From owner-freebsd-hackers Mon Dec 14 16:13:43 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id QAA07124 for freebsd-hackers-outgoing; Mon, 14 Dec 1998 16:13:43 -0800 (PST) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from bright.fx.genx.net (bright.fx.genx.net [206.64.4.154]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id QAA07112 for ; Mon, 14 Dec 1998 16:13:37 -0800 (PST) (envelope-from bright@hotjobs.com) Received: from localhost (bright@localhost) by bright.fx.genx.net (8.9.1/8.9.1) with ESMTP id TAA99453; Mon, 14 Dec 1998 19:16:00 -0500 (EST) (envelope-from bright@hotjobs.com) X-Authentication-Warning: bright.fx.genx.net: bright owned process doing -bs Date: Mon, 14 Dec 1998 19:16:00 -0500 (EST) From: Alfred Perlstein X-Sender: bright@bright.fx.genx.net To: "David G. Andersen" cc: spork , Bernd Walter , Kevin Day , hackers@FreeBSD.ORG Subject: Re: NFS thoughts In-Reply-To: <13941.42539.360724.653774@torrey.cs.utah.edu> 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 Mon, 14 Dec 1998, David G. Andersen wrote: > Lo and Behold, Alfred Perlstein said: > > > > > > I also haven't seen the "I've mounted soft and intr, yet things still > > > hang" behaviour using version 2 and udp. Any consensus on that? > > > > why would you mount _both_ soft and intr? to me they seem mutually > > exclusive. > > > > 'intr' allows you to intrupt a hung NFS proc so that it recives a > > transient error on a filesystem call, the process will hang forever unless > > NFS comes back, or you ^C it > > > > 'soft' automates that with a timeout however signals won't work, but after > > some time the process will unhang and get an error on the filesystem call. > > > > Are you trying to get an auto-timeout like mount with that ability to ^C? > > > > generally intr is best, the idea of many processes timeing out on NFS > > mounts should the server crash, makes my stomach turn. > > Note that until kern/8732 is fixed, this can be dangerous (hang > your machine dangerous). Mounting with 'intr' and then sending > SIGINTR while a process is attempting to flush a dirty block will > cause the machine to wedge. > > -Dave > there's about 5 suggested fixes in that PR, can someone take a look? maybe Matt "the PR killer" Dillon can take a look? :) breaking semantics a bit is a far better solution than a wedged system which allows for zero semantics. :) Alfred Perlstein - Programmer, HotJobs Inc. - www.hotjobs.com -- There are operating systems, and then there's FreeBSD. -- http://www.freebsd.org/ 3.0-current To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message