Re: [IPFIX] [IANA #579482] General Request for Assignment (ipfix)

Andrew,

>> 361 and 362 are port numbers. Notice that the semantics for all the 
>> other port elements (sourceTransportPort, destinationTransportPort, 
>> udpSourcePort, udpDestinationPort, tcpSourcePort, tcpDestinationPort, 
>> collectorTransportPort, exporterTransportPort, 
>> postNAPTSourceTransportPort, and postNAPTDestinationTransportPort) 
>> are all "identifier", since the value identifies a port.
>
> I agree that 361 and 362 are port numbers which are identifiers.  My 
> thinking, however, was that none of these are really identifiers 
> individually.  Information Element Ids are identifiers too, but 
> informationElementRangeBegin (342) and informationElementRangeEnd 
> (343) have semantics of 'quantity' (which is neither 'default' nor 
> 'identifier'.  :-)


5101 says:


        3.2.1 <http://tools.ietf.org/html/rfc5102#section-3.2.1>. quantity

    A quantity value represents a discrete 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.*If no semantic qualifier is given, the
    Information Elements that have an integral data type should behave as
    a quantity.*


        3.2.4 <http://tools.ietf.org/html/rfc5102#section-3.2.4>. identifier

    An integral value that serves as an identifier.  Specifically,
    *mathematical operations on two identifiers (aside from the equality
    operation) are meaningless.*   For example, Autonomous System ID 1 *
    Autonomous System ID 2 is meaningless.


(My bolding).

So, per 3.2.1, the default semantic is "quantity". However, since 
mathematical operations on port numbers are meaningless (other than 
equality and comparisons), my feeling is that "identifier" is more 
appropriate.

Thus the clue is in the definition, "The port number *identifying* the 
start of a range of ports." -> it's a number which identifies a port, 
rather than the value of some physical property of the port.

However, if the WG disagrees then I'm not strongly opposed to changing it.


> Should informationElementRangeBegin and informationElementRangeEnd 
> have semantics of 'default'?

Good question. It seems there should be consistency here.


> Now that I am typing this I think we covered some of this ground 
> before in this thread 
> (https://www.ietf.org/mail-archive/web/ipfix/current/msg06269.html).

What was the outcome of that thread? I found, "lets discuss this in 
Paris". However, I don't recall it being on the WG agenda.


> I don't yet have a strong opinion on what the right answer is.  I just 
> got curious while I was reviewing information elemnts that had been 
> added since the last time I pulled the list.

How often do you update? I have multiple tickets open with IANA. I'll 
try to announce to the WG list when I request new fields. However, note 
that WG discussion / consensus isn't part of the 5102 / expert review 
process.

P.

_______________________________________________
IPFIX mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ipfix