Re: [Teas] [Teas-ns-dt] New Version Notification for draft-nsdt-teas-ietf-network-slice-definition-00.txt

[Adrian Farrel <[e-mail]>]

I hope this debate is not dancing on the head of a pin, but it does seem to
be quite fundamental to what we are trying to achieve.

On the one hand, a network contains many "resources" such as storage,
compute, virtual functions, and access to content. If the process of slicing
is to 'share out' the network resources by building "virtual networks," then
those virtual networks surely need to include all the sorts of things that a
physical network includes.

On the other hand, the primary need is clearly connectivity. Without
connectivity we have nothing.

Maybe we have two concepts:
- IETF network slicing
- IETF connectivity network slicing

The second of these is a subset of the first.
The second is what Kiran is seeking to discuss in this draft (I think,
reading her emails and having talked with her).
The first is, I think, what Med is talking about.

Notwithstanding, all of this, Kiran's approach is the right one for this
draft at the moment. That is, fix the most inflammatory issues first; pick
up a pile of nits along the way; reduce the text to only that which is
necessary (work in progress); and then come up with something we can debate
a bit more clearly.

Cheers,
Adrian

-----Original Message-----
From: Teas-ns-dt <[e-mail]> On Behalf Of Kiran Makhijani
Sent: 28 October 2020 15:33
To: [e-mail]; [e-mail]; TEAS WG
<[e-mail]>
Subject: Re: [Teas-ns-dt] New Version Notification for
draft-nsdt-teas-ietf-network-slice-definition-00.txt

Hi Med, 
Well! while working on this major revision authors did not feel that primary
scope need to be adjusted. Although a lot of re-write has happened.
Deflating the tension with naming and addressing major comments  were the
first steps. 

Our proposed definition has 2 parts, first one is connectivity (because we
concern with network specific characteristics) and second is meeting service
specific requirements, which allows consumer to specify its
application/service needs. One could extend those objectives if they can be
well-described through NBI and are feasible in the realized network.

I still do not know what is that "something else" you are alluding to?
Terminology document should be very clear so it will be good to know what
other things you have in mind that have been overlooked.

You are also asking whether required attributes to characterize an IETF
network slice can be used beyond slicing. Without clearly knowing what you
want, I see them being usable in the context of end-to-end slice (last
section before security) which is broader than the IETF network slice. 

