Re: [Teas] Moving forward with draft-ietf-teas-ietf-network-slices

[Adrian Farrel <[e-mail]>]

Hi Kenichi,

I'm cleaning up the comments to produce a new revision.

>I agree with Loa, moving the definition of how the term "customer" is interpreted in
> the document to the Terms and Abbreviations section seems very reasonable.
> +1

OK. I'm doing this.

> I don't stick to "customer", but a discussion how about rephrasing "provider" with
> "producer" may also rise when I choose non-ietf-standard "consumer" term

Two things:
- Yes, the discussion of alternatives to "provider" is also still open
- I am keeping a list of people who feel "strongly" about customer/consumer/client
  At the moment I don't feel able to draw any conclusions. I don't think this will 
  change the text on the definition, and we can change the name later if we can
  reach agreement.

> Related to this, should we rephrase " IETF Network Slice user" if there is any
> intention to leave?

Good catch.

> the expectation currently is that the IETF Network Slice Service will use
> something along the lines of the service models (L2SM and L3SM).  It seems
> reasonable to enhance those with something about service functions, quite
> possibly along the lines you are already working on.

Yes. I hope to bring some text on "slice as a service" for discussion SOON.

> [TS] (snip) Alternatively, does it help if the NBI request expresses such 
> inclusions/exclusions policy using/referencing a  customer-centric 
> topology that a provider may have furnished earlier to the customer,
> and which the provider can readily map to their
>
> FYI, L3SM tried to let the customer define such policy against PE site in
> sec. 6.6.
> I know this is non-exhaustive for IETF Network Slice.

Thanks for the pointer to L3SM.

> Editorials:
> - I'm not sure if recent RFCs should refer SNMP as an NBI in the sense of
>   https://www.ietf.org/about/groups/iesg/statements/writable-mib-module/ .

I agree with you. I'll remove SNMP from the list in 5.2 unless someone shouts (see separate email).

> - multipoint-to-point twice
> .2.  IETF Network Slice Endpoints
> As noted in Section 3.1, an IETF Network Slice describes connectivity between multiple
> endpoints across the underlying network.  These connectivity types are: point-to-point,
> point-to-multipoint,  multipoint-to-point, multipoint-to-point, or multipoint-to-multipoint.

Thanks.

Cheers,
Adrian