Re: [Ibnemo] 答复: 答复: 答复: 答复: Defining a Common Model for intent

"Susan Hares" <shares@ndzh.com> Thu, 04 June 2015 18:23 UTC

Return-Path: <shares@ndzh.com>
X-Original-To: ibnemo@ietfa.amsl.com
Delivered-To: ibnemo@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D730C1A87F0 for <ibnemo@ietfa.amsl.com>; Thu, 4 Jun 2015 11:23:30 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -98.154
X-Spam-Level:
X-Spam-Status: No, score=-98.154 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DOS_OUTLOOK_TO_MX=2.845, HTML_MESSAGE=0.001, J_CHICKENPOX_45=0.6, MIME_8BIT_HEADER=0.3, USER_IN_WHITELIST=-100] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id PqBBwlrCllb2 for <ibnemo@ietfa.amsl.com>; Thu, 4 Jun 2015 11:23:25 -0700 (PDT)
Received: from hickoryhill-consulting.com (hhc-web3.hickoryhill-consulting.com [64.9.205.143]) by ietfa.amsl.com (Postfix) with ESMTP id E77B31A8833 for <ibnemo@ietf.org>; Thu, 4 Jun 2015 11:23:24 -0700 (PDT)
X-Default-Received-SPF: pass (skip=loggedin (res=PASS)) x-ip-name=184.157.80.157;
From: "Susan Hares" <shares@ndzh.com>
To: "'zhangyali \(D\)'" <zhangyali369@huawei.com>, "'PEDRO ANDRES ARANDA GUTIERREZ'" <pedroa.aranda@telefonica.com>, <nfvrg@irtf.org>
References: <D19315B6.1D624%pedroa.aranda@telefonica.com> <A747A0713F56294D8FBE33E5C6B8F5812951121A@szxeml513-mbx.china.huawei.com> <01b901d09d8b$96a057d0$c3e10770$@ndzh.com> <A747A0713F56294D8FBE33E5C6B8F581295113F0@szxeml513-mbx.china.huawei.com> <003001d09e69$178548f0$468fdad0$@ndzh.com> <A747A0713F56294D8FBE33E5C6B8F58129512109@szxeml513-mbx.china.huawei.com>
In-Reply-To: <A747A0713F56294D8FBE33E5C6B8F58129512109@szxeml513-mbx.china.huawei.com>
Date: Thu, 4 Jun 2015 14:23:18 -0400
Message-ID: <013101d09ef3$87fe83f0$97fb8bd0$@ndzh.com>
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="----=_NextPart_000_0132_01D09ED2.00F79250"
X-Mailer: Microsoft Outlook 14.0
Thread-Index: AQIyszWq5a5j3lzymrlGmCaVlNwkHQJHhrxxAbH20dsBPwGu8gKlCArWAnaVWDichcdUsA==
Content-Language: en-us
X-Authenticated-User: skh@ndzh.com
Archived-At: <http://mailarchive.ietf.org/arch/msg/ibnemo/CiB8sYsfedSshttEXDDAaDCKOo4>
Cc: draft-xia-ibnemo-icim@tools.ietf.org, ibnemo@ietf.org
Subject: Re: [Ibnemo] =?utf-8?b?562U5aSNOiAg562U5aSNOiAg562U5aSNOiDnrZTlpI06?= =?utf-8?q?__Defining_a_Common_Model_for_intent?=
X-BeenThere: ibnemo@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: "Discussion of Nemo, an intent-based North Bound \(NB\) interface consisting of an application protocol running over HTTP \(RESTful interfaces\) to exchange intent-based primitives between applications and meta-controllers controlling virtual network resources \(networks, storage, CPU\)." <ibnemo.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/ibnemo>, <mailto:ibnemo-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/ibnemo/>
List-Help: <mailto:ibnemo-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/ibnemo>, <mailto:ibnemo-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 04 Jun 2015 18:23:31 -0000

Yali: 

 

I agree that context is the external background information

user  à intent à context.

I struggle with the line between background and intent.  If intent is decomposed into object, result and constraint, the real question is whether the intent is pure or mixed.  If we are adding constraints then it is impure or constraint based.   My difficultly is that I know that constraints operate within a context.  In your example, costs operate within the context of a monetary evaluation ($1000, $2000) and line costs operate within a market (price goes up and down).  

 

I can agree that intent is object, result and constraint.   Howevere there is a a lot of context in constraint.  Have you observed this? 

 

Sue 

 

 

From: Ibnemo [mailto:ibnemo-bounces@ietf.org] On Behalf Of zhangyali (D)
Sent: Wednesday, June 03, 2015 11:03 PM
To: Susan Hares; 'PEDRO ANDRES ARANDA GUTIERREZ'; nfvrg@irtf.org
Cc: draft-xia-ibnemo-icim@tools.ietf.org; ibnemo@ietf.org
Subject: [Ibnemo] 答复: 答复: 答复: 答复: Defining a Common Model for intent

 

Sue:

 

Sorry for having different understanding for context of this example. 

