC (QSPEC RFC) cannot contain
"non-standard" QOS SLAs.

Best regards,
Radhika
 
  _____  

From: nsis-bounces@ietf.org on behalf of Ash, Gerald R (Jerry), ALABS
Sent: Thu 11/23/2006 8:59 AM
To: john.loughney@nokia.com
Cc: Ash, Gerald R (Jerry), ALABS; Morton, Alfred C, JR (Al), ALABS;
nsis@ietf.org
Subject: RE: [NSIS] Alternatives for
supportingdomain-localadmissioncontroloperation in NSLP/QSPEC



Hi John,

Happy Thanksgiving (for U.S. people anyway)!

> -----Original Message-----
> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com
<mailto:john.loughney@nokia.com> ]
> Sent: Thursday, November 23, 2006 2:55 AM
> To: Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
> Subject: RE: [NSIS] Alternatives for supporting
> domain-localadmissioncontroloperation in NSLP/QSPEC
>
> Hi all,
>
> >It seems to me there is benefit in keeping the Y.1541 class in
> >the QSPEC, because the Available class number could be
> >modified along the reservation path. Even if the individual
> >parameters change in the "QoS Available", it would be
> >important recognize that a (for example, delay) threshold has
> >been crossed and that a lesser class is now all that is
> >available. This seems consistent with:
> >
> >>    QoS Available: QSPEC object containing parameters describing the
> >>    available resources.  They are used to collect information along
> >>    a reservation path.
> >
> >(I had wanted to make this comment during the meeting, but was
> >busy taking notes for my friend, Hannes)

> At this point, I don't think I have seen any strong consensus for
> keeping the Y.1541 class in the QSPEC. I think it would be useful when
> Y.1541 QoS models are involved in the e2e signaling, but its use as a
> general QoS parameter is debatable.  Additionally, I don't think there
> has been support for this from other QoS models, nor many people
> speaking up in favor for keeping this as a general parameter.

I'll of course abide by your call here, and also remove <PHB Class> and
<DSTE Class Type> as per Roland's suggestion.  I'll note however

1. These parameters have been included in the QSPEC document from the
beginning, that is, from the -00 version over 2 years ago.
2. The document passed WGLC last February with these parameters in; IMO
that is consensus to leave in.
3. We just recently had the suggestion to remove these parameters, and
maybe 2 votes of support; it doesn't seem like a strong consensus to
remove.
4. Removing these parameters seems counter to support for DiffServ, as
originally mandated for the QSPEC, especially removing <PHB Class>.

Thanks,
Regards,
Jerry

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis
<https://www1.ietf.org/mailman/listinfo/nsis> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 27 02:15:04 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GoahC-0006wk-12; Mon, 27 Nov 2006 02:14:30 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GoahA-0006wb-BK
	for nsis@ietf.org; Mon, 27 Nov 2006 02:14:28 -0500
Received: from mgw-ext13.nokia.com ([131.228.20.172])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Goah8-0001rk-Od
	for nsis@ietf.org; Mon, 27 Nov 2006 02:14:28 -0500
Received: from esebh107.NOE.Nokia.com (esebh107.ntc.nokia.com [172.21.143.143])
	by mgw-ext13.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kAR7E3sP031665; Mon, 27 Nov 2006 09:14:20 +0200
Received: from esebh101.NOE.Nokia.com ([172.21.138.177]) by
	esebh107.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 27 Nov 2006 09:13:57 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh101.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 27 Nov 2006 09:13:57 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for
	supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
Date: Mon, 27 Nov 2006 09:13:55 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCC016B184B@esebe199.NOE.Nokia.com>
In-Reply-To: <62D92A9A02BCC845B202323D49A48426E1CF3E@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for
	supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
Thread-Index: AccRvnPB2GC0XSeVRTOF+QaqhhnWLgANMsgQ
From: <john.loughney@nokia.com>
To: <RADHIKA.R.ROY@saic.com>
X-OriginalArrivalTime: 27 Nov 2006 07:13:57.0909 (UTC)
	FILETIME=[9B085C50:01C711F3]
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 93e7fb8fef2e780414389440f367c879
Cc: acmorton@att.com, nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi Radhika,

I think we know your position on this issue.  I've filled you position
on this issue, so you do not have repeat your position on this.

thanks,
John

>-----Original Message-----
>From: ext Roy, Radhika R. [mailto:RADHIKA.R.ROY@saic.com]=20
>Sent: 27 November, 2006 00:15
>To: Loughney John (Nokia-NRC/Helsinki)
>Cc: Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
>Subject: RE: [NSIS] Alternatives for=20
>supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>
>John:
>=20
>You are right to say, "I have seen any strong consensus for keeping the
>Y.1541 class in the QSPEC."=20
>=20
>I am adding the following as well:
>
>*=09
>
>	The so-called Y.1541, DSTE, and 3GPP/Y.1541 are=20
>"NON-Standard" QOS models. They are rather value-based=20
>implementation-specific QOS SLAs and have no properties that=20
>belong to a QOS model.
>*=09
>
>	Per IETF, Y.1541, and 3GPP/Y.1541  are "Non-Standards."
>*=09
>
>	Per IETF, only RSVP, IntServ, and DiffServ are=20
>"standard" QOS Models.
>*	As explained by many, the mixing of QOS models (RSVP,=20
>IntServ, and
>DiffServ) and QOS SLAs (Y.1541, DSTE, and 3GPP/Y.1541) in the=20
>same draft means mixing between "apples" and "oranges."
>*	If PHBs are included for sake of DiffServ, the QSPECT=20
>RFC must NOT
>contain anything specific that relates to "non-standard"=20
>Y.1541, DSTE, and
>3GPP/Y.1541 QOS SLAs.
>*	Moreover, a standard track RFC (QSPEC RFC) cannot contain
>"non-standard" QOS SLAs.
>
>Best regards,
>Radhika
>=20
>  _____ =20
>
>From: nsis-bounces@ietf.org on behalf of Ash, Gerald R (Jerry), ALABS
>Sent: Thu 11/23/2006 8:59 AM
>To: john.loughney@nokia.com
>Cc: Ash, Gerald R (Jerry), ALABS; Morton, Alfred C, JR (Al),=20
>ALABS; nsis@ietf.org
>Subject: RE: [NSIS] Alternatives for
>supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>
>
>
>Hi John,
>
>Happy Thanksgiving (for U.S. people anyway)!
>
>> -----Original Message-----
>> From: john.loughney@nokia.com [mailto:john.loughney@nokia.com
><mailto:john.loughney@nokia.com> ]
>> Sent: Thursday, November 23, 2006 2:55 AM
>> To: Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
>> Subject: RE: [NSIS] Alternatives for supporting=20
>> domain-localadmissioncontroloperation in NSLP/QSPEC
>>
>> Hi all,
>>
>> >It seems to me there is benefit in keeping the Y.1541 class in the=20
>> >QSPEC, because the Available class number could be modified=20
>along the=20
>> >reservation path. Even if the individual parameters change in the=20
>> >"QoS Available", it would be important recognize that a=20
>(for example,=20
>> >delay) threshold has been crossed and that a lesser class=20
>is now all=20
>> >that is available. This seems consistent with:
>> >
>> >>    QoS Available: QSPEC object containing parameters=20
>describing the
>> >>    available resources.  They are used to collect=20
>information along
>> >>    a reservation path.
>> >
>> >(I had wanted to make this comment during the meeting, but was busy=20
>> >taking notes for my friend, Hannes)
>
>> At this point, I don't think I have seen any strong consensus for=20
>> keeping the Y.1541 class in the QSPEC. I think it would be=20
>useful when
>> Y.1541 QoS models are involved in the e2e signaling, but its=20
>use as a=20
>> general QoS parameter is debatable.  Additionally, I don't=20
>think there=20
>> has been support for this from other QoS models, nor many people=20
>> speaking up in favor for keeping this as a general parameter.
>
>I'll of course abide by your call here, and also remove <PHB=20
>Class> and <DSTE Class Type> as per Roland's suggestion.  I'll=20
>note however
>
>1. These parameters have been included in the QSPEC document=20
>from the beginning, that is, from the -00 version over 2 years ago.
>2. The document passed WGLC last February with these=20
>parameters in; IMO that is consensus to leave in.
>3. We just recently had the suggestion to remove these=20
>parameters, and maybe 2 votes of support; it doesn't seem like=20
>a strong consensus to remove.
>4. Removing these parameters seems counter to support for=20
>DiffServ, as originally mandated for the QSPEC, especially=20
>removing <PHB Class>.
>
>Thanks,
>Regards,
>Jerry
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
><https://www1.ietf.org/mailman/listinfo/nsis>=20
>
>
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Mon Nov 27 03:44:23 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Goc5T-0000mo-Hb; Mon, 27 Nov 2006 03:43:39 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Goc5S-0000mj-84
	for nsis@ietf.org; Mon, 27 Nov 2006 03:43:38 -0500
Received: from mgw-ext13.nokia.com ([131.228.20.172])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Goc5R-0000wk-BK
	for nsis@ietf.org; Mon, 27 Nov 2006 03:43:38 -0500
Received: from esebh107.NOE.Nokia.com (esebh107.ntc.nokia.com [172.21.143.143])
	by mgw-ext13.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kAR83Hee005213; Mon, 27 Nov 2006 10:03:26 +0200
Received: from esebh103.NOE.Nokia.com ([172.21.143.33]) by
	esebh107.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 27 Nov 2006 10:03:20 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh103.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Mon, 27 Nov 2006 10:03:20 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Date: Mon, 27 Nov 2006 10:03:19 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCC016B1967@esebe199.NOE.Nokia.com>
In-Reply-To: <62D92A9A02BCC845B202323D49A48426E1CF3C@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Thread-Index: AccP5ceRsNCbNznGQaCPBDUG9f2U+wAqQFkgAESpuhAAFiDIYA==
From: <john.loughney@nokia.com>
To: <RADHIKA.R.ROY@saic.com>
X-OriginalArrivalTime: 27 Nov 2006 08:03:20.0041 (UTC)
	FILETIME=[8099D590:01C711FA]
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 32a65c0bf5eb4ec26489239c7cdd0636
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1089646234=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1089646234==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C711FA.80509758"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C711FA.80509758
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

