Re: [Detnet] Initial DP split documents available

[Lou Berger <[e-mail]>]

On 4/23/2018 10:43 AM, Andrew G. Malis wrote:
> Lou,
>
> I prefer option (b). The reason is that the IP flows sourced from an 
> IP node are already covered in section 6 of 
> draft-ietf-detnet-dp-sol-ip-00,
Humm, I just looked at this draft a bit more.

(as chair), based on the latest direction discussed in the WG - I think 
section 6 should be come the meat of this document and *all* references 
to MPLS in the earlier sections need to be removed.

> while section 7 of that draft is the proper place for the draft-malis 
> text, and already contains a placeholder for it. It will indeed 
> reference the MPLS draft for the definition of the S-Label, Control 
> Word, etc.

(as ??? - perhaps all hats) I really don't think MPLS has a place in 
this document.  Just like RFC 791 doesn't define MPLS or XYZ over IP...

Lou

>
> Cheers,
> Andy
>
>
> On Mon, Apr 23, 2018 at 10:11 AM, Lou Berger <[e-mail] 
> <mailto:[e-mail]>> wrote:
>
>     Andy,
>
>     One comment below.
>
>     On 4/23/2018 9:36 AM, Andrew G. Malis wrote:
>
>
>         2. The text from draft-malis-detnet-ip-dp-00 has not yet been
>         incorporated into draft-ietf-detnet-dp-sol-ip-00. This can
>         happen in the next revision.
>
>
>     This draft really is about carrying MPLS over an IP PSN and not
>     native IP flows sourced from any IP node, including hosts, right?
>     As such I think this falls in the category of an interworking
>     function and we should decide if it belongs in (a) the MPLS
>     solution doc and references the IP solution doc, (b) in the IP
>     solution doc and references the MPLS solution doc or (c) in it's
>     own document which references both solutions docs.  My inclination
>     is either (a) as IP only devices won't care about MPLS,  or (c) as
>     then we have an RFC that makes it clear that a conformant
>     implementation supports the IWF.
>
>     What do you/others think?
>
>     Lou
>
>