PSINet/Cogent Latency

Tue Jul 23 03:51:39 UTC 2002

On Mon, Jul 22, 2002 at 11:34:44PM -0400, Phil Rosenthal wrote:
> 
> My point exactly -- I guess some people disagree...
> Probably with any sort of queuing there will only be minimal packet loss
> at 40mbit, but at any point one more stream can push it up to 43mbit,
> and then queuing might no longer be enough... (and even if it is, can we
> say lag?)

Efficient packet loss is still packet loss. Just because you manage to 
make the link "look good" by slowing down TCP before your queueing latency 
starts going up doesn't make your network any less ghetto.

IMHO the biggest problem in peering is getting the other side to actively 
upgrade links to prevent congestion. If you're not in a position where you 
can dictate terms to your peer, move traffic off it and let economics take 
care of the rest. Leaving a congested peer up for your own benefit at the 
expense of your customers is one of the surest ways to lose customers to 
someone who doesn't.

I'd rather have a noncongested gige public peer than a ds3 private peer
any day.

-- 
Richard A Steenbergen <ras at e-gerbil.net>       http://www.e-gerbil.net/ras
PGP Key ID: 0x138EA177  (67 29 D7 BC E8 18 3E DA  B2 46 B3 D8 14 36 FE B6)