The QSPEC document is an Informational RFC, so your comment about this
is really not very relevant to the discussion. =20
=20
thanks,
John


________________________________

	From: ext Roy, Radhika R. [mailto:RADHIKA.R.ROY@saic.com]=20
	Sent: 26 November, 2006 23:50
	To: Roland Bless; Georgios Karagiannis
	Cc: Loughney John (Nokia-NRC/Helsinki); Morton, Alfred C, JR
(Al), ALABS; nsis@ietf.org; cdvorak@research.att.com
	Subject: RE: [NSIS] Alternatives for supporting domain-local
admission controloperation in NSLP/QSPEC
=09
=09
	Ronald and Georgios:
	=20
	Why is it too late to clean out the document as Ronald has
explained?
	=20
	NSIS WG in particular and IETF in general have waited almost two
years to root out these "technically" inconsistencies. It is better NOT
to publish a standard track RFC like this.
	=20
	If needed, NSIS WG Chair needs to appoint a "single" EDITOR who
can clean out this document.
	=20
	Why does the NSIS WG Chair need to appoint 3 (Three) editors?
Is it not the problems of appointing 3 editors for the same simple RFC?
Is it not because of 3 editors who have put those things defying a
"simple" common sense that to keep the document is the most important
thing from the IETF point of view?
	=20
	Is it not that people still trying to force so-called QOS SLAs
like Y.1541, DSTE, and 3GPP/Y.1541 in this draft as "standard" QOS
models?
	=20
	A question to Georgios: Would you please provide any technical
justifications why so called "non-standard" QOS SLAs like Y.1541, DSTE,
and 3GPP/Y.1541 in this "standard" track RFC in the same level of IETF
"standard-based" RSVP, IntServ, and DiffServ QOS Model? Is it not mixing
between "apples" and "oranges?"
	=20
	If Georgios cannot justify the above, references of so called
"non-standard" QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 as even
examples for encoding of QSPEC parameters must NOT be included.
	=20
	If it is done despite objections, we would think that this QSPEC
RFC must NOT be a "standard" track RFC because a standard track RFC must
NOT contain "non-standard" QOS models. This will also be brought to the
notice of the NSIS WG Chairs and IESG.
	=20
	Best regards,
	Radhika

________________________________

	From: nsis-bounces@ietf.org on behalf of Ash, Gerald R (Jerry),
ALABS
	Sent: Sat 11/25/2006 7:44 AM
	To: Roland Bless; Georgios Karagiannis
	Cc: john.loughney@nokia.com; Ash, Gerald R (Jerry), ALABS;
Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org;
cdvorak@research.att.com
	Subject: RE: [NSIS] Alternatives for supporting domain-local
admission controloperation in NSLP/QSPEC
=09
=09

	Hi Roland, Georgios,
=09
	See below.
=09
	> -----Original Message-----
	> From: Roland Bless [mailto:bless@tm.uka.de]
	> Sent: Friday, November 24, 2006 11:30 AM
	> To: Georgios Karagiannis
	> Cc: Ash, Gerald R (Jerry), ALABS; john.loughney@nokia.com;
	> Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
	> Subject: Re: [NSIS] Alternatives for
	> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
	>
	> Hi Georgios,
	>
	> see comments inline.
	>
	> Georgios Karagiannis wrote:
	> > What about my proposal, sent in a previous e-mail!
	> >
	> > "But the QSPEC template draft is not a generic QSPEC draft.
	> > The Token Bucket and the constraints can be seen as
	> something that can be
	> > common and can be used for the interoperability and the rest
of the
	> > encoded parameters can just be used as guidelines on how the
QoS
	> > models can encode their parameters.
	> >
	> > Thus maybe Jerry could separate the QSPEC template draft
into:
	> > A section/part that describes the Token Bucket +
constraints.
	> > This part could then emphasize that the described parameters
	> > can be used for interoperability purposes.
	> >
	> > Another part of the QSPEC template draft will describe
	> > the other parameters.
	> > The goal of this part is to provide guidelines to QoS models
	> > on how to encode the different parameters, with the
currently
	> > available parameters used as examples."
	>
	> As I said at the meeting:
	> the current QSPEC template draft actually contains both
aspects
	> since we started with the template idea first and then added
the
	> interoperability aspects. An even cleaner separation would be
	> a split into two separate documents:
	> - one (template focus) explaining the general idea and layout
	> of QSPECs,
	>   which could be used by other QoS models as guidance to
create their
	>   own QSPECs, and
	> - one (interoperability aspects) proposing a generic QSPEC as
	>   an example of a QSPEC instance
	> However, it's also ok for me to describe both aspects in one
document
	> as it is probably now too late and too disturbing to make this
	> separation.
=09
	OK, Georgios' proposal sounds like a good compromise.
=09
	Thanks,
	Regards,
	Jerry
=09
	>
	> Regards,
	>  Roland
=09


------_=_NextPart_001_01C711FA.80509758
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML dir=3Dltr><HEAD><TITLE>RE: [NSIS] Alternatives for supporting =
domain-local admission controloperation in NSLP/QSPEC</TITLE>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<META content=3D"MSHTML 6.00.2900.2995" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D485225907-27112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>The QSPEC document is an Informational RFC, so =
your comment=20
about this is really not very relevant to the discussion.&nbsp;=20
</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D485225907-27112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D485225907-27112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>thanks,<BR>John</FONT></SPAN></DIV><BR>
<BLOCKQUOTE dir=3Dltr=20
style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT: #0000ff 2px =
solid; MARGIN-RIGHT: 0px">
  <DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> ext Roy, Radhika R.=20
  [mailto:RADHIKA.R.ROY@saic.com] <BR><B>Sent:</B> 26 November, 2006=20
  23:50<BR><B>To:</B> Roland Bless; Georgios Karagiannis<BR><B>Cc:</B> =
Loughney=20
  John (Nokia-NRC/Helsinki); Morton, Alfred C, JR (Al), ALABS; =
nsis@ietf.org;=20
  cdvorak@research.att.com<BR><B>Subject:</B> RE: [NSIS] Alternatives =
for=20
  supporting domain-local admission controloperation in=20
  NSLP/QSPEC<BR></FONT><BR></DIV>
  <DIV></DIV>
  <DIV id=3DidOWAReplyText23077 dir=3Dltr>
  <DIV dir=3Dltr><FONT face=3DArial color=3D#000000 size=3D2>Ronald and=20
  Georgios:</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>Why is it too late to clean =
out the=20
  document as Ronald has explained?</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>NSIS WG in particular and =
IETF in general=20
  have waited almost two years to root out these "technically" =
inconsistencies.=20
  It is better NOT to publish a standard track RFC like =
this.</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>If needed, NSIS WG Chair =
needs to appoint=20
  a "single" EDITOR who can clean out this document.</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>Why does the NSIS WG Chair =
need to=20
  appoint 3 (Three) editors?&nbsp;&nbsp;Is it not the problems of =
appointing 3=20
  editors for the same simple RFC? Is it not because of 3 editors who =
have put=20
  those things defying a "simple" common sense that to keep the document =
is the=20
  most important thing from the IETF point of view?</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>Is it not that people still =
trying to=20
  force so-called QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 in this =
draft as=20
  "standard" QOS models?</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>A question to Georgios: =
Would you please=20
  provide&nbsp;any technical justifications why so called "non-standard" =
QOS=20
  SLAs like Y.1541, DSTE, and 3GPP/Y.1541 in this "standard" track RFC =
in the=20
  same level of IETF "standard-based" RSVP, IntServ, and DiffServ QOS =
Model? Is=20
  it not mixing between "apples" and "oranges?"</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>If Georgios cannot justify =
the above,=20
  references of so called "non-standard" QOS SLAs like Y.1541, DSTE, and =

