[e2e] Can feedback be generated more fast in ECN?

David P. Reed dpreed at reed.com
Fri Feb 16 11:21:44 PST 2001


At 10:42 AM 2/15/01 +0000, Lloyd Wood wrote:
>On Wed, 14 Feb 2001, David P. Reed wrote:
>
> > I'm usually pretty precise in my language. Confusing terms can lead 
> to  bad
> > design - it's like trying to prove theorems based on inconsistent axioms.
> >
> > Congestion is the condition of a network that is operating with queues 
> near
> > full on some paths.
>
>If (in some hypothetical network with near-full queues) no packets are
>ever dropped, is it still congestion?

Assuming it's a serious question... the answer is yes and no.

Yes, because the same network could operate with empty queues, and 
consequently much lower congestion (the only way I know that a net can 
operate with near full queues and no dropped packets is if the flows have 
zero variance in rate of arrival at the fully queued routers. Variance 
would cause drops, for sure.  And a situation that had no variance in 
arrival rate needs no queues)

No, because the network is giving predictable service.


- David
--------------------------------------------
WWW Page: http://www.reed.com/dpr.html





More information about the end2end-interest mailing list