Re: [Anima] I-D Action: draft-ietf-anima-grasp-distribution-07.txt

[Michael Richardson <[e-mail]>]

Hi, I recalled that I'm the document shepherd for this document.
I did my shepherd review a year ago on version -04, and I also did a review
in 2020 when we adopted it:
  https://mailarchive.ietf.org/arch/msg/anima/hCv5bZxBrzSzA6BjA5DY7_RDo6U/

I am updating my shepherd review, but let me highlight some comments from
January 2022 review, and my January 2020 comments that have still not been
acted upon.

I think that the use cases are not well supported.
I can not comment on whether SBA is a reasonable use case, if it were, then I
would expect a 3GPP document to cite this one.  Has that occured?

I do not find the INC use case credible.
I can see that yes, the INC needs data backup and data aggregation, but I
don't see why they would do that via the ACP.  I can see that perhaps there
is an ASA that would run over an ACP that would provision where computing is
supposed to go, but the actual movement of the data seems to belong in the
production network.  Then the data might move by rsync, etc.

I don't find the V2X communications use credible either.
This is because I don't think V2X will deploy an ACP, or for that matter
*can* deploy such a thing as it would cross the boundaries of authorities.

The thing that I **do** find credible is software updates, particularly within
Enterprises and ISPs for their core switching equipment, but also for IoT
devices that might be connected at the edge of the switching network.
The firmware updates can be transfered across a high-speed network to the
edge where slow/sleepy IoT devices can pick them up via one-hop
communications.

I did not find the Smart Home use case credible.
Amazon, Google and Apple have invested in CSA/MATTER, and it has it's own
communication fabric.  That spec does not cite this spec, so I don't see how
this helps.


--
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-                      *I*LIKE*TRAINS*