  3GPP/Y.1541 as even examples for encoding of QSPEC parameters must NOT =
be=20
  included.</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>If it is done despite =
objections, we=20
  would think that this QSPEC RFC must NOT be a "standard" track RFC =
because a=20
  standard track RFC must NOT contain "non-standard" QOS models. This =
will also=20
  be&nbsp;brought to the notice of the NSIS WG Chairs and =
IESG.</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>Best regards,</FONT></DIV>
  <DIV dir=3Dltr><FONT face=3DArial size=3D2>Radhika</FONT></DIV></DIV>
  <DIV dir=3Dltr><BR>
  <HR tabIndex=3D-1>
  <FONT face=3DTahoma size=3D2><B>From:</B> nsis-bounces@ietf.org on =
behalf of Ash,=20
  Gerald R (Jerry), ALABS<BR><B>Sent:</B> Sat 11/25/2006 7:44 =
AM<BR><B>To:</B>=20
  Roland Bless; Georgios Karagiannis<BR><B>Cc:</B> =
john.loughney@nokia.com; Ash,=20
  Gerald R (Jerry), ALABS; Morton, Alfred C, JR (Al), ALABS; =
nsis@ietf.org;=20
  cdvorak@research.att.com<BR><B>Subject:</B> RE: [NSIS] Alternatives =
for=20
  supporting domain-local admission controloperation in=20
  NSLP/QSPEC<BR></FONT><BR></DIV>
  <DIV>
  <P><FONT size=3D2>Hi Roland, Georgios,<BR><BR>See below.<BR><BR>&gt;=20
  -----Original Message-----<BR>&gt; From: Roland Bless [<A=20
  href=3D"mailto:bless@tm.uka.de">mailto:bless@tm.uka.de</A>]<BR>&gt; =
Sent:=20
  Friday, November 24, 2006 11:30 AM<BR>&gt; To: Georgios =
Karagiannis<BR>&gt;=20
  Cc: Ash, Gerald R (Jerry), ALABS; john.loughney@nokia.com;<BR>&gt; =
Morton,=20
  Alfred C, JR (Al), ALABS; nsis@ietf.org<BR>&gt; Subject: Re: [NSIS]=20
  Alternatives for<BR>&gt; =
supportingdomain-localadmissioncontroloperation in=20
  NSLP/QSPEC<BR>&gt;<BR>&gt; Hi Georgios,<BR>&gt;<BR>&gt; see comments=20
  inline.<BR>&gt;<BR>&gt; Georgios Karagiannis wrote:<BR>&gt; &gt; What =
about my=20
  proposal, sent in a previous e-mail!<BR>&gt; &gt;<BR>&gt; &gt; "But =
the QSPEC=20
  template draft is not a generic QSPEC draft.<BR>&gt; &gt; The Token =
Bucket and=20
  the constraints can be seen as<BR>&gt; something that can be<BR>&gt; =
&gt;=20
  common and can be used for the interoperability and the rest of =
the<BR>&gt;=20
  &gt; encoded parameters can just be used as guidelines on how the =
QoS<BR>&gt;=20
  &gt; models can encode their parameters.<BR>&gt; &gt;<BR>&gt; &gt; =
Thus maybe=20
  Jerry could separate the QSPEC template draft into:<BR>&gt; &gt; A=20
  section/part that describes the Token Bucket + constraints.<BR>&gt; =
&gt; This=20
  part could then emphasize that the described parameters<BR>&gt; &gt; =
can be=20
  used for interoperability purposes.<BR>&gt; &gt;<BR>&gt; &gt; Another =
part of=20
  the QSPEC template draft will describe<BR>&gt; &gt; the other=20
  parameters.<BR>&gt; &gt; The goal of this part is to provide =
guidelines to QoS=20
  models<BR>&gt; &gt; on how to encode the different parameters, with =
the=20
  currently<BR>&gt; &gt; available parameters used as =
examples."<BR>&gt;<BR>&gt;=20
  As I said at the meeting:<BR>&gt; the current QSPEC template draft =
actually=20
  contains both aspects<BR>&gt; since we started with the template idea =
first=20
  and then added the<BR>&gt; interoperability aspects. An even cleaner=20
  separation would be<BR>&gt; a split into two separate =
documents:<BR>&gt; - one=20
  (template focus) explaining the general idea and layout<BR>&gt; of=20
  QSPECs,<BR>&gt;&nbsp;&nbsp; which could be used by other QoS models as =

  guidance to create their<BR>&gt;&nbsp;&nbsp; own QSPECs, and<BR>&gt; - =
one=20
  (interoperability aspects) proposing a generic QSPEC =
as<BR>&gt;&nbsp;&nbsp; an=20
  example of a QSPEC instance<BR>&gt; However, it's also ok for me to =
describe=20
  both aspects in one document<BR>&gt; as it is probably now too late =
and too=20
  disturbing to make this<BR>&gt; separation.<BR><BR>OK, Georgios' =
proposal=20
  sounds like a good=20
  compromise.<BR><BR>Thanks,<BR>Regards,<BR>Jerry<BR><BR>&gt;<BR>&gt;=20
  Regards,<BR>&gt;&nbsp; =
Roland<BR></FONT></P></DIV></BLOCKQUOTE></BODY></HTML>

------_=_NextPart_001_01C711FA.80509758--


--===============1089646234==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1089646234==--




From nsis-bounces@ietf.org Mon Nov 27 10:24:52 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GoiLA-00059K-R4; Mon, 27 Nov 2006 10:24:16 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GoiL9-00058k-Qa
	for nsis@ietf.org; Mon, 27 Nov 2006 10:24:15 -0500
Received: from mclmx.mail.saic.com ([149.8.64.10])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GoiL7-0000Yr-Er
	for nsis@ietf.org; Mon, 27 Nov 2006 10:24:15 -0500
Received: from 0015-its-ieg01.mail.saic.com ([149.8.64.21] [149.8.64.21]) by
	mclmx.mail.saic.com; Mon, 27 Nov 2006 10:24:04 -0500
Received: from mcl-its-exbh01.mail.saic.com ([149.8.64.11])
	by 0015-its-ieg01.mail.saic.com (SMSSMTP 4.0.5.66) with SMTP id
	M2006112710240402813 ; Mon, 27 Nov 2006 10:24:04 -0500
Received: by mcl-its-exbh01.mail.saic.com with Internet Mail Service
	(5.5.2657.72) id <XWKP5QD9>; Mon, 27 Nov 2006 10:24:04 -0500
Message-Id: <62D92A9A02BCC845B202323D49A48426230EF6@0591-ITS-EXMP02.us.saic.com>
From: "Roy, Radhika R." <RADHIKA.R.ROY@saic.com>
To: john.loughney@nokia.com
Subject: RE: [NSIS] Alternatives for supporting domain-local admission con
	troloperation in NSLP/QSPEC
Date: Mon, 27 Nov 2006 10:23:59 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2657.72)
content-class: urn:content-classes:message
x-mimeole: Produced By Microsoft Exchange V6.5
x-originalarrivaltime: 27 Nov 2006 15:24:00.0122 (UTC)
	FILETIME=[101E05A0:01C71238]
Content-Type: text/plain;
	charset="iso-8859-1"
X-Spam-Score: 0.0 (/)
X-Scan-Signature: ff0adf256e4dd459cc25215cfa732ac1
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

John:

 

Thanks with a great relief that it is good that this is an Informational
RFC. If it is so, why does this not mentioned in the draft as it was the
case in the beginning? Did I miss it?

 

The next questions that are relevant for this "Informational" RFC are as
follows as this draft has been taken as the WG item:

*	Is this informational RFC supposed to create "Interworking" problems
between IETF "standard-based" QOS models (RSVP, IntServ, and DiffServ) and
so-called "non-standards" QOS models (Y.1541, DSTE, and 3GPP/Y.1541) through
mixing between "apples" and "oranges" in the same draft?
*	Why QOS models and QOS SLAs need to be treated in the same level
just because of the "informational" RFC?
*	If there are separate individual QOSM drafts, why this QSPEC draft
needs to deal with all those things of all QOS models and QOS SLAs that are
"incomplete" and create problems of technical problems including
"non-interoperability?"
*	Why is it too late to clean the draft as Ronald mentioned just
because it is an "Informational" RFC?
*	As you have already pointed out to all NSIS WG members that there is
no WG consensus to include Y.1451 (i.e., QOS SLAs) in this draft despite it
is an "Informational" RFC?

