Re: [nm-wg] Response Schema question

[Dan Gunter <dkgunter@lbl.gov>]

Eric Boyd wrote:
> Greetings,
>
> In the relaxed response schema, delay (RT or OW) is defined as:
>
> xsd:double, attribute units { token }
>
> Could someone (Dan?) please explain what units are implied here, if any, 
> and explain exactly what this means? (Variable units? Fixed fine-grained 
> units, implying integers are OK? Fixed large units, implying fixed point 
> numbers are required?)
Something representing time. The RFC uses seconds, so that's what I 
would use.

But I would like to "set" some of these units, to avoid interop problems 
down the road. Volunteers? The old(er) XSD schema default units provide 
a starting point.

> In addition, it seems like there should also be a way to specify one of 
> two special values for delay.
>
>         a) "Undefined"
>         b) "Infinite"
>
> Can this be done in the current schema?
>
> The latter ("Infinite") needs the context of the measurement methodology 
> ... what is your loss threshold before you "give up".
>
> Can this be done in the current schema?
a) maybe this should be reported somewhere else, i.e. "measurement 
failed"? is that the intent?

b) I think any value over the lossThreshold param is considered infinite 
(i.e. not a delay, a loss). note that this implies that lossThreshold 
should have the same type as delay, which it currently doesn't (it's an 
int).

> --Eric
>
>
> Eric L. Boyd
> eboyd@internet2.edu
> 734-352-7032 (work)
> 734-834-6708 (cell)
>
> Office 230
> Internet2
> 3025 Boardwalk
> Ann Arbor, MI 48108

--
  ((    Dan Gunter
   ))   Computer Scientist, Lawrence Berkeley National Lab
C|   |  M/S 50B-2239, 1 Cyclotron Road, Berkeley, CA, 94720
 |___|  Phone:510/495-2504 Fax:510/486-6363