Re: [Actn] ACTN progress

[윤빈영 <[e-mail]>]

Hi Daniele,


Your answer makes me rethink about the 2-level model.

I think that ACTN needs to cover the 2-level model CONCEPTUALLY as well as the 3-level model to provide for a complete transport SDN solution, keeping up the PNC/MDSC/CNC architecture. Looking back the discussion on the PNC/MDSC interface and transparency, I think that we need to clearly define the relationship between the SDN controller(widely accepted among other groups) and the PNC/MDSC, which ,in my opinion, caused the debate on the recursive interface issue. The following is my view on the definition and a way to support the 2-level model based on the 3-level model.



It is generally said that the SDN controller should have all of the four functions the same as the MDSC. However, ACTN defines that the PNC has only two functions, so that it is not sufficient to call it a complete SDN controller. So the PNC are needed to combine with the MDSC to implement a complete T-SDN controller based on physical transport networks to support the four functions. We also need to define another transport controller running over virtual transport networks for the recursive SDN controller model. So we can define two types of a transport SDN controller to support all the four functions as follows.



-      T-SDN Controller(based on physical networks): Non-recursive =  PNC + MDSC

-      T-SDN Controller(based on virtual networks): Recursive =  MDSC



The MDSC can interface with a single PNC or multiple PNCs. For the single PNC, the multi-domain coordination function of the MDSC is disabled, three functions being kept.

Regareds,
Bin

________________________________
보낸 사람 : "Daniele Ceccarelli" <[e-mail]>
보낸 날짜 : 2015-01-18 18:45:36 ( +09:00 )
받는 사람 : Arashmid Akhavain <[e-mail]>, 윤빈영 <[e-mail]>, Leeyoung <[e-mail]>, Igor Bryskin <[e-mail]>, [e-mail] <[e-mail]>
참조 :
제목 : RE: [Actn] ACTN progress

Hi Arashmid, all,

Up to now we have always considered the 3-level model but it is possible to say that is also include the 2-level model. From an implementation point of view we can have a CNC interfacing with a single controller which implements MDSC and PNC functionalities (basically able to speak with CNC and with nodes at the same time).
I don’t see any issue in allowing for a controller to be both PNC and MDSC at the same time.

BR
Daniele

From: ACTN [mailto:[e-mail]] On Behalf Of Arashmid Akhavain
Sent: venerdì 16 gennaio 2015 19:55
To: 윤빈영; Leeyoung; Igor Bryskin; [e-mail]
Subject: Re: [Actn] ACTN progress

I don’t think Young’s comment meant to collapse the 3-level model to a 2-level model. The interface design will be the same. But I think that is different from saying that CNC can bypass the MDSC and talk to PNC directly.

Regards,
Arashmid


From:ACTN [mailto:[e-mail]] On Behalf Of ???
Sent: 16 January 2015 03:23
To: Leeyoung; Igor Bryskin; [e-mail]<mailto:[e-mail]>
Subject: Re: [Actn] ACTN progress

Hi Young,

You newly define the term of “transparency (direct talk between CNC and PNC) ”.
It makes sense that the CNC directly interfaces with the PNC.
To me, it is like that MDSC is not used any more for this case.
So we can define two types of ACTN control hierarchy as follows.

- Three-level model: CNC-MDSC-PNC(current model)
– Two-level model: CNC-PNC(for the case that MDSC is transparent)

If it is right, the framework document should be modified to support the two-level model in terms of functionalities and architecture, as it currently describes only the three-level model.
In the current framework document, the PNC of the three-level model supports only two functionalities including “multi domain coordination function” and “virtualization/abstraction function”.
However, the PNC of the two-level model needs to support four functionalities for the direct interface to the CNC, as the CNC additionally requires “customer VN mapping function” and “virtual service coordination function”. Therefore, the document should be in line with the two-level model.

Correct me if I am wrong.

Sincerely yours

From: Leeyoung [mailto:[e-mail]]
Sent: Friday, January 16, 2015 2:00 AM
To: Igor Bryskin; 윤빈영; [e-mail]<mailto:[e-mail]>
Subject: Re: [Actn] ACTN progress

Hi Bin and Igor,

I agree with Igor on his short reply to Bin. Please see inline for my comment, Bin.