 

Best regards,

Radhika

 

PS: 

 

1.	If QOS SLAs are removed from this draft, why can this draft be a
"standard-track" RFC as NSIS QOS signaling messages will almost be blind or
unworkable to carry the requisite information of the QOS parameters?
2.	Even the "guidelines" for "encoding" of QSPEC by different QOS
Models and QOS SLAs in this QSPEC draft will be "incomplete" and will create
more problems than solving this because the individual QOSM drafts are
better place to do so as they will be complete for each QOS model and each
QOS SLA. However, the "generic" encoding "guidelines" without mentioning any
particular QOS models or QOS SLAs will be OK in this draft.

  _____  

From: john.loughney@nokia.com [mailto:john.loughney@nokia.com] 
Sent: Monday, November 27, 2006 3:03 AM
To: RADHIKA.R.ROY@saic.com
Cc: nsis@ietf.org
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

 

The QSPEC document is an Informational RFC, so your comment about this is
really not very relevant to the discussion.  

 

thanks,
John

 


  _____  


From: ext Roy, Radhika R. [mailto:RADHIKA.R.ROY@saic.com] 
Sent: 26 November, 2006 23:50
To: Roland Bless; Georgios Karagiannis
Cc: Loughney John (Nokia-NRC/Helsinki); Morton, Alfred C, JR (Al), ALABS;
nsis@ietf.org; cdvorak@research.att.com
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

Ronald and Georgios:

 

Why is it too late to clean out the document as Ronald has explained?

 

NSIS WG in particular and IETF in general have waited almost two years to
root out these "technically" inconsistencies. It is better NOT to publish a
standard track RFC like this.

 

If needed, NSIS WG Chair needs to appoint a "single" EDITOR who can clean
out this document.

 

Why does the NSIS WG Chair need to appoint 3 (Three) editors?  Is it not the
problems of appointing 3 editors for the same simple RFC? Is it not because
of 3 editors who have put those things defying a "simple" common sense that
to keep the document is the most important thing from the IETF point of
view?

 

Is it not that people still trying to force so-called QOS SLAs like Y.1541,
DSTE, and 3GPP/Y.1541 in this draft as "standard" QOS models?

 

A question to Georgios: Would you please provide any technical
justifications why so called "non-standard" QOS SLAs like Y.1541, DSTE, and
3GPP/Y.1541 in this "standard" track RFC in the same level of IETF
"standard-based" RSVP, IntServ, and DiffServ QOS Model? Is it not mixing
between "apples" and "oranges?"

 

If Georgios cannot justify the above, references of so called "non-standard"
QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 as even examples for encoding of
QSPEC parameters must NOT be included.

 

If it is done despite objections, we would think that this QSPEC RFC must
NOT be a "standard" track RFC because a standard track RFC must NOT contain
"non-standard" QOS models. This will also be brought to the notice of the
NSIS WG Chairs and IESG.

 

Best regards,

Radhika

 


  _____  


From: nsis-bounces@ietf.org on behalf of Ash, Gerald R (Jerry), ALABS
Sent: Sat 11/25/2006 7:44 AM
To: Roland Bless; Georgios Karagiannis
Cc: john.loughney@nokia.com; Ash, Gerald R (Jerry), ALABS; Morton, Alfred C,
JR (Al), ALABS; nsis@ietf.org; cdvorak@research.att.com
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
controloperation in NSLP/QSPEC

Hi Roland, Georgios,

See below.

> -----Original Message-----
> From: Roland Bless [mailto:bless@tm.uka.de <mailto:bless@tm.uka.de> ]
> Sent: Friday, November 24, 2006 11:30 AM
> To: Georgios Karagiannis
> Cc: Ash, Gerald R (Jerry), ALABS; john.loughney@nokia.com;
> Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org
> Subject: Re: [NSIS] Alternatives for
> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>
> Hi Georgios,
>
> see comments inline.
>
> Georgios Karagiannis wrote:
> > What about my proposal, sent in a previous e-mail!
> >
> > "But the QSPEC template draft is not a generic QSPEC draft.
> > The Token Bucket and the constraints can be seen as
> something that can be
> > common and can be used for the interoperability and the rest of the
> > encoded parameters can just be used as guidelines on how the QoS
> > models can encode their parameters.
> >
> > Thus maybe Jerry could separate the QSPEC template draft into:
> > A section/part that describes the Token Bucket + constraints.
> > This part could then emphasize that the described parameters
> > can be used for interoperability purposes.
> >
> > Another part of the QSPEC template draft will describe
> > the other parameters.
> > The goal of this part is to provide guidelines to QoS models
> > on how to encode the different parameters, with the currently
> > available parameters used as examples."
>
> As I said at the meeting:
> the current QSPEC template draft actually contains both aspects
> since we started with the template idea first and then added the
> interoperability aspects. An even cleaner separation would be
> a split into two separate documents:
> - one (template focus) explaining the general idea and layout
> of QSPECs,
>   which could be used by other QoS models as guidance to create their
>   own QSPECs, and
> - one (interoperability aspects) proposing a generic QSPEC as
>   an example of a QSPEC instance
> However, it's also ok for me to describe both aspects in one document
> as it is probably now too late and too disturbing to make this
> separation.

OK, Georgios' proposal sounds like a good compromise.

Thanks,
Regards,
Jerry

>
> Regards,
>  Roland


_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 30 02:17:16 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gpg9h-0004Xi-QS; Thu, 30 Nov 2006 02:16:25 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gpg9g-0004XP-8e
	for nsis@ietf.org; Thu, 30 Nov 2006 02:16:24 -0500
Received: from mgw-ext03.nokia.com ([131.228.20.95])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gpg9f-0007rm-CA
	for nsis@ietf.org; Thu, 30 Nov 2006 02:16:24 -0500
Received: from esebh106.NOE.Nokia.com (esebh106.ntc.nokia.com [172.21.138.213])
	by mgw-ext03.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kAU7GHPC029305; Thu, 30 Nov 2006 09:16:21 +0200
Received: from esebh101.NOE.Nokia.com ([172.21.138.177]) by
	esebh106.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:15:28 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh101.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:15:28 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: RE: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Date: Thu, 30 Nov 2006 09:15:27 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCC01727115@esebe199.NOE.Nokia.com>
In-Reply-To: <62D92A9A02BCC845B202323D49A48426230EF6@0591-ITS-EXMP02.us.saic.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Alternatives for supporting domain-local admission
	controloperation in NSLP/QSPEC
Thread-Index: AccS2lGE8ydKvS2OToGbFEphF8kfiABdPf4Q
From: <john.loughney@nokia.com>
To: <RADHIKA.R.ROY@saic.com>
X-OriginalArrivalTime: 30 Nov 2006 07:15:28.0783 (UTC)
	FILETIME=[506FE1F0:01C7144F]
X-eXpurgate-Category: 1/0
X-eXpurgate-ID: 149371::061130091621-70531BB0-2B825726/0-0/0-0
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 3f3e54d3c03ed638c06aa9fa6861237e
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

If you read the charter, it says:

Jun 2006	  	Submit 'NSIS QoS Specification Template' to IESG
for publication as an Informational RFC

