From owner-freebsd-isp Sun Jan 4 16:04:30 1998 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA08722 for isp-outgoing; Sun, 4 Jan 1998 16:04:30 -0800 (PST) (envelope-from owner-freebsd-isp) Received: from ns2.cetlink.net (root@ns2.cetlink.net [209.54.54.20]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id QAA08715 for ; Sun, 4 Jan 1998 16:04:25 -0800 (PST) (envelope-from jak@cetlink.net) Received: from unknown (ts1-cltnc-29.cetlink.net [209.54.58.29]) by ns2.cetlink.net (8.8.5/8.8.5) with SMTP id TAA25643; Sun, 4 Jan 1998 19:04:14 -0500 (EST) From: jak@cetlink.net (John Kelly) To: Atipa Cc: freebsd-isp@FreeBSD.ORG Subject: PPPD terminal server on FreeBSD, NATD Date: Mon, 05 Jan 1998 01:04:56 GMT Message-ID: <34b02cb1.200036@mail.cetlink.net> References: In-Reply-To: X-Mailer: Forte Agent 1.01/16.397 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by hub.freebsd.org id QAA08717 Sender: owner-freebsd-isp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 4 Jan 1998 14:51:32 -0700 (MST), Atipa wrote: >> >> I'm doing some simple testing to see if FreeBSD is feasible as a >> >> terminal server. My first test of 2.2.5-stable, using only two ports >> >> simultaneously, indicates there is considerable efficiency loss due >> >> to transmit underruns. >> >What getty are you using? >> >> None whatsoever. This is a PPPD-only terminal server. My ttys >> entries look like this: >> >> ttyd3 "/usr/sbin/pppd +pap -detach 57600" unknown on secure > >I use both pppd and userland PPP (ijppp), and I have gotten a bit more >consistent results with ijppp. I ran a another test with 2.2.2-release, a different machine, and a different multiport serial card. Although I got the same result, I learned something new. The first port is connected to the Internet, ISDN 128k with 230k DTE speed, and the second port is null modem to another local machine at 56k. While downloading from the Internet to the test machine, and simultaneously downloading from the test machine to another local machine, I can saturate both serial ports at full speed, with no apparent efficiency loss. The ISDN port runs about 13kb/s, and the null modem port runs about 5.4kb/s. But when I try to download *through* the test box, the null modem port drops to about 4.0kb/s. Since I don't have static IP addresses for this test setup, I'm using NATD to get connectivity to the outside world, via the test box. So NATD seems to be causing the efficiency loss. Without NATD in the mix, a FreeBSD terminal server might scale to 24 or 32 ports, since it scaled to 2 ports with no apparent loss. But more testing is needed to know for sure. John