[6786] in www-talk@info.cern.ch
Re: More on HTTP-NG timings
daemon@ATHENA.MIT.EDU (Karl Auerbach)
Tue Nov 22 21:08:00 1994
Date: Wed, 23 Nov 1994 02:57:09 +0100
Errors-To: listmaster@www0.cern.ch
Reply-To: karl@cavebear.com
From: Karl Auerbach <karl@cavebear.com>
To: Multiple recipients of list <www-talk@www0.cern.ch>
>> This is true except... why did the one TCP packet get lost? Virtually all
>> packet loss on the Internet is caused by congestion.
There are still too many boxes out there that drop the main packet
when doing an ARP. This hurts UDP stuff more than TCP flows.
> I spent a
> good chunk of time using the Internet on a regular basis from Tromsoe
> Norway where it took 10 .. 17 hops to get to my destinations. Lots
> of packets were lost in their travels, some days it approached 10%
> of the packets, and it was consistantly 2% loss.
Some of us are (unfortunately) behind some major US providers which
experience as much as a 50% packet loss rate on bad days and 20% and
above is quite common.
Now, if TCP/IP engines would do Path MTU discovery and quit sending
those little 500+ byte packets whenever something goes off their
subnet...
--karl--