John=20

>-----Original Message-----
>From: ext Roy, Radhika R. [mailto:RADHIKA.R.ROY@saic.com]=20
>Sent: 27 November, 2006 17:24
>To: Loughney John (Nokia-NRC/Helsinki)
>Cc: nsis@ietf.org
>Subject: RE: [NSIS] Alternatives for supporting domain-local=20
>admission controloperation in NSLP/QSPEC
>
>John:
>
>=20
>
>Thanks with a great relief that it is good that this is an=20
>Informational RFC. If it is so, why does this not mentioned in=20
>the draft as it was the case in the beginning? Did I miss it?
>
>=20
>
>The next questions that are relevant for this "Informational"=20
>RFC are as follows as this draft has been taken as the WG item:
>
>*	Is this informational RFC supposed to create=20
>"Interworking" problems
>between IETF "standard-based" QOS models (RSVP, IntServ, and=20
>DiffServ) and so-called "non-standards" QOS models (Y.1541,=20
>DSTE, and 3GPP/Y.1541) through mixing between "apples" and=20
>"oranges" in the same draft?
>*	Why QOS models and QOS SLAs need to be treated in the same level
>just because of the "informational" RFC?
>*	If there are separate individual QOSM drafts, why this=20
>QSPEC draft
>needs to deal with all those things of all QOS models and QOS=20
>SLAs that are "incomplete" and create problems of technical=20
>problems including "non-interoperability?"
>*	Why is it too late to clean the draft as Ronald mentioned just
>because it is an "Informational" RFC?
>*	As you have already pointed out to all NSIS WG members=20
>that there is
>no WG consensus to include Y.1451 (i.e., QOS SLAs) in this=20
>draft despite it is an "Informational" RFC?
>
>=20
>
>Best regards,
>
>Radhika
>
>=20
>
>PS:=20
>
>=20
>
>1.	If QOS SLAs are removed from this draft, why can this draft be a
>"standard-track" RFC as NSIS QOS signaling messages will=20
>almost be blind or unworkable to carry the requisite=20
>information of the QOS parameters?
>2.	Even the "guidelines" for "encoding" of QSPEC by different QOS
>Models and QOS SLAs in this QSPEC draft will be "incomplete"=20
>and will create more problems than solving this because the=20
>individual QOSM drafts are better place to do so as they will=20
>be complete for each QOS model and each QOS SLA. However, the=20
>"generic" encoding "guidelines" without mentioning any=20
>particular QOS models or QOS SLAs will be OK in this draft.
>
>  _____ =20
>
>From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]
>Sent: Monday, November 27, 2006 3:03 AM
>To: RADHIKA.R.ROY@saic.com
>Cc: nsis@ietf.org
>Subject: RE: [NSIS] Alternatives for supporting domain-local=20
>admission controloperation in NSLP/QSPEC
>
>=20
>
>The QSPEC document is an Informational RFC, so your comment=20
>about this is really not very relevant to the discussion. =20
>
>=20
>
>thanks,
>John
>
>=20
>
>
>  _____ =20
>
>
>From: ext Roy, Radhika R. [mailto:RADHIKA.R.ROY@saic.com]
>Sent: 26 November, 2006 23:50
>To: Roland Bless; Georgios Karagiannis
>Cc: Loughney John (Nokia-NRC/Helsinki); Morton, Alfred C, JR=20
>(Al), ALABS; nsis@ietf.org; cdvorak@research.att.com
>Subject: RE: [NSIS] Alternatives for supporting domain-local=20
>admission controloperation in NSLP/QSPEC
>
>Ronald and Georgios:
>
>=20
>
>Why is it too late to clean out the document as Ronald has explained?
>
>=20
>
>NSIS WG in particular and IETF in general have waited almost=20
>two years to root out these "technically" inconsistencies. It=20
>is better NOT to publish a standard track RFC like this.
>
>=20
>
>If needed, NSIS WG Chair needs to appoint a "single" EDITOR=20
>who can clean out this document.
>
>=20
>
>Why does the NSIS WG Chair need to appoint 3 (Three) editors? =20
>Is it not the problems of appointing 3 editors for the same=20
>simple RFC? Is it not because of 3 editors who have put those=20
>things defying a "simple" common sense that to keep the=20
>document is the most important thing from the IETF point of view?
>
>=20
>
>Is it not that people still trying to force so-called QOS SLAs=20
>like Y.1541, DSTE, and 3GPP/Y.1541 in this draft as "standard"=20
>QOS models?
>
>=20
>
>A question to Georgios: Would you please provide any technical=20
>justifications why so called "non-standard" QOS SLAs like=20
>Y.1541, DSTE, and
>3GPP/Y.1541 in this "standard" track RFC in the same level of=20
>IETF "standard-based" RSVP, IntServ, and DiffServ QOS Model?=20
>Is it not mixing between "apples" and "oranges?"
>
>=20
>
>If Georgios cannot justify the above, references of so called=20
>"non-standard"
>QOS SLAs like Y.1541, DSTE, and 3GPP/Y.1541 as even examples=20
>for encoding of QSPEC parameters must NOT be included.
>
>=20
>
>If it is done despite objections, we would think that this=20
>QSPEC RFC must NOT be a "standard" track RFC because a=20
>standard track RFC must NOT contain "non-standard" QOS models.=20
>This will also be brought to the notice of the NSIS WG Chairs and IESG.
>
>=20
>
>Best regards,
>
>Radhika
>
>=20
>
>
>  _____ =20
>
>
>From: nsis-bounces@ietf.org on behalf of Ash, Gerald R (Jerry), ALABS
>Sent: Sat 11/25/2006 7:44 AM
>To: Roland Bless; Georgios Karagiannis
>Cc: john.loughney@nokia.com; Ash, Gerald R (Jerry), ALABS;=20
>Morton, Alfred C, JR (Al), ALABS; nsis@ietf.org;=20
>cdvorak@research.att.com
>Subject: RE: [NSIS] Alternatives for supporting domain-local=20
>admission controloperation in NSLP/QSPEC
>
>Hi Roland, Georgios,
>
>See below.
>
>> -----Original Message-----
>> From: Roland Bless [mailto:bless@tm.uka.de <mailto:bless@tm.uka.de> ]
>> Sent: Friday, November 24, 2006 11:30 AM
>> To: Georgios Karagiannis
>> Cc: Ash, Gerald R (Jerry), ALABS; john.loughney@nokia.com; Morton,=20
>> Alfred C, JR (Al), ALABS; nsis@ietf.org
>> Subject: Re: [NSIS] Alternatives for
>> supportingdomain-localadmissioncontroloperation in NSLP/QSPEC
>>
>> Hi Georgios,
>>
>> see comments inline.
>>
>> Georgios Karagiannis wrote:
>> > What about my proposal, sent in a previous e-mail!
>> >
>> > "But the QSPEC template draft is not a generic QSPEC draft.
>> > The Token Bucket and the constraints can be seen as
>> something that can be
>> > common and can be used for the interoperability and the=20
>rest of the=20
>> > encoded parameters can just be used as guidelines on how the QoS=20
>> > models can encode their parameters.
>> >
>> > Thus maybe Jerry could separate the QSPEC template draft into:
>> > A section/part that describes the Token Bucket + constraints.
>> > This part could then emphasize that the described=20
>parameters can be=20
>> > used for interoperability purposes.
>> >
>> > Another part of the QSPEC template draft will describe the other=20
>> > parameters.
>> > The goal of this part is to provide guidelines to QoS=20
>models on how=20
>> > to encode the different parameters, with the currently available=20
>> > parameters used as examples."
>>
>> As I said at the meeting:
>> the current QSPEC template draft actually contains both=20
>aspects since=20
>> we started with the template idea first and then added the=20
>> interoperability aspects. An even cleaner separation would=20
>be a split=20
>> into two separate documents:
>> - one (template focus) explaining the general idea and layout of=20
>> QSPECs,
>>   which could be used by other QoS models as guidance to create their
>>   own QSPECs, and
>> - one (interoperability aspects) proposing a generic QSPEC as
>>   an example of a QSPEC instance
>> However, it's also ok for me to describe both aspects in one=20
>document=20
>> as it is probably now too late and too disturbing to make this=20
>> separation.
>
>OK, Georgios' proposal sounds like a good compromise.
>
>Thanks,
>Regards,
>Jerry
>
>>
>> Regards,
>>  Roland
>
>
>_______________________________________________
>nsis mailing list
>nsis@ietf.org
>https://www1.ietf.org/mailman/listinfo/nsis
>

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 30 02:39:34 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GpgVH-0001Ku-5K; Thu, 30 Nov 2006 02:38:43 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GpgVF-0001Ko-6M
	for nsis@ietf.org; Thu, 30 Nov 2006 02:38:41 -0500
Received: from mgw-ext13.nokia.com ([131.228.20.172])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1GpgVD-0004CO-N9
	for nsis@ietf.org; Thu, 30 Nov 2006 02:38:41 -0500
Received: from esebh106.NOE.Nokia.com (esebh106.ntc.nokia.com [172.21.138.213])
	by mgw-ext13.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kAU7cYmr007376; Thu, 30 Nov 2006 09:38:36 +0200
Received: from esebh104.NOE.Nokia.com ([172.21.143.34]) by
	esebh106.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:38:22 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh104.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:38:22 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Date: Thu, 30 Nov 2006 09:38:20 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCC0172715F@esebe199.NOE.Nokia.com>
In-Reply-To: <5E7777BA-D9C1-4E4E-90E1-187F09A6236B@cisco.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Consensus check on: domain-local admission controloperation in
	NSLP/QSPEC
Thread-Index: AccEMXelUz7Wd5yeQwWjk/DFNc2c8AQIK+lA
From: <john.loughney@nokia.com>
To: <oran@cisco.com>, <nsis@ietf.org>
X-OriginalArrivalTime: 30 Nov 2006 07:38:22.0664 (UTC)
	FILETIME=[83557C80:01C71452]
X-eXpurgate-Category: 1/0
X-eXpurgate-ID: 149371::061130093837-1C6A1BB0-453EC547/0-0/0-1
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: 0bc60ec82efc80c84b8d02f4b0e4de22
Cc: 
Subject: [NSIS] Consensus check on: domain-local admission controloperation
	in NSLP/QSPEC
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Dave proposed the following:

>1. Do complete QSPEC stacking, and deal with the issues around=20
>what constraints have to be met by the stacked QSPEC, whether=20
>the QSPEC lower in the "stack" is opacified, and what the=20
>rules are for pushing, swapping, and popping.
>
>2. Allow domain-parameters to be added to a QSPEC on domain=20
>entry and removed on exit. In this option one has to describe=20
>whether anything is opacified and which existing QSPEC=20
>parameters get "re-interpreted" =20
>if any.
>
>3. Recurse at the NSLP level, have a separate domain-private=20
>NSLP session, and that would contain a domain-private set of=20
>paramteres instead of the end-to-end QSPEC. I'll note that=20
>this approach allows (but not requires) the two signaling=20
>state machines (end-end and edge- edge through interior nodes)=20
>to be decoupled.

