Re: [Teas] Some comments on draft-ali-teas-spring-ns-building-blocks

Hi PSF,

See inline:

> -----Original Message-----
> From: peng.shaofu@zte.com.cn [mailto:peng.shaofu@zte.com.cn]
> Sent: Tuesday, August 3, 2021 6:49 PM
> To: Huzhibo <huzhibo@huawei.com>
> Cc: Dongjie (Jimmy) <jie.dong@huawei.com>; teas@ietf.org;
> draft-ali-teas-spring-ns-building-blocks@ietf.org
> Subject: Re:[Teas] Some comments on draft-ali-teas-spring-ns-building-blocks
> 
> Hi Zhibo,
> 
> See in-line [PSF].
> 
> Regards,
> PSF
> 
> 
> 
> ------------------原始邮件------------------
> 发件人：Huzhibo
> 收件人：彭少富10053815;
> 抄送人：Dongjie
> (Jimmy);teas@ietf.org;draft-ali-teas-spring-ns-building-blocks@ietf.org;
> 日 期 ：2021年08月03日 17:20
> 主 题 ：RE: [Teas]  Some comments on draft-ali-teas-spring-ns-building-blocks
> Hi PSF:
> 
> Both the VTN and AII identify the underlay of the slice.
> [PSF] that is right.
> 
> 
> However, the VTN-ID is not used as an identifier of the control plane topology.
> Instead, the topology of the slice is defined by referencing an existing
> multi-topology technology.
> VTN-ID defines the topology and resources of a slice by decoupling the
> topology and resources, which are different from AII.
> [PSF] Whether it is customizing slice's own topology (such as AII) or referencing
> an existing topology (such as VTN-ID), there is a logical topology for slice. So
> VTN-ID is also actually an identifier of logical topology. And, both AII and
> VTN-ID are used to identify resouce. So, the smallest difference for AII and
> VTN-ID is the flavor to create the logical topology. Note that
> draft-bestbar-teas-ns-packet contains both customizing and referencing flavor.

[Jie] The key difference is that AII reinvents the wheel by defining a new topology identifier, while VTN refers to the well-defined MT or Flex-Algo for its topology attribute. This is why VTN allows the decoupling of topology and resource attributes to provide better scalability, while AII has limitation in the supported network scale. 

> As mentioned in the previous email, the agreement is to find a common new
> term for network slice realization.
> [PSF] Agree, alignment with some new term, but not VTN according to
> Dongjie's original mail.

[Jie] Apparently you didn't read my mail carefully. Here is what I suggested in my email:

> 1. As discussed on the TEAS meeting, the terminologies in several drafts
> related to network slicing realization will be aligned, thus it is suggested the
> terminology in this document also aligns with the "new term"  to be
> proposed.

-Jie

> Best regards,
> Zhibo
> -----Original Message-----
> From: Teas [mailto:teas-bounces@ietf.org] On Behalf Of
> peng.shaofu@zte.com.cn
> Sent: Tuesday, August 3, 2021 12:37 PM
> To: Huzhibo <huzhibo@huawei.com>
> Cc: Dongjie (Jimmy) <jie.dong@huawei.com>; teas@ietf.org;
> draft-ali-teas-spring-ns-building-blocks@ietf.org
> Subject: Re: [Teas] Some comments on draft-ali-teas-spring-ns-building-blocks
> 
> Hi Zhibo,
> 
> First, when aii was proposed in draft-peng-teas-network-slicing to introduce
> slice identifier to underlay network, it was not good just to change its name as
> vtn-id in redundant draft, for example, just open
> https://www.ietf.org/archive/id/draft-zch-lsr-isis-network-slicing-06.txt and
> https://www.ietf.org/archive/id/draft-dong-lsr-sr-enhanced-vpn-06.txt to see
> the repeat and overlap.
> Second, SA-ID defined in draft-bestbar-teas-ns-packet has completely covered
> aii and vtn-id, especially when the mapping of slice and Slice-Aggregate is 1:1.
> So that there is not any reason for you to continue to update vtn-id concept,
> and even ask other proposals to be alignment with VTN.
> 
> Regards,
> PSF
> 
> ------------------原始邮件------------------
> 发件人：Huzhibo
> 收件人：彭少富10053815;Dongjie (Jimmy);
> 抄送人：draft-ali-teas-spring-ns-building-blocks@ietf.org;teas@ietf.org;
> 日 期 ：2021年08月02日 20:21
> 主 题 ：RE: Re:[Teas] Some comments on
> draft-ali-teas-spring-ns-building-blocks
> Hi PSF,
> The discussion Jie referred to was the about the terminology alignment
> between VTN, Slice Aggregate, etc. in several active network slice related
> drafts. With the "merging" of draft-peng-teas-network-slicing into
> draft-bestbar-teas-ns-packet, it is not clear whether the term "AII" has been
> totally replaced by "Slice Aggregate", or you still plan to work on it separately?
> After the discussion among the authors of several drafts, the agreement is to
> find a common new term for network slice realization, so that those drafts
> could refer to that new term. This is also what we suggested to the authors of
> draft-ali-teas-spring-ns-building-blocks for better terminology alignment.
> Best regards,
> Zhibo
> -----Original Message-----
> From: peng.shaofu@zte.com.cn [mailto:peng.shaofu@zte.com.cn]
> Sent: Monday, August 2, 2021 5:39 PM
> To: Dongjie (Jimmy) <jie.dong@huawei.com>
> Cc: draft-ali-teas-spring-ns-building-blocks@ietf.org; teas@ietf.org
> Subject: Re:[Teas] Some comments on draft-ali-teas-spring-ns-building-blocks
> Hi Dongjie,
> Seeing that you openly mention VTN-ID, I have to remind you again that the
> VTN-ID is just the A-I-I of draft-peng-teas-network-slicing which analyzes the
> necessity of introducing slice identifier into underlay network. I believe you
> never thought about dealing with the overlap of VTN-ID and A-I-I, so how can
> you ask other drafts to do so?
> Again, you may say that the A-I-I related scheme described by
> draft-peng-teas-network-slicing has scalability problems ... ... so that VTN-ID is
> not A-I-I... ...
> Regards,
> PSF
> ------------------原始邮件------------------
> 发件人：Dongjie(Jimmy)
> 收件人：draft-ali-teas-spring-ns-building-blocks@ietf.org;
> 抄送人：TEAS WG;
> 日 期 ：2021年08月02日 17:07
> 主 题 ：[Teas] Some comments on draft-ali-teas-spring-ns-building-blocks
> _______________________________________________
> Teas mailing list
> Teas@ietf.org
> https://www.ietf.org/mailman/listinfo/teas
> Hi authors,
> Thanks for the presentation in the TEAS session in last week. Here are some
> comments on this draft:
> 1.      As discussed on the TEAS meeting, the terminologies in several drafts
> related to network slicing realization will be aligned, thus it is suggested the
> terminology in this document also aligns with the "new term"  to be
> proposed.
> 2.      As this document lists the SR technologies which can be used for
> network slice realization in SR networks, the suggestion is it should also
> describe and reference draft-ietf-spring-resource-aware-segments and
> draft-ietf-spring-sr-for-enhanced-vpn which specifies the extensions to SR
> segments and the mechanisms to provide SR based VTNs.
> 3.      Section 8 of this document describes the "stateless network slice ID"
> concept and the mechanisms with different data planes. The general
> mechanism of introducing dedicated VTN identifier in data packet for per-VTN
> packet processing was described in
> draft-dong-teas-enhanced-vpn-vtn-scalability, thus there is some overlap in
> this part which needs to be solved in future versions.
> Best regards,
> Jie