 

In this example, price, as the context,  is marked by the provider of pipelines which is identical to all the users. For example, common circuit is marked as $1000, and dedicated line is marked $2000, etc. So context is the background information which could not changed as the result of intent. It’s free for users to describe context or not when expressing intent. Though context can be obtained by system automatically, it is important factor for the implementation of intent. In this example, when system choose the least cost, it will know the price of each way. So intent is the background information of intent, rather than part of intent’ content.

 

Therefore, the fundamental of intent always be: user  à intent à context. 

User is the subject to express intent (the enterprise, in this example), and intent is the result he desired (i.e. connect two sites with the least cost), and context is the external background information (i.e. the price of each line)

 

Though this example, maybe we can dig up the content of intent. The real intent is connect two sites with the least cost. So, intent can be decomposed into object and result. In this example, objects are these two sites, and result is the connectivity between these objects with the least money. If we treat the constraint as a individual part of intent, intent can be decomposed into object, result and constraint. In this example, objects are also these two sites, result is the connectivity, and constraint is the least cost.

 

Are you agree for the definition of context? And do you think this decomposition is suitable? Or any comments?

 

Yali

 

发件人: Susan Hares [mailto:shares@ndzh.com] 
发送时间: 2015年6月4日 9:52
收件人: zhangyali (D); 'PEDRO ANDRES ARANDA GUTIERREZ'; nfvrg@irtf.org
抄送: draft-xia-ibnemo-icim@tools.ietf.org; ibnemo@ietf.org
主题: RE: [Ibnemo] 答复: 答复: 答复: Defining a Common Model for intent

 

Yali:

 

I this example really helps.  I think the intent is the connection and the price a context that the user may want to have as a constraint. 

 

Intent à users à context --- constraint. 

 

Does this make sense? 

 

Sue 

 

From: Ibnemo [mailto:ibnemo-bounces@ietf.org] On Behalf Of zhangyali (D)
Sent: Tuesday, June 02, 2015 11:44 PM
To: Susan Hares; 'PEDRO ANDRES ARANDA GUTIERREZ'; nfvrg@irtf.org
Cc: draft-xia-ibnemo-icim@tools.ietf.org; ibnemo@ietf.org
Subject: [Ibnemo] 答复: 答复: 答复: Defining a Common Model for intent

 

Hi Sue:

 

It’s very good for coming back to the fundamental, even though discussing about classification of intent is also a meaningful topic. 

 

For a user, intent is the result he desired or some operations he want to take in a context. For example, a end-user want to make the communication between two sites is the minimum. For this intent, price is the context. Though context is omitted usually, it is really a important factor to affect the decision. Do you think this example is suitable to express general intent?

 

Best,

Yali

 

 

发件人: Susan Hares [mailto:shares@ndzh.com] 
发送时间: 2015年6月3日 7:27
收件人: zhangyali (D); 'PEDRO ANDRES ARANDA GUTIERREZ'; nfvrg@irtf.org
抄送: draft-xia-ibnemo-icim@tools.ietf.org; ibnemo@ietf.org
主题: RE: [Ibnemo] 答复: 答复: Defining a Common Model for intent

 

Yali:

 

While I believe the lower layers exist to serve the upper layers, I think Intent is a quality of declarative policy which provides a prescriptive rather than descriptive description.   It is not upper/lower layers – for that reduces the concept. 

 

I still come back to you fundamental: User (of the layer) à intent à context. 

 

Do you have any other examples? 

 

Sue 

 

 

From: Ibnemo [mailto:ibnemo-bounces@ietf.org] On Behalf Of zhangyali (D)
Sent: Tuesday, June 02, 2015 8:29 AM
To: PEDRO ANDRES ARANDA GUTIERREZ; Susan Hares; nfvrg@irtf.org
Cc: draft-xia-ibnemo-icim@tools.ietf.org; ibnemo@ietf.org
Subject: [Ibnemo] 答复: 答复: Defining a Common Model for intent

 

Hi Pedro,

 

That’s very good for enumerating these two different viewpoints when discuss the layer of intent. Definitely, these two viewpoints are two important perspectives for intent representation.

 

I am not sure if my understanding is consistent with you, and I assume I have gotten what you said. So if my understanding is wrong, please point it out without any hesitation.

 

For viewpoint 1, intent is divided into hierarchical relationship according a specific implementation procedure which lower layers are completely used to fulfill the requirements of upper layers. At this point, the intent of lower layer serve as the implementation and the rendering layer by layer lead to the realization of highest layer’ intent. This confirms to implementation procedure of a specific affair, however, if we consider the responsibility of the whole layer, there are some intent just serve itself, i.e. A administrator want to do a backup line to ensure the stabilization. Another problem is that there may be some extra intent independent of present upper layer intent.

