Re: [I2nsf] Yangdoctors last call review of draft-ietf-i2nsf-consumer-facing-interface-dm-07

[tom petch <[e-mail]>]

On 06/09/2020 12:26, Mr. Jaehoon Paul Jeong wrote:
> Hi Tom,
> I have reflected your two comments in the revision:
> https://tools.ietf.org/html/draft-ietf-i2nsf-consumer-facing-interface-dm-11
>
> Please see my answers inline below.

Paul

I like the prefix:-)

With the type 'time' I had something rather more complex in mind and 
think now that it is rather too complex for this situation. (Namely that 
you can import a specific revision of a module so that you can import 
the types from 6991-bis rather than 6991 so you then get 'time' 
automatically and do not have to define anything but it does introduce 
side effects which I now think outweigh the benefits)

So, I would revert to what you had and just define 'time' as you do at 
present and omit the references to 6991-bis.

Tom Petch


> On Sat, Aug 29, 2020 at 1:31 AM tom petch <[e-mail]> wrote:
>
>> On 28/08/2020 14:46, Mr. Jaehoon Paul Jeong wrote:
>>> Hi Tom,
>>> I have addressed all your comments in the following revision:
>>>
>> https://tools.ietf.org/html/draft-ietf-i2nsf-consumer-facing-interface-dm-10
>>>
>>
>> Inline
>>
>>
>>> Here are my answers for your comments:
>>>
>> ------------------------------------------------------------------------------------
>>> Some more minor tweaks
>>> s.5.1 /gorup/group/
>>> => The replacement is done.
>>>
>>> YANG module
>>>
>>> WG Chairs are not usually listed in the module - they used to be
>>> => The information of WG Chairs is removed.
>>>
>>> description is a bit terse - some quote the Abstract
>>> => I have improved the descriptions in the YANG module.
>>>
>>> YARA, SNORT, SURICATA would benefit from references; they are not ones I
>>> see in TLS or SSH!
>>> => I have added the references to YARA, SNORT, and SURICATA.
>>>
>>> typedef time I see in RFC6991bis
>>> => I used typedef time in RFC6991bis.
>>
>> See my other note about importing from 6991-bis rather than from 6991
>>
>
>   => I replaced 6991 with 6991-bis.
>
>>>
>>> does the ipv6 addresss ever need the interface?
>>> => Yes, the IPv6 address needs the CFI interface.
>>>      I added an XML example using IPv6 addresses.
>>>
>>> start/end ipv4/ipv6 could do with a must end > start
>>> => I put a description that an IPv4/IPv6 start address is lower than
>>>      an IPv4/IPv6 end address.
>>>
>>>      "A range match for IPv4 addresses is provided.  Note that the
>>>       start IPv4 address must be lower than the end IPv4 address.";
>>>
>>>      "A range match for IPv6 addresses is provided.  Note that the
>>>       start IPv6 address must be lower than the end IPv4 address.";
>>>
>>> geo-ip could do with a reference
>>> => I added a reference to geo-ip as follows.
>>>      RFC8805: A Format for Self-Published IP Geolocation Feeds
>>>
>>> s.9.1 221.159 is not a documentation address - see RFC5737
>>> => I used documentation addresses for IPv4 from RFC5737.
>>>      I also used documentation addresses for IPv6 from RFC3849.
>>
>> good
>>
>>
>>> IESG often expect an ipv6 example alongside ipv4
>>> => I added an XML example using IPv6 addresses in Figure 19.
>>>
>>> s.12 Registrant should be IESG
>>> => I modified the IANA considerations section such that Registrant is the
>>> IESG.
>>>
>>> prefix is not that of the module
>>> => I am not sure of this comment. I think we use the correct prefix of
>>> "cfi-policy".
>>>      CFI stands for Consumer-Facing Interface.
>>
>> Yes indeed you do - my mistake.  What I had intended to say, looking at
>> other NSF modules, was that there are a number of NSF modules and the
>> chosen prefix have nothing in common.  Bear in mind that all the YANG
>> modules get mixed up together on the box so while the prefix need to be
>> compact, there is something to be said for them to be meaningful so
>> RTGWG modules could start rt... or MPLS ones mpls... or PCE ones pce..
>> and so on so you could consider using a prefix of nsf... such as nsfcfi
>> or if there are several such nsfcfi-p or some such (but that is getting
>> a bit long)
>>
>
>   => I used nsfcfi for the prefix for Consumer-Facing Interface (CFI).
>
>       Thanks.
>
>       Best Regards,
>       Paul
>>
>> Tom Petch
>>
>>>
>> ------------------------------------------------------------------------------------
>>>
>>> Thanks for your valuable comments.
>>>
>>> Best Regards,
>>> Paul
>>>
>>> On Fri, Jul 17, 2020 at 8:07 PM tom petch <[e-mail]>
>> wrote:
>>>
>>>> On 17/07/2020 10:51, tom petch wrote:
>>>>> On 11/07/2020 08:44, Mr. Jaehoon Paul Jeong wrote:
>>>>>> Hi Jan and Tom,
>>>>>> I have revised our I2NSF Consumer-Facing Interface (CFI) Data Model
>>>> Draft
>>>>>> according to both your comments.
>>>>>>
<snip>