Thanks
Kiran


> -----Original Message-----
> From: [e-mail] <[e-mail]>
> Sent: Monday, October 26, 2020 2:27 AM
> To: Kiran Makhijani <[e-mail]>; [e-mail]; TEAS WG
> <[e-mail]>
> Subject: RE: New Version Notification for
draft-nsdt-teas-ietf-network-slice-
> definition-00.txt
> 
> Hi Kiran,
> 
> > Let ask naïve questions on (1) Can you explain what's misleading?
> 
> Sure.
> 
> Changing the name without adjusting the scope is what is misleading. The
> actual description is more about connectivity which is exactly what you
had in
> previous versions of the draft but the use of "network slice" suggests
this is not
> exclusively restricted to connectivity but can include "something else".
There is
> a disconnect if you will between the scope as described in the draft and
the
> current name. This discussion is meant to hopefully clarify this.
> 
> > and " **specific** to slicing vs generic ones."?  How do you see
> 
> The question is whether the required attributes to characterize the
connectivity
> part of a slice can be used beyond slicing or there are attributes that
are "tied"
> to slicing. We need to call out these exclusive attributes, if any. This
is
> important from a modelling standpoint.
> 
> > CPP/7297 relates to IETF network slices?
> 
> This depends on the answer to the previous question, but from the current
> draft, the connectivity component of an "IETF Network Slice" can be
expressed
> as a CPP.
> 
> Cheers,
> Med
> 
> > -----Message d'origine-----
> > De : Kiran Makhijani [mailto:[e-mail]] Envoyé : vendredi
> > 23 octobre 2020 19:47 À : BOUCADAIR Mohamed TGI/OLN
> > <[e-mail]>; teas- [e-mail]; TEAS WG
> > <[e-mail]> Objet : RE: New Version Notification for
> > draft-nsdt-teas-ietf- network-slice-definition-00.txt
> >
> > Hi Med,
> > Many thanks for reviewing the updated text.
> >
> > Let ask naïve questions on (1) Can you explain what's misleading?
> > and " **specific** to slicing vs generic ones."?  How do you see
> > CPP/7297 relates to IETF network slices?
> >
> > On (2) and (3): Terminology document is motivated to establish minimal
> > common understanding independently - upon which further work can
> > progress, as well as previous efforts can relate to.
> >
> > With that in mind, there are 2 ways to tie down RFC7297 -  we have
> > applicability section in framework and appropriate text on how CPP
> > applies can go there. Another option is to establish CPP relationship
> > briefly with SLOs in section  4.1.1 to slices with a reference to
> > 7297. In both cases we need your help to provide the right text.
> >
> > But that's secondary. Lets first cover (1).
> > -Kiran
> >
> >
> > > -----Original Message-----
> > > From: [e-mail]
> <[e-mail]>
> > > Sent: Thursday, October 22, 2020 4:41 AM
> > > To: Kiran Makhijani <[e-mail]>; [e-mail];
> > TEAS
> > > WG <[e-mail]>
> > > Subject: RE: New Version Notification for
> > > draft-nsdt-teas-ietf-network-slice-
> > > definition-00.txt
> > >
> > > Hi Kiran, all,
> > >
> > > Thank you for sharing this updated version.
> > >
> > > (1)
> > >
> > > It seems that the scope is still connectivity:
> > >
> > >    An IETF Network Slice is a well-defined structure of
> > connectivity
> > >    requirements and associated network behaviors.  IETF Network
> > Slices
> > >    are defined such that they are independent of the underlying
> > >    infrastructure connectivity and technologies used.  This is to
> > allow
> > >    an IETF Network Slice consumer to describe their network
> > connectivity
> > >    and relevant objectives in a common format, independent of the
> > >    underlying technologies used.
> > >
> > > Which is fine by me but the use of "network slice" is misleading.
> > >
> > > (2)
> > >
> > > I already made this comment during the call for adoption, but I
> > don't
> > > see it addressed in this version: It would be really cool if we
> > can
> > > identify attributes that are **specific** to slicing vs generic
> > ones.
> > > I'm particularly referring to the CPP defined in RFC7297:
> > >
> > > ====
> > >    3.  Connectivity Provisioning Profile (CPP)
> > >      3.1.  Customer Nodes Map
> > >      3.2.  Scope
> > >      3.3.  QoS Guarantees
> > >      3.4.  Availability
> > >      3.5.  Capacity
> > >      3.6.  Conformance Traffic
> > >      3.7.  Overall Traffic Guarantees
> > >      3.8.  Traffic Isolation
> > >      3.9.  Flow Identification
> > >      3.10. Routing and Forwarding
> > >      3.11. Activation Means
> > >      3.12. Invocation Means
> > >      3.13. Notifications
> > > ====
> > >
> > > (3)
> > >
> > > Both clarifications are important to be worked out for the
> > following reasons:
> > > * If the "IETF Network slice" is more than connectivity, then its
> > > connectivity component does not need to signal explicitly this is
> > > about a "slice" because its presence in the "IETF Network slice"
> > is sufficient to infer that.
> > >
> > > * If there are no connectivity-related attributes that are
> > specific to
> > > slicing, then we need to factorize and use a generic modelling for
> > the
> > > connectivity component. For example, an ABNF inspired from RFC7297
> > would look like:
> > >
> > >    <NETWORK_SLICE> ::=
> > >                  <Some_Non_Connectivity_Component> ...
> > >                  <Connectivity Provisioning Component> ...
> > >    <Connectivity Provisioning Component> ::=
> > >                               <CONNECTIVITY_PROVISIONING_PROFILE>
> > ...
> > >    <CONNECTIVITY_PROVISIONING_PROFILE> ::=
> > >                               <Customer Nodes Map>
> > >                               <Scope>
> > >                               <QoS Guarantees>
> > >                               <Availability>
> > >                               <Capacity>
> > >                               <Traffic Isolation>
> > >                               <Conformance Traffic>
> > >                               <Flow Identification>
> > >                               <Overall Traffic Guarantees>
> > >                               <Routing and Forwarding>
> > >                               <Activation Means>
> > >                               <Invocation Means>
> > >                               <Notifications>
> > >                               <Optional Information Element> ...
> > >
> > > Cheers,
> > > Med
> > >
> > > > -----Message d'origine-----
> > > > De : Teas [mailto:[e-mail]] De la part de Kiran
> > > > Makhijani Envoyé : jeudi 22 octobre 2020 08:12 À :
> > > > [e-mail]; TEAS WG <[e-mail]> Objet : [Teas] FW:
> > New
> > > > Version Notification for
> > > > draft-nsdt-teas- ietf-network-slice-definition-00.txt
> > > >
> > > > Hello Teas and teas-ns-dt,
> > > > FYI: Please find new version of  IETF network slices (previously
> > > > called transport slices) definition document.
> > > >
> > > > This is still a work in progress document but several comments
> > and
> > > > feedback received till now have been addressed. We want to share
> > > > updates so far and look forward to further comments and
> > discussion.
> > > >
> > > > Thanks
> > > > -Authors
> > > >
> > >
> 
> 
> ___________________________________________________________________
> ______________________________________________________
> 
> Ce message et ses pieces jointes peuvent contenir des informations
> confidentielles ou privilegiees et ne doivent donc pas etre diffuses,
exploites ou
> copies sans autorisation. Si vous avez recu ce message par erreur,
veuillez le
> signaler a l'expediteur et le detruire ainsi que les pieces jointes. Les
messages
> electroniques etant susceptibles d'alteration, Orange decline toute
> responsabilite si ce message a ete altere, deforme ou falsifie. Merci.
> 
> This message and its attachments may contain confidential or privileged
> information that may be protected by law; they should not be distributed,
used
> or copied without authorisation.
> If you have received this email in error, please notify the sender and
delete this
> message and its attachments.
> As emails may be altered, Orange is not liable for messages that have been
> modified, changed or falsified.
> Thank you.

-- 
Teas-ns-dt mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/teas-ns-dt