Thanks,
Young

From: Igor Bryskin [mailto:[e-mail]]
Sent: Thursday, January 15, 2015 7:54 AM
To: 윤빈영; Leeyoung; [e-mail]<mailto:[e-mail]>
Subject: RE: [Actn] ACTN progress

Hi Bin,
Your understanding is correct. Conceptually, the interfaces are the same. Practically, a given interface may support only a sub-set of functionality.

Igor

From: ACTN [mailto:[e-mail]] On Behalf Of ???
Sent: Wednesday, January 14, 2015 11:51 PM
To: Leeyoung; [e-mail]<mailto:[e-mail]>
Subject: Re: [Actn] ACTN progress


Hi Young,



I like to catch up with you.

So I like to translate your saying into my words for my clear understanding as follows.



For future ACTN work, we will keep the same architecture consisting of CNC-MDSC-PNC,

and the interfaces between them should be the same as Igor insists.

Each controller has the same functions conceptually,

but some functions may not be used depending on the type of controller(MDSC/PNC).



YOUNG>> Yes.



In order to implement the models(Provisioning, OAM, etc.) required for transport SDN,

ACTN will describe the interface of each controller based on the architecture of CNC-MDSC-PNC.

The interfaces between the controllers may be different practically due to some functions not used.



YOUNG>> Yes, but it would great if we can design the interfaces transparently so that one interface design will accommodate all functions.



We can say like the following if those above are true.

Some can implement a super(general) transport SDN controller to support all the functions,

while other can implement specific purposed(targeted) controllers such as PNC and MDSC.

For the former, the CMI is the same as the MPI.

For the latter, the CMI may not be the same as the MPI due to some functions not used.

ACTN will follow the latter for T-SDN implementation incorporating existing control systems like GMPLS and EMS/NMS.



YOUNG>> For the latter, we can design the MDSC-PNC interface as if CNC talking to PNC(s) directly, (which is termed as ‘transparency’) so that one interface design would fit to all cases. There may be some functions not used, which is fine, but we would say it is one interface design.



Correct me if there is any misunderstanding.



Thanks,

Bin

From:Leeyoung [mailto:[e-mail]]
Sent: Thursday, January 15, 2015 4:12 AM
To: [e-mail]<mailto:[e-mail]>
Subject: [Actn] ACTN progress

Hi All,

Just wanted to share some private emails exchanged among a limited interested parties in the past week with a permission with Igor.

In a nutshell, I think we are converging with a common view on ACTN interfaces and architecture. Please check the following email thread. Please comment if you have any question.

Thanks,
Young

From:Igor Bryskin [mailto:[e-mail]]
Sent: Thursday, January 08, 2015 2:13 PM
To: Leeyoung; Daniele Ceccarelli
Cc: AshwoodsmithPeter
Subject: RE: Recap this morning's call

Hi Young,
You’ve captured all pretty accurately. I nominate you for the ACTN scribe job :=)

Igor

From: Leeyoung [mailto:[e-mail]]
Sent: Thursday, January 08, 2015 3:09 PM
To: Igor Bryskin; Daniele Ceccarelli
Cc: AshwoodsmithPeter
Subject: Recap this morning's call

Hi Igor,

Thanks for this morning’s call. I just wanted to recap what was agreed upon. Please feel free to correct if anything needs to be corrected.

We have identified several models to be implemented in ACTN control hierarchy (CNC-MDSC-PNC). Among them are, but not limited to:


1.      Topology Model

2.      Provisioning Model

3.      Service Model

4.      OAM Model

5.      Client Mapping/Policy Model

There could be more to the list. We can identify and add later on if we will.

Now, with these models, different components (i.e., different controller type) will implement whatever relevant models and support its interfaces. For instance, PNC may need not support Client mapping/policy model while CNC may not need to support Provisioning model.

We also agree that MDSC can interface another MDSC or PNCs with transparency. MDSC should not be able to distinguish if it interfaces MDSC or PNCs (i.e., the same interface).

And you also said, CNC interfacing MDSC should be essentially the same interface as MDSC interfacing another MDSC or PNCs.

Let us start from here. Then we can discuss further.

Thanks,
Young