From owner-freebsd-hackers Wed Mar 29 21:54:23 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id VAA15805 for hackers-outgoing; Wed, 29 Mar 1995 21:54:23 -0800 Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id VAA15792 for ; Wed, 29 Mar 1995 21:54:08 -0800 Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id PAA30807; Thu, 30 Mar 1995 15:45:54 +1000 Date: Thu, 30 Mar 1995 15:45:54 +1000 From: Bruce Evans Message-Id: <199503300545.PAA30807@godzilla.zeta.org.au> To: bugs@ns1.win.net, hsu@cs.hut.fi Subject: Re: Mail... (fwd) Cc: freebsd-hackers@freefall.cdrom.com Sender: hackers-owner@FreeBSD.org Precedence: bulk > Using a terminal server front end will allow you to switch users based > on incoming protocol type, load, or software version. (you also don't > have to futz with serial drivers or burden your box with serial interrupts). >386-33 with 16 V34's connected with a cyclades card doesn't seem to cause >any visible load (load average stays under 15%). Haven't tried to server >PPP yet, but will do so soon. This is with Linux, 8M, dedicated to be a >terminal server. That's surprising. I've (carefully) measured the following loads for 115.2Kbps raw mode input on a 486DX2/66: FreeBSD-1.1.5-16550 5.6% Linux-1.1.late-16550 6.6% FreeBSD-1.1.5-16450 15.0% FreeBSD-2.1-cyclades 19.8% Linux-1.1.late-16450 23.1% Linux-1.2-16450 23.9% FreeBSD-2.1-16450 25.4% Linux-1.2-cyclades 29.0% In raw mode, the cyclades should be slightly more efficient than a dumb 16550 board. It is less efficient now because of 14/3 times as many interrupts and higher overhead per interrupt. In cooked mode, the cyclades should be much more efficient. Bruce