From owner-freebsd-hackers Wed Oct 28 13:04:57 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA18413 for freebsd-hackers-outgoing; Wed, 28 Oct 1998 13:04:57 -0800 (PST) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from sasami.jurai.net (sasami.jurai.net [207.153.65.3]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA18387 for ; Wed, 28 Oct 1998 13:04:51 -0800 (PST) (envelope-from winter@jurai.net) Received: from localhost (winter@localhost) by sasami.jurai.net (8.8.8/8.8.7) with SMTP id QAA11871; Wed, 28 Oct 1998 16:04:12 -0500 (EST) Date: Wed, 28 Oct 1998 16:04:10 -0500 (EST) From: "Matthew N. Dodd" To: Kris Kirby cc: hackers@FreeBSD.ORG Subject: Re: dropping to single user mode on a telnet connection In-Reply-To: <3637788B.7912F11A@airnet.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Wed, 28 Oct 1998, Kris Kirby wrote: > Matthew N. Dodd wrote: > > From the standpoint of someone observing the operation of 100s of machines > > installed world wide I would have to disagree with you. > And you've obviously not read their web page, have you? You've not been following the thread have you. He said "serial consoles add complexity to the system which tends to make it more unstable/less reliable." I asserted that "based on experience, I don't believe this is the case regarding serial consoles." (for other things added complexity may indeed make for less reliable/more unstable operation. We're talking about plain old simple serial consoles so everything outside of serial consoles is moot.) > > > 2) because serial consoles are going away on the next round of > > > motherboards > > > > Poor design decision? > > On the contrary. They are doing this in a large cluster. > He more or less implied that "we're using serial consoles now, but the next rev of the system we won't be." I labeled this choice as "a poor design decision." referring to the lack of serial consoles on the new systems. == I work at a place that uses a client/server console management system based on Vixie's rtty package. (I believe local mods are Kerberos 5 auth and encryption and real client/server support instead of local access.) Basically you type 'console ' from any machine as any user listed in the ACLs and you are connected to the console port of hostname. Multiple users can connect to the same console and all i/o is logged. Quite nice. I'm pretty sure that Ron Minnich desire to perpetuate this hack due to prolonged exposure to typing 'telnet sometermserver xxxx' to get console access. (Or maybe he drags out a laptop/dumbterm). I'm still trying to duplicate the above setup for myself at home but a bit of code is required to match the stated feature list. Since someone is sure to bring up the fact that more hardware (term servers and a console management server) is less stable I'll answer now by saying "less, but to no noticeable degree." I'm fairly certain the only time I've seen 2511 term-servers reboot is when their power cords get tripped over. A lightly loaded console management server isn't likely to be loaded enough to tickle any bugs, even if it was hosted on a SCO box. :) -- | Matthew N. Dodd | 78 280Z | 75 164E | 84 245DL | FreeBSD/NetBSD/Sprite/VMS | | winter@jurai.net | This Space For Rent | ix86,sparc,m68k,pmax,vax | | http://www.jurai.net/~winter | Are you k-rad elite enough for my webpage? | To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message