[ippm] Comments on draft-tempia-ippm-p3m-02

["Alexander Clemm (alex)" <[e-mail]>]

Hi,

I have a few comments on draft-tempia-ippm-p3m-02.  I find this a very relevant draft that I hope will get adopted by the the WG.  That said, I do have a couple of comments and suggestions:


-          It would be a good idea to establish some terminology, and add a corresponding terminology section.  For example, we need a term for the "flow blocks" in which a particular marking applies, for the data records that are generated for each flow block, for the coloring intervals and block duration, etc.

-          Section 2:  It would be useful to indicate what you do with the colored "blocks" respectively the measurements associated with them, and what the potential interoperability issues are: for example, are you looking at a standard way of controlling the marking, or a standard way of exporting the measures being taken?  Even if you consider this outside the scope of the document, it would be good to still mention that (even if just to state this is out of scope).

-          It would be good to add a paragraph specifically on the problem of how to determine when a block is terminated, probably around section 3.1.  It would be good to add a little further discussion on assumptions choices, most specifically the lower bound that an interval can have for measurements to still be accurate.  Presumably that would be something like 2*max jitter that can occur (to account for the possibilility of laggards).

-          It would be useful to add an additional diagram, showing the deployment of how/where you perform the measurements (i.e. where the packets are captured, time stamped, etc)

-          It would be useful to add a section at the end regarding interoperability considerations, per the earlier comment, e.g. the control of the markings, the counting, the exporting of the records.  This can be accompanied by an explanation that this is out of scope, but mentioning this would be good.

-          The draft is written as too Telecom Italia specific.  This needs to be framed as a general-purpose problem that is of broader interest across the industry.  There are various associated edits that should be made, perhaps most importantly in the abstract and the introduction.    If you would want to refer to Telecom Italia specifics, you might consider simply referencing a publication (in a conference, journal or such) in the reference section instead.

Kind regards
--- Alex