Re-reading responses to this thread, I am left slightly confused
as to what people are supporting.  I think we need to break this
up into two questions, one that affects the QSPEC document
and one that affects the QoS NLSP.  I will send separate mails
asking for consensus on this issue.

thanks,
John

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 30 02:55:31 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gpgkw-0000PA-08; Thu, 30 Nov 2006 02:54:54 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gpgku-0000P3-JX
	for nsis@ietf.org; Thu, 30 Nov 2006 02:54:52 -0500
Received: from mgw-ext14.nokia.com ([131.228.20.173])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gpgkt-0000JE-2P
	for nsis@ietf.org; Thu, 30 Nov 2006 02:54:52 -0500
Received: from esebh105.NOE.Nokia.com (esebh105.ntc.nokia.com [172.21.138.211])
	by mgw-ext14.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kAU7sdhv031716 for <nsis@ietf.org>; Thu, 30 Nov 2006 09:54:49 +0200
Received: from esebh101.NOE.Nokia.com ([172.21.138.177]) by
	esebh105.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:54:39 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh101.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:54:40 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Thu, 30 Nov 2006 09:54:38 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCC0172718C@esebe199.NOE.Nokia.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Support for local domain information in the QSPEC
Thread-Index: AccUVMjBCgWWu3X5T6KN+j72Q8Re5Q==
From: <john.loughney@nokia.com>
To: <nsis@ietf.org>
X-OriginalArrivalTime: 30 Nov 2006 07:54:40.0084 (UTC)
	FILETIME=[C9EBF140:01C71454]
X-eXpurgate-Category: 3/0
X-eXpurgate-ID: 149371::061130095449-7B433BB0-3A41BA16/0-0/0-17
X-Nokia-AV: Clean
X-Spam-Score: 0.3 (/)
X-Scan-Signature: 386e0819b1192672467565a524848168
Subject: [NSIS] Support for local domain information in the QSPEC
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1003699062=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1003699062==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C71454.C9616715"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C71454.C9616715
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

I'm trying to tease through all of the mails on this, and so if my
interpretation is incorrect, please suggest corrective text.

At the previous IETF, Dave Oran proposed that we have a mathmatically
complete description of the traffic source.  I believe we have consensus
on that.=20

The open question is how to handle QoS model specific parameters:
traffic class, diff serv classes, etc.
1) Include a common set of them as optional parameters in the QSPEC
document and allow QoSM documents define additional parameters via an
IANA registry. =20
2) Allow QoSM documents to define their own complete QSPECs, which are
used within the QoSM domain, essentially defining an e2e QSPEC and a
local QSPEC.

I think Georgios summarized these quite well in:
http://www1.ietf.org/mail-archive/web/nsis/current/msg07164.html

I believe that more people were favoring #2.  Of course, the devil is
always in the details, as there have been discussions about what
parameters belong where.

Comments?
John

------_=_NextPart_001_01C71454.C9616715
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Dus-ascii">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
6.5.7650.21">
<TITLE>Support for local domain information in the QSPEC</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/rtf format -->

<P><FONT SIZE=3D2 FACE=3D"Arial">I'm trying to tease through all of the =
mails on this, and so if my interpretation is incorrect, please suggest =
corrective text.</FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">At the previous IETF, Dave Oran =
proposed that we have a mathmatically complete description of the =
traffic source.&nbsp; I believe we have consensus on that. </FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">The open question is how to handle QoS =
model specific parameters: traffic class, diff serv classes, etc.</FONT>

<BR><FONT SIZE=3D2 FACE=3D"Arial">1) Include a common set of them as =
optional parameters in the QSPEC document and allow QoSM documents =
define additional parameters via an IANA registry.&nbsp; </FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">2) Allow QoSM documents to define their =
own complete QSPECs, which are used within the QoSM domain, essentially =
defining an e2e QSPEC and a local QSPEC.</FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">I think Georgios summarized these quite =
well in:</FONT>

<BR><A =
HREF=3D"http://www1.ietf.org/mail-archive/web/nsis/current/msg07164.html"=
><U><FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial">http://www1.ietf.org/mail-archive/web/nsis/current/msg0716=
4.html</FONT></U></A>
</P>

<P><FONT SIZE=3D2 FACE=3D"Arial">I believe that more people were =
favoring #2.&nbsp; Of course, the devil is always in the details, as =
there have been discussions about what parameters belong =
where.</FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">Comments?</FONT>

<BR><FONT SIZE=3D2 FACE=3D"Arial">John</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C71454.C9616715--


--===============1003699062==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1003699062==--




From nsis-bounces@ietf.org Thu Nov 30 02:59:43 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gpgp0-0001TK-QO; Thu, 30 Nov 2006 02:59:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gpgoy-0001RE-Pz
	for nsis@ietf.org; Thu, 30 Nov 2006 02:59:04 -0500
Received: from mgw-ext03.nokia.com ([131.228.20.95])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gpgox-00013A-8p
	for nsis@ietf.org; Thu, 30 Nov 2006 02:59:04 -0500
Received: from esebh107.NOE.Nokia.com (esebh107.ntc.nokia.com [172.21.143.143])
	by mgw-ext03.nokia.com (Switch-3.1.10/Switch-3.1.10) with ESMTP id
	kAU7wqNq024183 for <nsis@ietf.org>; Thu, 30 Nov 2006 09:59:02 +0200
Received: from esebh102.NOE.Nokia.com ([172.21.138.183]) by
	esebh107.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:59:00 +0200
Received: from esebe199.NOE.Nokia.com ([172.21.138.143]) by
	esebh102.NOE.Nokia.com with Microsoft SMTPSVC(6.0.3790.1830); 
	Thu, 30 Nov 2006 09:58:59 +0200
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Date: Thu, 30 Nov 2006 09:58:58 +0200
Message-ID: <BAA65A575825454CBB0103267553FCCC01727194@esebe199.NOE.Nokia.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: QoS NSLP: Stacking vs. tunneling vs. 
Thread-Index: AccUVWO0cOBR4oZwSS2jn/HRA5ih0w==
From: <john.loughney@nokia.com>
To: <nsis@ietf.org>
X-OriginalArrivalTime: 30 Nov 2006 07:58:59.0684 (UTC)
	FILETIME=[64A7C240:01C71455]
X-eXpurgate-Category: 3/0
X-eXpurgate-ID: 149371::061130095902-5DDCFBB0-7E8768B5/0-0/0-17
X-Nokia-AV: Clean
X-Spam-Score: 0.2 (/)
X-Scan-Signature: b5d20af10c334b36874c0264b10f59f1
Subject: [NSIS] QoS NSLP: Stacking vs. tunneling vs. 
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============0778459554=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============0778459554==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C71455.645FA10E"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C71455.645FA10E
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Dave Oran proposed 3 options for supporting domain-local admission
control operation in NSLP/QSPEC in the following mail:
http://www1.ietf.org/mail-archive/web/nsis/current/msg07178.html

As stated, these are meant as complete solutions:

1. Do complete QSPEC stacking, and deal with the issues around what
constraints have to be met by the stacked QSPEC, whether the QSPEC lower
in the "stack" is opacified, and what the rules are for pushing,
swapping, and popping.

2. Allow domain-parameters to be added to a QSPEC on domain entry and
removed on exit. In this option one has to describe whether anything is
opacified and which existing QSPEC parameters get "re-interpreted" if
any.

3. Recurse at the NSLP level, have a separate domain-private NSLP
session, and that would contain a domain-private set of paramteres
instead of the end-to-end QSPEC. I'll note that this approach allows
(but not requires) the two signaling state machines (end-end and edge-
edge through interior nodes) to be decoupled.