 

For viewpoint 2, intent is divided according to roles differentiation which have different knowledge, responsibility, etc. Users with different roles can express their intent separately, and different roles do not have the relationship of serve-client. All users with different roles have the liberty to express their own intent without worrying about what is the other users’ intent, even though the function is used by them in some specific affair. This classification give users with different roles liberty to express their intent which may involve brief vision or detailed function.

 

By comparing these two viewpoints, I think the classification of intent according to the difference of roles is a better way. Do you have same opinion or any comments?

 

Best Regards,

Yali

 

 

发件人: PEDRO ANDRES ARANDA GUTIERREZ [mailto:pedroa.aranda@telefonica.com] 
发送时间: 2015年6月2日 14:38
收件人: zhangyali (D); Susan Hares; nfvrg@irtf.org
抄送: draft-xia-ibnemo-icim@tools.ietf.org; ibnemo@ietf.org
主题: Re: 答复: [Ibnemo] Defining a Common Model for intent

 

Hi Yali,

 

let’s keep in the networking domain :-) I strongly believe that we need different levels or layers of intent. However, there are (at least) two different viewpoints:

 

If you follow the infrastructure view(which is where I feel more comfortable), I hope we agree that it is a completely different situation when you are designing a network element-by-element than when you are designing the network at a sub-domain level (for example levels in an IS-IS based network or areas if you use OSPF) or if you are defining the interconnections of a service provider’s AS (and dealing with BGP-4 policies) or if you are defining an end-to-end service. Although at the end, the upper layers will use all the features provided by the lower layers.

 

Now, I’m sure we can find the equivalent layering from a role point of view:

 

The user wants to access a service (for example a Web page), the provider of that Web page wants it to be served with the best quality of experience and so he chooses a specific provider – normally a CDN). The CDN provider will choose a given carrier to get access to the user’s service provider. The user’s service provider will dimension his network to fulfill a series of criteria. Within the service provider, the operators sitting at the Network Operations Centre will have to fulfill a series of KPIs, etc. 

 

As you see, I’m more a ‘box’ thinker. However, if someone can complete the ‘role’ example we can compare both approaches and try to identify if we can do any mapping between the two views. Maybe we could come up at end with a set of common denominators we can use to continue this discussion.

 

Best,

/PA

 

 

De: "zhangyali (D)" <zhangyali369@huawei.com>
Fecha: martes, 2 de junio de 2015 05:52
Para: PEDRO ANDRES ARANDA GUTIERREZ <pedroa.aranda@telefonica.com>om>, Sue Hares <shares@ndzh.com>om>, "nfvrg@irtf.org" <nfvrg@irtf.org>
CC: "draft-xia-ibnemo-icim@tools.ietf.org" <draft-xia-ibnemo-icim@tools.ietf.org>rg>, "ibnemo@ietf.org" <ibnemo@ietf.org>
Asunto: 答复: [Ibnemo] Defining a Common Model for intent
Nuevo envío de: <zhangyali369@huawei.com>
Nuevo envío para: <draft-xia-ibnemo-icim@ietf.org>
Fecha de nuevo envío: martes, 2 de junio de 2015 05:52

 

Hi Pedro,

 

Thanks for reviewing the draft and giving modification.

 

The question you have mentioned is a very important point for the abstraction of intent model. Maybe we can propose the transport market as a analogy.

1.       A customer wants to transport his goods from A to B. So his intent is getting his goods from A to B without carrying about how to do it. Then his intent is transferred to the transportation system. 

2.       This system analyzes customer’s requirement, and choose a suitable way to complete the requirement. For example, the system choose truck as the means. So the intent of transportation system is transferring the goods with truck.

3.       The driver of this truck analyze the path from A to B, and choose a most appropriate path to complete this order which will save more time. So  the intent of driver may be transferring the goods with the least time. Then the driver will start the engine, step on the gas, etc.

 

>From this analogy, the ultimate effect is the same, namely, transfer the goods from A to B. But the specific intent of different roles has some differences which depends on user’ role, knowledge, responsibility, etc. For example, transportation system is responsible for transporting goods, and he know the various ways. So he can form his intent by rendering the upper customer’s intent.

 

Supposing we divide users into different layers according to the implementation series, users in upper layer expresses his intent as what he want without having the knowledge about how to do it. Then the how procedure will be transferred to what in the lower layer according to knowledge and context. These transfer procedure lead to the completion of requirement. Same with the example in draft. Although the ultimate effect is same, the focus is different which will bring out the differentiation of intent.

 

This is just my immature opinion about intent. Do you think the differentiation of intent to complete the same thing is important and reasonable?

