On Fri, Jul 18, 2014 at 11:39 AM, Leeyoung 
<[email protected]<mailto:[email protected]>> wrote:
Hi Qin and Richard,

Thanks for putting a good quality slide together. I think it is good with this 
version.

On one question:
–      Do we allow more flexible statistics operators (e.g., mean, avg, 
x-percentile, variance)
•       In current draft, delay and delay jitter are both on delay, with one 
reflect mean and the other variance

Do you mean, “Do we allow more flexible statistics for/from/?? operator?” Not 
sure what you mean. If the network operators are the ones that provide these 
stats to ALTO application, then we may say, “Do we need more flexible 
statistics for some applications?” for instance. May be I am missing something 
here.
[Qin]: Yes, you misunderstand what it said here. The operator is not service 
provider but more related to how you measure the TE metrics and how you 
calculate TE metrics. The measurement unit or measurement method may indicate 
what statistics operator we are using, e.g., delay use microseconds as unit and 
choose the mean value as the value of cost metric.

Flexible stat such as percentile and variance are all good, yet there might be 
scaling issues and usability of such data and the networks that provide such 
data do not normally have in their TED  except jitter variance. This could 
imply an overhead for networks.
[Qin]: Yes, bandwidth related metrics  are not provided in the unit of 
percentile in TED becos bandwidth related metrics all use the same measurement 
unit and they don’t want to change unit to lose measurement precision. But in 
alto context, alto client maybe more care about how much percentage of the 
bandwidth are used, how much percentage of the bandwidth are not reserved 
rather than the total number of bytes per sec for each bandwidth related metric.
I am not sure we really need new semantic to indicate what kind of statistics 
operator are supported, but measurement unit or measurement method may be 
impacted.

Thanks,
Young

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

Reply via email to