There seems to be consensus that 3 is mandatory, but 2 is still
desirable.  Does this mean that 2 would be optional in the QoS NSLP? =20

John

------_=_NextPart_001_01C71455.645FA10E
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Dus-ascii">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
6.5.7650.21">
<TITLE>QoS NSLP: Stacking vs. tunneling vs. </TITLE>
</HEAD>
<BODY>
<!-- Converted from text/rtf format -->

<P><FONT SIZE=3D2 FACE=3D"Arial">Dave Oran proposed 3 options for =
supporting domain-local admission control operation in NSLP/QSPEC in the =
following mail:</FONT></P>

<P><A =
HREF=3D"http://www1.ietf.org/mail-archive/web/nsis/current/msg07178.html"=
><U><FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial">http://www1.ietf.org/mail-archive/web/nsis/current/msg0717=
8.html</FONT></U></A>
</P>

<P><FONT SIZE=3D2 FACE=3D"Arial">As stated, these are meant as complete =
solutions:</FONT>
</P>

<P><FONT SIZE=3D2 FACE=3D"Arial">1. Do complete QSPEC stacking, and deal =
with the issues around what constraints have to be met by the stacked =
QSPEC, whether the QSPEC lower in the &quot;stack&quot; is opacified, =
and what the rules are for pushing, swapping, and popping.</FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">2. Allow domain-parameters to be added =
to a QSPEC on domain entry and removed on exit. In this option one has =
to describe whether anything is opacified and which existing QSPEC =
parameters get &quot;re-interpreted&quot; if any.</FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">3. Recurse at the NSLP level, have a =
separate domain-private NSLP session, and that would contain a =
domain-private set of paramteres instead of the end-to-end QSPEC. I'll =
note that this approach allows (but not requires) the two signaling =
state machines (end-end and edge- edge through interior nodes) to be =
decoupled.</FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">There seems to be consensus that 3 is =
mandatory, but 2 is still desirable.&nbsp; Does this mean that 2 would =
be optional in the QoS NSLP?&nbsp; </FONT></P>

<P><FONT SIZE=3D2 FACE=3D"Arial">John</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C71455.645FA10E--


--===============0778459554==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============0778459554==--




From nsis-bounces@ietf.org Thu Nov 30 04:11:39 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gphwg-0000Sj-Og; Thu, 30 Nov 2006 04:11:06 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gphwd-0000Lu-Pa
	for nsis@ietf.org; Thu, 30 Nov 2006 04:11:03 -0500
Received: from courier.cs.helsinki.fi ([128.214.9.1] helo=mail.cs.helsinki.fi)
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gphwc-0008F3-DR
	for nsis@ietf.org; Thu, 30 Nov 2006 04:11:03 -0500
Received: from sbz-30.cs.helsinki.fi (sbz-30.cs.helsinki.fi [128.214.9.98])
	(TLS: TLSv1/SSLv3,256bits,AES256-SHA)
	by mail.cs.helsinki.fi with esmtp; Thu, 30 Nov 2006 11:11:01 +0200
	id 00080898.456EA025.00007035
Date: Thu, 30 Nov 2006 11:11:01 +0200 (EET)
From: Jukka MJ Manner <jmanner@cs.Helsinki.FI>
To: john.loughney@nokia.com
Subject: Re: [NSIS] QoS NSLP: Stacking vs. tunneling vs.
In-Reply-To: <BAA65A575825454CBB0103267553FCCC01727194@esebe199.NOE.Nokia.com>
Message-ID: <Pine.LNX.4.64.0611301109090.22464@sbz-30.cs.Helsinki.FI>
References: <BAA65A575825454CBB0103267553FCCC01727194@esebe199.NOE.Nokia.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 52e1467c2184c31006318542db5614d5
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org


Item 2 does not affect the QoS NSLP in any way. That was the main point. 
It is left to QOSM designers and the RMF. The NSLP doesn't care what the 
contents of the QSPEC is, and whether it is updated in some QNE or not.

Jukka

On Thu, 30 Nov 2006, john.loughney@nokia.com wrote:

> Dave Oran proposed 3 options for supporting domain-local admission
> control operation in NSLP/QSPEC in the following mail:
> http://www1.ietf.org/mail-archive/web/nsis/current/msg07178.html
> 
> As stated, these are meant as complete solutions:
> 
> 1. Do complete QSPEC stacking, and deal with the issues around what
> constraints have to be met by the stacked QSPEC, whether the QSPEC lower
> in the "stack" is opacified, and what the rules are for pushing,
> swapping, and popping.
> 
> 2. Allow domain-parameters to be added to a QSPEC on domain entry and
> removed on exit. In this option one has to describe whether anything is
> opacified and which existing QSPEC parameters get "re-interpreted" if
> any.
> 
> 3. Recurse at the NSLP level, have a separate domain-private NSLP
> session, and that would contain a domain-private set of paramteres
> instead of the end-to-end QSPEC. I'll note that this approach allows
> (but not requires) the two signaling state machines (end-end and edge-
> edge through interior nodes) to be decoupled.
> 
> There seems to be consensus that 3 is mandatory, but 2 is still
> desirable.  Does this mean that 2 would be optional in the QoS NSLP?  
> 
> John
> 

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 30 07:01:19 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1Gpkai-0007IS-3q; Thu, 30 Nov 2006 07:00:36 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1Gpkah-0007IM-DQ
	for nsis@ietf.org; Thu, 30 Nov 2006 07:00:35 -0500
Received: from rotterdam.ewi.utwente.nl ([130.89.10.5])
	by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1Gpkag-0007ac-0J
	for nsis@ietf.org; Thu, 30 Nov 2006 07:00:35 -0500
Received: from utip105 (utip105.ewi.utwente.nl [130.89.13.76])
	by rotterdam.ewi.utwente.nl (8.13.6/8.13.6) with ESMTP id
	kAUC0PTJ014622; Thu, 30 Nov 2006 13:00:29 +0100 (MET)
From: "Georgios Karagiannis" <karagian@cs.utwente.nl>
To: "'Jukka MJ Manner'" <jmanner@cs.Helsinki.FI>, <john.loughney@nokia.com>
Subject: RE: [NSIS] QoS NSLP: Stacking vs. tunneling vs.
Date: Thu, 30 Nov 2006 13:00:21 +0100
Message-ID: <000a01c71477$1f8a0070$4c0d5982@dynamic.ewi.utwente.nl>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Office Outlook 11
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2962
In-Reply-To: <Pine.LNX.4.64.0611301109090.22464@sbz-30.cs.Helsinki.FI>
Thread-Index: AccUX+6MUDt8n0k1QU2/yJu3J3ffkgAFvkbg
X-Spam-Score: 0.122 () AWL
X-Scanned-By: MIMEDefang 2.52 on 130.89.10.5
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-3.0rc3
	(rotterdam.ewi.utwente.nl [130.89.10.5]);
	Thu, 30 Nov 2006 13:00:30 +0100 (MET)
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 4b800b1eab964a31702fa68f1ff0e955
Cc: nsis@ietf.org
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Errors-To: nsis-bounces@ietf.org

Hi John

Option 2 will be described in the QSPEC templete draft and in the QoS models
that require 
to use this feature.


Best Regards,
Georgios
 

> -----Original Message-----
> From: Jukka MJ Manner [mailto:jmanner@cs.Helsinki.FI] 
> Sent: donderdag 30 november 2006 10:11
> To: john.loughney@nokia.com
> Cc: nsis@ietf.org
> Subject: Re: [NSIS] QoS NSLP: Stacking vs. tunneling vs.
> 
> 
> Item 2 does not affect the QoS NSLP in any way. That was the 
> main point. 
> It is left to QOSM designers and the RMF. The NSLP doesn't 
> care what the contents of the QSPEC is, and whether it is 
> updated in some QNE or not.
> 
> Jukka
> 
> On Thu, 30 Nov 2006, john.loughney@nokia.com wrote:
> 
> > Dave Oran proposed 3 options for supporting domain-local admission 
> > control operation in NSLP/QSPEC in the following mail:
> > http://www1.ietf.org/mail-archive/web/nsis/current/msg07178.html
> > 
> > As stated, these are meant as complete solutions:
> > 
> > 1. Do complete QSPEC stacking, and deal with the issues around what 
> > constraints have to be met by the stacked QSPEC, whether the QSPEC 
> > lower in the "stack" is opacified, and what the rules are 
> for pushing, 
> > swapping, and popping.
> > 
> > 2. Allow domain-parameters to be added to a QSPEC on domain 
> entry and 
> > removed on exit. In this option one has to describe whether 
> anything 
> > is opacified and which existing QSPEC parameters get 
> "re-interpreted" 
> > if any.
> > 
> > 3. Recurse at the NSLP level, have a separate domain-private NSLP 
> > session, and that would contain a domain-private set of paramteres 
> > instead of the end-to-end QSPEC. I'll note that this 
> approach allows 
> > (but not requires) the two signaling state machines 
> (end-end and edge- 
> > edge through interior nodes) to be decoupled.
> > 
> > There seems to be consensus that 3 is mandatory, but 2 is still 
> > desirable.  Does this mean that 2 would be optional in the QoS NSLP?
> > 
> > John
> > 
> 
> _______________________________________________
> nsis mailing list
> nsis@ietf.org
> https://www1.ietf.org/mailman/listinfo/nsis
> 



