Re: [IPFIX] Full or Truncated EHs RE: Some comments on draft-ietf-opsawg-ipfix-tcpo-v6eh

[Benoit Claise <[e-mail]>]

Hi Paul,

I looked at this one below.

On 10/19/2023 11:02 PM, [e-mail] wrote:
>
>
> 3.2 Data Type Semantics:
>
> - this is not an identifier. It seems be a new type consisting of 
> (type, count) tuples.
>
>
> */[Med] Will double check this one. /*
>
>
>
You are right that ipv6ExtensionHeaderCount consists of (type, count) 
tuples.
I tried too look at similar IPFIX IEs in IANA
The closest one I could find might be anonymizationFlags, which 
categorized as "flags"
So which "Data Type Semanitcs" should be have? Flags?

Then I investigated further..._

_1. https://datatracker.ietf.org/doc/html/rfc7012#section-2.1

        dataTypeSemantics - The_integral _types are qualified by additional
           semantic details.  Valid values for the data type semantics are
           either specified inSection 3.2  of this document or will be
           specified in a future extension of the information model.

2. Looking at IPFIX IANA, I saw as series of integral IPFIX IEs with 
"Data Type Semantics" with no value.
See sourceIPv4PrefixLength

3. I see a single integral IPFIX IE, "srhSegmentIPv6LocatorLength" 
(document 
https://datatracker.ietf.org/doc/draft-ietf-opsawg-ipfix-srv6-srh/14/ in 
AUTH48 right now), with "Data Type Semantics" with "default" value.
That seems wrong to me.

Interestingly, I don't see this "Default" in 
https://datatracker.ietf.org/doc/html/rfc7012#section-3.2, as a valid 
Data Type Semantic.
BUT searching around, I found this: 
https://datatracker.ietf.org/doc/html/rfc5610#section-3.6

                              +-------+--------------+
                              | Value | Description  |
                              +-------+--------------+
                              | 0     | default      |
                              | 1     | quantity     |
                              | 2     | totalCounter |
                              | 3     | deltaCounter |
                              | 4     | identifier   |
                              | 5     | flags        |
                              +-------+--------------+

                              Table 2: IE Semantics Values

See how "default" is different than "quantity", which is inconsistent 
with the following text
https://datatracker.ietf.org/doc/html/rfc7012#section-3.2.1


    3.2.1 <https://datatracker.ietf.org/doc/html/rfc7012#section-3.2.1>.
    quantity

        "quantity" is a numeric (integral or floating point) value
        representing a measured value pertaining to the record.  This is
        distinguished from counters that represent an ongoing measured value
        whose "odometer" reading is captured as part of a given record._This is the default semantic type of all numeric data types._


Question:  Data Type Semantic => is "default" different than "quantity"?
Question: if yes, what's the difference?
Question: if not different, errata in RFC5610?
Question: should srhSegmentIPv6LocatorLength be changed to quantity? I 
think so (remember: doc in AUTH48)

4. Then, coming back to ipv6ExtensionHeaderCount , I wonder, is the 
"Data Type Semantics" optional?
https://datatracker.ietf.org/doc/html/rfc7012#section-3

        Abstract data types unsigned8, unsigned16, unsigned32, unsigned64,
        signed8, signed16, signed32, and signed64 are integral data types.
        As described inSection 3.2  <https://datatracker.ietf.org/doc/html/rfc7012#section-3.2>, their data type semantics_can be further specified_, for example, by 'totalCounter', 'deltaCounter',
        'identifier', or 'flags'.

"can be" => seems optional to me.
However, https://datatracker.ietf.org/doc/html/rfc7012#section-3.2 is 
not clear about it

So proposal for ipv6ExtensionHeaderCount => remove the Data Type Semantic.
At least, that would consistent with the current IANA entries.
If not, 
https://datatracker.ietf.org/doc/html/draft-ietf-opsawg-ipfix-fixes 
might need a long series of updates (I am not searching for more work 
:-) ), to add "default" for integral IPFIX IE with no values.

Regards, Benoit