From owner-freebsd-isp  Mon Oct 14 08:58:02 1996
Return-Path: owner-isp
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id IAA05758
          for isp-outgoing; Mon, 14 Oct 1996 08:58:02 -0700 (PDT)
Received: from labinfo.iet.unipi.it (labinfo.iet.unipi.it [131.114.9.5])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id IAA05702;
          Mon, 14 Oct 1996 08:57:25 -0700 (PDT)
Received: from localhost (luigi@localhost) by labinfo.iet.unipi.it (8.6.5/8.6.5) id QAA13670; Mon, 14 Oct 1996 16:30:16 +0100
From: Luigi Rizzo <luigi@labinfo.iet.unipi.it>
Message-Id: <199610141530.QAA13670@labinfo.iet.unipi.it>
Subject: Re: ppp over vty possible ?
To: dg@root.com
Date: Mon, 14 Oct 1996 16:30:15 +0100 (MET)
Cc: exidor@nimbus.superior.net, question@FreeBSD.org, isp@FreeBSD.org
In-Reply-To: <199610141531.IAA20691@root.com> from "David Greenman" at Oct 14, 96 08:31:00 am
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text
Sender: owner-isp@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

> >large block to attempt compression. But -- apart for this being a
> >symptom of poor design in the modem -- what is intriguing me is
> >the delay -- 200ms -- so close to what is a TCP_FASTTIMEOUT.
> 
>    fastimeout can't be involved with pings since they use ICMP, not TCP. 14.4K

I was not using ping, I have a modified kernel which prints out
the current RTT and bandwidth for every connection (I am doing some
experiments on congestion control stuff right now). The same could
be seen with tcpdump, I guess, just a bit more annoying to do time
differences.

BTW the same kernel showed that using "telnet", a large number of
telnet options (31 bytes if I remember well) is exchanged at every
newline. Not that one should care much (at least, with a symmetrical
channel), but it is impressive how many things occur behind (or
below) you.

> modems usually have large delays due to buffering and compression delays. Some
> are better than others. I used to see 170ms on certain Intel modes, and as low
> as 110ms on certain 28.8K Hayes modems. Supra's are usually about the worst
> at near 200ms.

Thanks for these data points. How did you get them, using ping or
something else ?

	Luigi
====================================================================
Luigi Rizzo                     Dip. di Ingegneria dell'Informazione
email: luigi@iet.unipi.it       Universita' di Pisa
tel: +39-50-568533              via Diotisalvi 2, 56126 PISA (Italy)
fax: +39-50-568522              http://www.iet.unipi.it/~luigi/
====================================================================