[e2e] Fundamental Questions about Router Queue in High Speed IP Networks

[David P. Reed]

At 05:21 PM 8/24/2001 +0100, Vos, E.W. wrote:
>I don't see why buffering is so wrong. Actually, a buffer should be large
>enough to accomodate incidental bursts of traffic. Aren't the effects of
>dropping a packet much worse than delaying one a little bit?

Not a well-formed question - some drops are essential, some are 
costly.  Dropping a packet is how you tell TCP to back off.  Delaying the 
dropping may lead to a larger catastrophe.  RED and ECN (and even SQ) are 
just simulated dropping, and increasing queue size also slows TCP's back-off.

What you want buffering to do is accomodate bursts that are very transient 
(so end-to-end congestion control can't possibly prevent or respond) but 
knowing that they are transient requires omniscience that probably exceeds 
the "speed of light".






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