Re: [DMM] Call for adoption of draft-clt-dmm-tn-aware-mobility-08 as a WG document - Objections

[Kaippallimalil John <[e-mail]>]

Joel,
This draft is not attempting to lay out a new architecture. The figure and architecture section are there to provide context for the reader.

The draft is also not asserting that these are the only ways to solve the issue. 
The last portion of the draft only refers to applicability since several IETF technology (and even non IETF data plane like L2) may be used E2E in the transport underlay corresponding to a 3GPP overlay (F1/W1, N3, N9). 
The draft is agnostic to any specific underlay. It is concerned with how the slice type/QoS properties between 3GPP provider and subscriber (UE) is realized as the data plane GTP packets (overlay) traverse one or more transport underlay segments on path.

The dmm working group seems to be a natural choice as folks there have background and expertise in both IETF and 3GPP technologies.  

Best Regards,
John


-----Original Message-----
From: dmm <[e-mail]> On Behalf Of Joel M. Halpern
Sent: Tuesday, January 05, 2021 6:10 PM
To: dmm <[e-mail]>
Subject: Re: [DMM] Call for adoption of draft-clt-dmm-tn-aware-mobility-08 as a WG document - Objections

Looking agin at the document, I have multiple problems with adoption of this document.

First, as just discussed, it seems to be laying out an archtiecture, but that architecture is not consistent with the other work going on in the IETF (specifically in the routing area, and known to many of the co-authors of this work.)

Second, the document states that it is Informational, and then is written as if it is defining the one and only way to structure the needed system.

Third, related to the second, this document seems to assert that there is a specific and correct way to use IETF technology to solve the problem.  Then, about three quarters of the way through it remembers that there are multiple alternatives for some parts.  The IETF rarely writes implementation specifications.  There are other SDOs that do so. 
  They can wrestle with the problem of defining something that can be utilized by multiple operators with varying constraints.  Why would we want to get into trying to have that fight?

For the WG chairs, I am trying to figure out how this even fits the DMM charter.  It is a LOT more than network exposure, which is the closest I can see in the charter.

Yours, unhappily,
Joel



_______________________________________________
dmm mailing list
[e-mail]
https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.ietf.org%2Fmailman%2Flistinfo%2Fdmm&amp;data=04%7C01%7Cjohn.kaippallimalil%40futurewei.com%7C3d167f9186e74cfca90008d8b1d78d79%7C0fee8ff2a3b240189c753a1d5591fedc%7C1%7C0%7C637454886651720444%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&amp;sdata=sTvL%2Bu5b7qLIC%2BAmJ%2FvaIgToUMLUdY4bWCcZBlFn9Kw%3D&amp;reserved=0