From owner-freebsd-questions Mon Apr 12 13:52:39 1999 Delivered-To: freebsd-questions@freebsd.org Received: from ds9.dreamhaven.org (dt0b4nfc.san.rr.com [24.94.22.252]) by hub.freebsd.org (Postfix) with ESMTP id DCB5E15634 for ; Mon, 12 Apr 1999 13:52:33 -0700 (PDT) (envelope-from data@ds9.dreamhaven.org) Received: from data (helo=localhost) by ds9.dreamhaven.org with local-esmtp (Exim 2.12 #2) id 10Wned-000G3f-00; Mon, 12 Apr 1999 13:50:03 -0700 Date: Mon, 12 Apr 1999 13:50:03 -0700 (PDT) From: Bryce Newall X-Sender: data@ds9.dreamhaven.org To: Doug White Cc: FreeBSD Questions List Subject: Re: FD_SETSIZE In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Mon, 12 Apr 1999, Doug White wrote: > > equivalent to FD_SETSIZE in 2.2.8 and below? When I have that option in a > > 3.1 kernel, config tells me it doesn't recognize that option. > > I think it's a sysctl now, and you generally have to fix this on a > per-program basis at build-time. Hmm... so it's no longer a kernel option that can be "global" to all programs like it used to be? Here's why I'm asking about this: We (DreamHaven) just upgraded one of our systems from FreeBSD 2.2.8-STABLE to 3.1-STABLE. We did a fresh installation on a new hard drive, installed all the programs we use, threw in the /home hard drive, and were off and running. One of the web sites we host is extremely high-volume, and I suspect that it may have been responsible for a system crash that occurred last night. Here is a sample of the pile of error messages that got dumped into /var/log/messages: Apr 11 18:59:08 calico /kernel: Apr 11 18:59:08 calico /kernel: 3>file: table is full Apr 11 18:59:08 calico /kernel: file: table is full Apr 11 18:59:41 calico last message repeated 1781 times Apr 11 18:59:41 calico syslogd: /var/run/utmp: Too many open files in system Apr 11 18:59:41 calico syslogd: /var/run/utmp: Too many open files in system Apr 11 18:59:41 calico /kernel: file: table is full Apr 11 19:00:00 calico last message repeated 3310 times Apr 11 19:00:00 calico /kernel: pid 5114 (cron), uid 0: exited on signal 11 Apr 11 19:00:00 calico /kernel: file: table is full Apr 11 19:00:00 calico last message repeated 12 times It wasn't a complete crash; one of our administrators was able to do a remote "shutdown -r now" after she attempted it several times and managed to "slip" the shutdown process into the process table. We're worried that something like this may happen again, though, and this one high-volume web server I mentioned before may be aggrivating the problem. We use Roxen Challenger for our web server, by the way. (Apache would die... :> ) None of us are programmers by any means, and sysctl is, well, something of a mystery to me as well. I appreciate any advice you may have to offer. :) Thanks! ********************************************************************** * Bryce Newall * Email: data@dreamhaven.net * * WWW: http://www.dreamhaven.net/~data * ICQ: 32620929 * * "Insanity takes its toll. Please have exact change." * ********************************************************************** To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message