From owner-freebsd-small Wed Jan 26 9: 7:36 2000 Delivered-To: freebsd-small@freebsd.org Received: from edna.bus.net (edna.bus.net [207.41.24.10]) by hub.freebsd.org (Postfix) with ESMTP id B554E14A2C for ; Wed, 26 Jan 2000 09:07:27 -0800 (PST) (envelope-from mso@bus.net) Received: from bus.net (17.ct9.dyn.connix.net [209.66.147.24]) by edna.bus.net (8.8.8/8.8.8) with ESMTP id MAA12273; Wed, 26 Jan 2000 12:07:15 -0500 (EST) (envelope-from mso@bus.net) Message-ID: <388F299C.8976DAC5@bus.net> Date: Wed, 26 Jan 2000 12:06:36 -0500 From: "Michael S. O'Donnell" X-Mailer: Mozilla 4.51 [en]C-DIAL (Win95; I) X-Accept-Language: en MIME-Version: 1.0 To: Mike Wade Cc: "Jeffrey S. Sharp" , Greg Lehey , freebsd-small Subject: Re: your mail References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-small@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > > > Also it would be nice to have an "accidental power off" safe file > > > system when using hard drives for embedded devices such as > > > Internet Appliances. i just started developing a hardware solution for this for a client who has servers going to trade shows where uniformed individuals occasionally take it upon themselves to 'push the button'. i don't mind losing the project to a software solution. i know that would be the solution the client would prefer (hardware would require me to build a new unit for every different server; the design doesn't change, but, the footprint and mounting does). if anyone knows how to do it, please tell me and/or the list. i will request that the client make the solution available under the FreeBSD license (my hardware solution will not be). the client runs _all_ systems under FreeBSD and i'm quite confident he would welcome the opportunity to contribute. thank you. Michael S. O'Donnell Mike Wade wrote: > > On 26 Jan 2000, Jeffrey S. Sharp wrote: > > > Therefore, I suggest something like what Warner has done (and that I am > > working on as time permits), where the flash is the root fs and /tmp, > > /var, and so on are mounted as small MFS filesystems. The flash is > > normally kept mounted read-only. Then, instead of running an update > > script, one simply remounts the flash read-write, makes changes, and > > remounts read-only. > > I've attempted this and I ended up with a filesystem of corrupted files > when mounting read-only, remounting read-write, then remounting read-only > several times. I ended up partitioning the flash and creating a read-only > binary partition and a read/write config partition that is mounted only on > update. > > >From the mount man page: > > BUGS > It is possible for a corrupted file system to cause a crash. > Switching a filesystem back and forth between asynchronous and normal > operation or between read/write and read/only access using ``mount > -u'' may gradually bring about severe filesystem corruption. > > It would be very nice to have this feature when dealing with flash. Also > it would be nice to have a "accidental power off" safe file system when > using hard drives for embedded devices such as Internet Appliances. > > --- > Mike Wade (mwade@cdc.net) > Director of Systems Administration > CDC Internet, Inc. > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-small" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-small" in the body of the message