From owner-freebsd-hackers Wed Aug 23 9: 7:52 2000 Delivered-To: freebsd-hackers@freebsd.org Received: from sentinel.office1.bg (sentinel.office1.bg [195.24.48.182]) by hub.freebsd.org (Postfix) with SMTP id DD66437B43C for ; Wed, 23 Aug 2000 09:07:35 -0700 (PDT) Received: (qmail 66409 invoked by uid 1001); 23 Aug 2000 16:02:23 -0000 Date: Wed, 23 Aug 2000 19:02:23 +0300 From: Peter Pentchev To: freebsd-hackers@freebsd.org Subject: mount_mfs /var on a diskless workstation Message-ID: <20000823190223.H63286@ringwraith.office1.bg> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Alright, so probably I'm missing something again, like in my previous post about mounting different subdirs in a filesystem (btw thanks to those who replied then, all replies have been most helpful). So.. Is there a way to cleanly shutdown (or reboot) a NFS-rooted workstation? rc.diskless2 MFS-mounts /var, and when init(8) tries to kill the running daemons, the /etc and /var mount_mfs's obstinately and rightly refuse to die - /etc has /etc/ttys in use, and /var has.. lots of files, most notably /var/db/mounttab. As a result, mount_mfs lives, init waits quite a while, tries to kill it with SIGKILL, fails again, and complains, advising a ps axl.. I wonder if there might be some kind of race condition here - some processes that received the SIGTERM, but not quite finished yet - but anyway, just how stupid would it be if I tried to teach the kernel to leave mount_mfs alone on a kill(-1, SIGTERM)? Would this break anything but init on shutdown? G'luck, Peter -- This sentence was in the past tense. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message