[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [ippm] I-D ACTION:draft-ietf-ippm-multimetrics-04.txt
Hi Ruediger,
See inline.
> -----Message d'origine-----
> De : Geib, Ruediger [mailto:Ruediger.Geib at t-systems.com]
> Envoyé : vendredi 20 juillet 2007 10:03
> À : STEPHAN Emile RD-CORE-LAN
> Cc : acmorton at att.com; ippm at ietf.org
> Objet : Re: [ippm] I-D ACTION:draft-ietf-ippm-multimetrics-04.txt
>
> Emile,
>
> I just noticed that the 04 version includes text on "Spatial Segments
> metrics definitions", which in the case of delay follows my proposal. I
> appreciate that.
I refurbished the whole draft to have it more consistent while avoiding duplication. I renamed 'subpath metric' to 'segment metric' to have this metric more close to operational needs.
>
> The section on "A Definition of a sample of Packet Loss of a segment of
> the path" correctly includes an
>
> o a value of Lk of 2 means that Bk.a has a value of 1 (not observed)
> and that Bk.b have a value of 0 (observed);
>
> The discussion qualifies this as
>
> o A value of Lk of 2 (1,0) corresponds to a mistake in the ordering
> of Ha and Hb over the path coming either from the configuration
> (asumption on the path) or from the processing of the vectors: bad
> scrutening of the path ordering information, or some other mistake
> in the measure or the reporting. It is not in the scope of this
> document to go in further details which are mostly implementation
> dependent. This value MUST not be used to compute packet lost
> statistics.
>
> After talking with Nicolai Leymann (having multicast expertise) I learned
> that after a rerouting event, a passive monitor along a path may indeed no
> longer be in it, if the new route is around (also a replication point may
> change). That means, a value of Lk of 2 can result from a usual re-routing
> event, it is not an anomaly. In that case, the results should be
> considered.
Spatial metrics are defined on the top of a one-to-one framework (one-to-many covered by the second part of the draft), just adding the knowledge of the Path digest. So the path digest is known for each metric result. But this does not preclude that the path does not change between 2 results of the same measure.
>
> I think all metric definitions should respect this case.
Not sure I understand. Which metrics ?
>
> I've only searched for the wird "reroute" in the current version and found
> it once. Sgould my comment be discussed already, I apologize for not
> having read that section.
>
> Regards
>
>
> Ruediger
>
>
Regards
emile
_______________________________________________
ippm mailing list
ippm at ietf.org
https://www1.ietf.org/mailman/listinfo/ippm