From owner-freebsd-stable Sat Jan 4 5: 1: 3 2003 Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id B58E737B401 for ; Sat, 4 Jan 2003 05:01:01 -0800 (PST) Received: from smtp-relay.omnis.com (smtp-relay.omnis.com [216.239.128.27]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3DC5743EE6 for ; Sat, 4 Jan 2003 05:01:01 -0800 (PST) (envelope-from wes@softweyr.com) Received: from [192.168.0.5] (dslpool2-019.networldnoc.net [209.63.227.179]) by smtp-relay.omnis.com (Postfix) with ESMTP id AC2B742FC6 for ; Sat, 4 Jan 2003 04:58:37 -0800 (PST) Subject: Re: FreeBSD Stability From: Wes Peters To: freebsd-stable@freebsd.org In-Reply-To: <20030103154323.GA454@laurel.tmseck.homedns.org> References: <200212170023.gBH0Nvlu000764@beast.csl.sri.com> <20030103000232.GA52181@blazingdot.com> <20030103062708.GA426@laurel.tmseck.homedns.org> <20030103084232.GA3371@localhost.bsd.net.il> <20030103154323.GA454@laurel.tmseck.homedns.org> Content-Type: text/plain Organization: Softweyr LLC Message-Id: <1041659893.9975.179.camel@zaphod.softweyr.com> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.2.0 Date: 04 Jan 2003 05:58:13 +0000 Content-Transfer-Encoding: 7bit Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Fri, 2003-01-03 at 15:43, Thomas Seck wrote: > * Nimrod Mesika (nimrod-me@bezeqint.net): > > [Uptime] > > > Think about compute servers. Our CAD servers can run simulations and > > other types of processes for ~40 hours. You definitely don't want to > > interrupt a running system and it finding some idle time for service > > gets really difficult. > > Of course not. But these are probably neither publicly accessible nor > 'monitored' by Netcraft and thus not subject to public 'uptime-size' > wars. > > > Would be nice if you could upgrade subsystems one at a time. This > > way one could, for example, shutdown the network subsystem, load > > the new version and restart it. > > Sounds like what microkernels were designed for. > > > And uptimes are not important. Downtimes *are*. > > Yes. Especially the unscheduled ones. Don't be silly, uptimes are terribly important when they're not long enough to be useful. They're no longer important when they've gotten long enough to last between system upgrades, which FreeBSD and a number of other systems are regularly capable of these days. I remember people being mightily impressed with VAX/VMS systems being able to stay up for 30 days at a time. I also more recently recall system administrators being very disappointed by Windows NT servers because they couldn't stay up for 6 days at a time and they had NO time in their schedule when the machines could be rebooted without disrupting workflow between 0400 Monday and 0400 Saturday. -- Where am I, and what am I doing in this handbasket? Wes Peters wes@softweyr.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message