Re: (no subject)

> PS I don't think there's a connection between Y.1541 and
> the various ITU-T Admission Recs (Y.1571, now Y.2171).
> Y.1541 specifies categories of performance objectives (mostly), and
> it seems like a Scavenger service would have "Unspecified" objectives
> (same as the Y.1541 Best Effort). Scavenger is more about *when* it
> works than about how well, IMO.

again, my fault for twisting the references around.  I meant to refer  
to Y.1571

as for scavenger service, it uses resources not used by best effort or  
better than best effort flows.  so, if we assume a transit node only  
supports best effort, then during bursty spikes in load will allocate  
available resources to BE at the expense of scavenger service.  that  
may be a case of "when", but the result is a service below that of BE  
since you are not relying on random drop or drop from head schemas.

outside of scavenger service, I know of another on-going proprietary  
effort at lower than best effort service.

regards,

-ken