Re: [Anima] "virtual out-of-band" ... or some minor non-ACP-number comments on Action: draft-ietf-anima-autonomic-control-plane-25.txt

["Bless, Roland (TM)" <[e-mail]>]

Hi,

Am 30.06.20 um 08:27 schrieb Michael H. Behringer:
> I still prefer the definition "virtual out of band".

Me, too.
Especially, if one reads RFC8368 it clearly makes the
point that the DCN/OAM networks are normally out-of-band,
whereas the GACP is realized as _in-band_ solution.
So GACP and ACP are "virtually out-of-band". I think that
the editorial change Michael Richardson referred to just
fixed that, because "virtually in-band" would not be
correct.

An overlay is a very generic concept (e.g., IP is an overlay
on top of layer 2 networks) and you can stack
them on top of each other nearly infinitely. So overlays
are nearly everywhere and I think it's also clear that ACP
is a establishing a control overlay.

> An "overlay" (secure or not) depends on correct configuration of the
> underlay. The ACP does NOT depend on configuration in the underlay, that
> is what makes it special.
> 
> I haven't seen the definition "virtual out of band" anywhere else, and
> it is the most precise way to describe it.

Regards
 Roland

> Michael
> 
> On 30/06/2020 00:06, Brian E Carpenter wrote:
>> Say "secure overlay" to emphasise the point, but yes.
>>
>> The draft I submitted yesterday "describes a simple method of forming
>> an ACP immediately above the transport layer" which is indeed
>> precisely a secure overlay.
>>
>> Regards
>>     Brian
>>
>> On 30-Jun-20 00:45, William Atwood wrote:
>>> Is "overlay" the right word?
>>>
>>> I agree that it is physically in-band, and virtually out-of-band.  Isn't
>>> that the definition of "overlay"?
>>>
>>>    Bill
>>>
>>> On 2020-06-28 11:02 p.m., Michael Richardson wrote:
>>>> Attention This email originates from outside the concordia.ca
>>>> domain. //
>>>> Ce courriel provient de l'exterieur du domaine de concordia.ca
>>>> On 2020-06-23 10:31 p.m., [e-mail] wrote:
>>>>> A diff from the previous version is available at:
>>>>>
>>>> https://www.ietf.org/rfcdiff?url2=draft-ietf-anima-autonomic-control-plane-25
>>>>
>>>>
>>>>
>>>> yes, I read the diffs :-)
>>>>
>>>> -   This document describes a modular design for a self-forming, self-
>>>> -   managing and self-protecting ACP, which is a virtual in-band
>>>> network
>>>> -   designed to be as independent as possible of configuration,
>>>>
>>>> +   This document describes a modular design for a self-forming, self-
>>>> +   managing and self-protecting ACP, which is a virtual out-of-band
>>>> +   network designed to be as independent as possible of configuration,
>>>>
>>>> This change from being a virtual in-band network to a virtual
>>>> out-of-band network must have been in response to some comments... It
>>>> seems a big change in some ways.  I guess it makes this text consistent
>>>> with the abstract which has said virtual out-of-band for awhile now.
>>>>
>>>> But, I do have to wonder if we are creating confusion by claiming that
>>>> this is an out-of-band mechanism, even though it's really an in-band
>>>> mechanism.  It's just virtually-out.
>>>>
>>>> I actually do want to start a bike-shed issue here?
>>>> Are we describing ourself wrong?  Maybe there is some portmanteau that
>>>> would be more accurate?  I think that the above sentence is essentially
>>>> the elevator pitch for all of ANIMA.
>>>>
>>>>
>>>> There is also a bunch of other text that has been added to the
>>>> Introduction, which I think confuses more than it enlightens.
>>>> Or at least needs a better copy-edit.
>>>>
>>>> A number of other new sections (9.4..) need a copy-edit to fix some
>>>> missing words.  I will try to help Toerless with that via github.