From owner-freebsd-hackers Tue Aug 8 12:20:03 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.11/8.6.6) id MAA11370 for hackers-outgoing; Tue, 8 Aug 1995 12:20:03 -0700 Received: from irbs.irbs.com (irbs.com [199.182.75.129]) by freefall.cdrom.com (8.6.11/8.6.6) with ESMTP id MAA11353 for ; Tue, 8 Aug 1995 12:19:58 -0700 Received: (from jc@localhost) by irbs.irbs.com (8.6.11/8.6.6) id PAA29384; Tue, 8 Aug 1995 15:18:30 -0400 From: John Capo Message-Id: <199508081918.PAA29384@irbs.irbs.com> Subject: Re: client & server ppp To: tony@thing.sunquest.com Date: Tue, 8 Aug 1995 15:18:29 -0400 (EDT) Cc: freebsd-hackers@freefall.cdrom.com (freebsd-hackers) In-Reply-To: <9508081845.AA09868@thing.sunquest.com> from "tony@thing.sunquest.com" at Aug 8, 95 11:45:47 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2286 Sender: hackers-owner@FreeBSD.org Precedence: bulk It sounds like you have the getty on the wrong port or you are dialing out on the wrong port. Getty should not wake up when you dial out on a /dev/cuxxx port. You may want to use ppp, aka iijppp, to dial into the Portmaster. I have been using it for uni-directional demand dialing for a year and used it for several months for bi-directional demand dialing. John Capo IRBS Engineering tony@thing.sunquest.com writes: > > > Question regarding client and server ppp on the same system (using pppd). > > For a while I had client pppd working just fine. > > This weekend, I set up a Livingston portmaster to do 'on-demand' ppp to > a remote FreeBSD system (V34 modem) > - The Livingston advertises a RIP route for the remote machine > - Whenever it receives an IP packet, dials the remote FreeBSD machine. > - The portmaster runs a send/expect sequence to logs into the FreeBSD > system on getty/ttyd1. The shell for the account is a script which > fires up PPPd. > > Everything is working quite well, after 15 mins of IP inactivity the Livingston > will reset the port, dropping CD and thus DSR drops causing pppd on FreeBSD > to terminate and the getty respawns. > > > The problem is that several other folks use the remote system and when server > PPP isn't running they still want to be able to initiate client side PPP. > > Problems here: > 1) AT echo/reply is disabled (E0Q1) since getty/autoanswer is enabled. > This makes it hard to reliably dial out, and really needs to be toggled back. > > 2) Kermit seems happy to dial the port even when getty is still connected, but > pppd detects that getty (or the login it spawns due to line activity) has > the device open and thus pppd fails. > > Clearly in the client.ppp.start and client.ppp.stop scripts I can: > "Off" getty in ttys, HUP init > "E1Q0" to the modem > > start client ppp session > end-ppp session > > "On" getty in ttys, HUP init > "E0Q1" to the modem" > > Is anyone aware of alternatives avoiding some or all of the above ? > > I have not been able to get 100% reliable dial/hup scripts for client PPP > and am worried that someone using client PPP _could_ leave the system in > a state where the Livingston will be unable to connect for server PPP. > > thanks > > tony >