From owner-freebsd-current Thu Jan 21 07:54:51 1999 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA19028 for freebsd-current-outgoing; Thu, 21 Jan 1999 07:54:51 -0800 (PST) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from cs.rpi.edu (mumble.cs.rpi.edu [128.213.8.16]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id HAA19022 for ; Thu, 21 Jan 1999 07:54:49 -0800 (PST) (envelope-from crossd@cs.rpi.edu) Received: from cs.rpi.edu (crossd@o2.cs.rpi.edu [128.113.96.156]) by cs.rpi.edu (8.9.1/8.9.1) with ESMTP id KAA28454 for ; Thu, 21 Jan 1999 10:54:38 -0500 (EST) Message-Id: <199901211554.KAA28454@cs.rpi.edu> To: current@FreeBSD.ORG Subject: mountd Date: Thu, 21 Jan 1999 10:54:36 -0500 From: "David E. Cross" Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I posted this awhile ago to -questions, but never received a reply. We have a number of FreeBSD NFS servers here. Occasionally we need to change the exports list on the servers and send mountd a SIGHUP. This leads to a condition that in many ways is much worse than a server reboot. What happens is for the duration of mountd reading the exports file it denies all NFS requests. This has a number of bad effects; 1) any user home and system directories become unavailable, with the error 'permission denied' 2) (and this is far worse), any process with a mapped .text segment off of the NFS server, should it branch to code not in the cache gets immediately killed. This include user processes that are running from home directories, and system processes (such as ssh). If we were to reboot the machine it would just hang those connections until the machine came back, without killing anyone. Is there a solution to this problem? I know that none of HP-UX, IRIX, or Solaris have this problem. -- David Cross To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message