_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis



From nsis-bounces@ietf.org Thu Nov 30 07:43:39 2006
Received: from [127.0.0.1] (helo=stiedprmman1.va.neustar.com)
	by megatron.ietf.org with esmtp (Exim 4.43)
	id 1GplFS-0001ol-69; Thu, 30 Nov 2006 07:42:42 -0500
Received: from [10.91.34.44] (helo=ietf-mx.ietf.org)
	by megatron.ietf.org with esmtp (Exim 4.43) id 1GplFP-0001Uj-Po
	for nsis@ietf.org; Thu, 30 Nov 2006 07:42:39 -0500
Received: from mail121.messagelabs.com ([216.82.241.195])
	by ietf-mx.ietf.org with smtp (Exim 4.43) id 1GplFO-0006Lq-CQ
	for nsis@ietf.org; Thu, 30 Nov 2006 07:42:39 -0500
X-VirusChecked: Checked
X-Env-Sender: gash@att.com
X-Msg-Ref: server-6.tower-121.messagelabs.com!1164890557!10467484!1
X-StarScan-Version: 5.5.10.7; banners=-,-,-
X-Originating-IP: [134.24.146.4]
Received: (qmail 8009 invoked from network); 30 Nov 2006 12:42:37 -0000
Received: from unknown (HELO attrh3i.attrh.att.com) (134.24.146.4)
	by server-6.tower-121.messagelabs.com with SMTP;
	30 Nov 2006 12:42:37 -0000
Received: from attrh.att.com (localhost [127.0.0.1])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id kAUCgaPK005072; 
	Thu, 30 Nov 2006 07:42:36 -0500 (EST)
Received: from kcclust06evs1.ugd.att.com (kcst11.ugd.att.com [135.38.164.88])
	by attrh3i.attrh.att.com (8.13.7/8.13.7) with ESMTP id
	kAUCgUEU005038; Thu, 30 Nov 2006 07:42:31 -0500 (EST)
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Subject: RE: [NSIS] Support for local domain information in the QSPEC
Date: Thu, 30 Nov 2006 06:42:02 -0600
Message-ID: <9473683187ADC049A855ED2DA739ABCA0DDC7FC8@KCCLUST06EVS1.ugd.att.com>
In-Reply-To: <BAA65A575825454CBB0103267553FCCC0172718C@esebe199.NOE.Nokia.com>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: [NSIS] Support for local domain information in the QSPEC
Thread-Index: AccUVMjBCgWWu3X5T6KN+j72Q8Re5QAJt/mw
From: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
To: <john.loughney@nokia.com>, <nsis@ietf.org>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 48472a944c87678fcfe8db15ffecdfff
Cc: "Ash, Gerald R \(Jerry\), ALABS" <gash@att.com>
X-BeenThere: nsis@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Next Steps in Signaling <nsis.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=unsubscribe>
List-Post: <mailto:nsis@ietf.org>
List-Help: <mailto:nsis-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/nsis>,
	<mailto:nsis-request@ietf.org?subject=subscribe>
Content-Type: multipart/mixed; boundary="===============1034227229=="
Errors-To: nsis-bounces@ietf.org

This is a multi-part message in MIME format.

--===============1034227229==
Content-class: urn:content-classes:message
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C7147C.EF32553A"

This is a multi-part message in MIME format.

------_=_NextPart_001_01C7147C.EF32553A
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable

Hi John,
=20
My impression was clearly agreement on option 1, which I think Georgios
originally described in his post at=20
http://www1.ietf.org/mail-archive/web/nsis/current/msg07213.html
and again at
http://www1.ietf.org/mail-archive/web/nsis/current/msg07232.html
=20
On the thread I think most agreed with Georgios' suggestion, see=20
http://www1.ietf.org/mail-archive/web/nsis/current/msg07237.html
and then
http://www1.ietf.org/mail-archive/web/nsis/current/msg07238.html
=20
Thanks,
Jerry

________________________________

From: john.loughney@nokia.com [mailto:john.loughney@nokia.com]=20
Sent: Thursday, November 30, 2006 2:55 AM
To: nsis@ietf.org
Subject: [NSIS] Support for local domain information in the QSPEC



I'm trying to tease through all of the mails on this, and so if my
interpretation is incorrect, please suggest corrective text.

At the previous IETF, Dave Oran proposed that we have a mathmatically
complete description of the traffic source.  I believe we have consensus
on that.=20

The open question is how to handle QoS model specific parameters:
traffic class, diff serv classes, etc.=20
1) Include a common set of them as optional parameters in the QSPEC
document and allow QoSM documents define additional parameters via an
IANA registry. =20

2) Allow QoSM documents to define their own complete QSPECs, which are
used within the QoSM domain, essentially defining an e2e QSPEC and a
local QSPEC.

I think Georgios summarized these quite well in:=20
http://www1.ietf.org/mail-archive/web/nsis/current/msg07164.html
<http://www1.ietf.org/mail-archive/web/nsis/current/msg07164.html> =20

I believe that more people were favoring #2.  Of course, the devil is
always in the details, as there have been discussions about what
parameters belong where.

Comments?=20
John=20


------_=_NextPart_001_01C7147C.EF32553A
Content-Type: text/html;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD><TITLE>Support for local domain information in the =
QSPEC</TITLE>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Dus-ascii">
<META content=3D"MSHTML 6.00.2900.2995" name=3DGENERATOR></HEAD>
<BODY>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Hi John,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>My impression was clearly agreement on option =
1, which I=20
think Georgios originally described in his post at </FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2><A=20
href=3D"http://www1.ietf.org/mail-archive/web/nsis/current/msg07213.html"=
>http://www1.ietf.org/mail-archive/web/nsis/current/msg07213.html</A></FO=
NT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>and again at</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2><A=20
href=3D"http://www1.ietf.org/mail-archive/web/nsis/current/msg07232.html"=
>http://www1.ietf.org/mail-archive/web/nsis/current/msg07232.html</A></FO=
NT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>On&nbsp;the thread I think most agreed with =
Georgios'=20
suggestion, see </FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2><A=20
href=3D"http://www1.ietf.org/mail-archive/web/nsis/current/msg07237.html"=
>http://www1.ietf.org/mail-archive/web/nsis/current/msg07237.html</A></FO=
NT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>and then</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2><A=20
href=3D"http://www1.ietf.org/mail-archive/web/nsis/current/msg07238.html"=
>http://www1.ietf.org/mail-archive/web/nsis/current/msg07238.html</A></FO=
NT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2></FONT></SPAN>&nbsp;</DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Thanks,</FONT></SPAN></DIV>
<DIV dir=3Dltr align=3Dleft><SPAN class=3D644543212-30112006><FONT =
face=3DArial=20
color=3D#0000ff size=3D2>Jerry</FONT></SPAN></DIV><BR>
<DIV class=3DOutlookMessageHeader lang=3Den-us dir=3Dltr align=3Dleft>
<HR tabIndex=3D-1>
<FONT face=3DTahoma size=3D2><B>From:</B> john.loughney@nokia.com=20
[mailto:john.loughney@nokia.com] <BR><B>Sent:</B> Thursday, November 30, =
2006=20
2:55 AM<BR><B>To:</B> nsis@ietf.org<BR><B>Subject:</B> [NSIS] Support =
for local=20
domain information in the QSPEC<BR></FONT><BR></DIV>
<DIV></DIV><!-- Converted from text/rtf format -->
<P><FONT face=3DArial size=3D2>I'm trying to tease through all of the =
mails on this,=20
and so if my interpretation is incorrect, please suggest corrective=20
text.</FONT></P>
<P><FONT face=3DArial size=3D2>At the previous IETF, Dave Oran proposed =
that we have=20
a mathmatically complete description of the traffic source.&nbsp; I =
believe we=20
have consensus on that. </FONT></P>
<P><FONT face=3DArial size=3D2>The open question is how to handle QoS =
model specific=20
parameters: traffic class, diff serv classes, etc.</FONT> <BR><FONT =
face=3DArial=20
size=3D2>1) Include a common set of them as optional parameters in the =
QSPEC=20
document and allow QoSM documents define additional parameters via an =
IANA=20
registry.&nbsp; </FONT></P>
<P><FONT face=3DArial size=3D2>2) Allow QoSM documents to define their =
own complete=20
QSPECs, which are used within the QoSM domain, essentially defining an =
e2e QSPEC=20
and a local QSPEC.</FONT></P>
<P><FONT face=3DArial size=3D2>I think Georgios summarized these quite =
well=20
in:</FONT> <BR><A=20
href=3D"http://www1.ietf.org/mail-archive/web/nsis/current/msg07164.html"=
><U><FONT=20
face=3DArial color=3D#0000ff=20
size=3D2>http://www1.ietf.org/mail-archive/web/nsis/current/msg07164.html=
</FONT></U></A>=20
</P>
<P><FONT face=3DArial size=3D2>I believe that more people were favoring =
#2.&nbsp; Of=20
course, the devil is always in the details, as there have been =
discussions about=20
what parameters belong where.</FONT></P>
<P><FONT face=3DArial size=3D2>Comments?</FONT> <BR><FONT face=3DArial=20
size=3D2>John</FONT> </P></BODY></HTML>

------_=_NextPart_001_01C7147C.EF32553A--


--===============1034227229==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
nsis mailing list
nsis@ietf.org
https://www1.ietf.org/mailman/listinfo/nsis

--===============1034227229==--