 

Best Regards,

 

Yali

 

发件人: PEDRO ANDRES ARANDA GUTIERREZ [mailto:pedroa.aranda@telefonica.com] 
发送时间: 2015年6月1日 17:15
收件人: Susan Hares; nfvrg@irtf.org
抄送: draft-xia-ibnemo-icim@tools.ietf.org; ibnemo@ietf.org
主题: Re: [Ibnemo] Defining a Common Model for intent

 

Hi,

 

A small clarification proposal for draft  <https://datatracker.ietf.org/doc/draft-xia-ibnemo-icim/> https://datatracker.ietf.org/doc/draft-xia-ibnemo-icim/ .

 

In section 2.4, I would leave the following as a paragraph

For example, in the network area the intent of end-users could be
safe connectivity between two sites which a technology independent
and device independent requirement. For business-based network
designers, the network connectivity can be selected which is device-
independent but technology specific. An example of the business-based
technology is the L3VPN. 
And change:
For network administrators, intent can be
specific operations on a set of devices such as configuring IP
addresses on network servers in a data center.

To

 

For network administrators, intent can be <new>defining a network topology like a router connected to a firewall, connected to a load balancer and this to two L2 networks where WWW servers sit or specifying the</new> operations on a set of devices such as configuring IP addresses on network servers in a data center.
 
Rationale behind this is again, that intent should be anything that is invariant and that expresses what a network operator/administrator may need to do, as opposed to how he would do that, i.e. The router is a HW device from vendor X or a virtual machine running a specific routing daemon over a given data-path implementation.
Best, /PA

---

Dr. Pedro A. Aranda Gutiérrez

 

Technology Exploration -

Network Innovation & Virtualisation

email: pedroa d0t aranda At telefonica d0t com

Telefónica, Investigación y Desarrollo

C/ D. Ramón de la Cruz,84

28006 Madrid, Spain

 

Fragen sind nicht da, um beantwortet zu werden.

Fragen sind da, um gestellt zu werden.

Georg Kreisler

 


  _____  



Este mensaje y sus adjuntos se dirigen exclusivamente a su destinatario, puede contener información privilegiada o confidencial y es para uso exclusivo de la persona o entidad de destino. Si no es usted. el destinatario indicado, queda notificado de que la lectura, utilización, divulgación y/o copia sin autorización puede estar prohibida en virtud de la legislación vigente. Si ha recibido este mensaje por error, le rogamos que nos lo comunique inmediatamente por esta misma vía y proceda a su destrucción.

The information contained in this transmission is privileged and confidential information intended only for the use of the individual or entity named above. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this transmission in error, do not read it. Please immediately reply to the sender that you have received this communication in error and then delete it.

Esta mensagem e seus anexos se dirigem exclusivamente ao seu destinatário, pode conter informação privilegiada ou confidencial e é para uso exclusivo da pessoa ou entidade de destino. Se não é vossa senhoria o destinatário indicado, fica notificado de que a leitura, utilização, divulgação e/ou cópia sem autorização pode estar proibida em virtude da legislação vigente. Se recebeu esta mensagem por erro, rogamos-lhe que nos o comunique imediatamente por esta mesma via e proceda a sua destruição

 

 

 

---

Dr. Pedro A. Aranda Gutiérrez

 

Technology Exploration -

Network Innovation & Virtualisation

email: pedroa d0t aranda At telefonica d0t com

Telefónica, Investigación y Desarrollo

C/ D. Ramón de la Cruz,84

28006 Madrid, Spain

 

Fragen sind nicht da, um beantwortet zu werden.

Fragen sind da, um gestellt zu werden.

Georg Kreisler

 

  _____  


Este mensaje y sus adjuntos se dirigen exclusivamente a su destinatario, puede contener información privilegiada o confidencial y es para uso exclusivo de la persona o entidad de destino. Si no es usted. el destinatario indicado, queda notificado de que la lectura, utilización, divulgación y/o copia sin autorización puede estar prohibida en virtud de la legislación vigente. Si ha recibido este mensaje por error, le rogamos que nos lo comunique inmediatamente por esta misma vía y proceda a su destrucción.

The information contained in this transmission is privileged and confidential information intended only for the use of the individual or entity named above. If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this transmission in error, do not read it. Please immediately reply to the sender that you have received this communication in error and then delete it.

Esta mensagem e seus anexos se dirigem exclusivamente ao seu destinatário, pode conter informação privilegiada ou confidencial e é para uso exclusivo da pessoa ou entidade de destino. Se não é vossa senhoria o destinatário indicado, fica notificado de que a leitura, utilização, divulgação e/ou cópia sem autorização pode estar proibida em virtude da legislação vigente. Se recebeu esta mensagem por erro, rogamos-lhe que nos o comunique imediatamente por esta mesma via e proceda a sua destruição