[ippm] AD Review #2 of RFC8321bis

[Martin Duke <[e-mail]>]

Hi Giuseppe,

This is looking good.

1) The one conceptual problem I'm having is what it means for a packet to
be "unmarked". If marking involves encapsulation, no problem: some packets
can simply not be encapsulated, or one can omit the header option or
whatever. But if it's passive measurement because there's a bit in the
header, this bit has to be set to something, right?  So how would one
mediate between flows that are to be measured or not measured in the latter
case?

2) Some minor rewording of the new fragmentation language in section 4.4:
OLD

Nodes that fragment packets within the measurement domain MUST NOT
replicate marks, but SHOULD mark the first fragment if they have the

capability to do so.


NEW

Nodes that fragment packets within the measurement domain SHOULD, if

they have the capability to do so, ensure that only one resulting

fragment carries the marking bit(s) of the original packet. Failure

to do so can introduce errors into the measurement.


I would propose the following steps.

A) address these points and publish draft-02. We'll save the diff to show
the delta between 8321 and 8321 bis.

B) Immediately reorganize the document as you intend and publish draft-03.

Then we'll take it through IETF Last Call.

Thanks
Martin