From owner-freebsd-stable Wed Dec 1 6:28:44 1999 Delivered-To: freebsd-stable@freebsd.org Received: from trooper.velocet.net (trooper.velocet.net [216.126.82.226]) by hub.freebsd.org (Postfix) with ESMTP id 4BCC414F9F for ; Wed, 1 Dec 1999 06:28:37 -0800 (PST) (envelope-from dgilbert@trooper.velocet.net) Received: (from dgilbert@localhost) by trooper.velocet.net (8.9.3/8.9.3) id JAA83318; Wed, 1 Dec 1999 09:28:33 -0500 (EST) (envelope-from dgilbert) From: David Gilbert MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <14405.12433.220410.215813@trooper.velocet.net> Date: Wed, 1 Dec 1999 09:28:33 -0500 (EST) To: Brad Knowles Cc: David Gilbert , stable@FreeBSD.ORG Subject: Re: vinum experiences. In-Reply-To: References: <14405.8810.777783.992833@trooper.velocet.net> X-Mailer: VM 6.75 under 20.4 "Emerald" XEmacs Lucid Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >>>>> "Brad" == Brad Knowles writes: Brad> At 8:28 AM -0500 1999/12/1, David Gilbert wrote: >> 1. Removing a device (at least, with the ahc controller) locks the >> bus even though I have a RAID hot-swap ready chassy (that properly >> isolates the bus between commands). In my test, I had a completely >> quiet SCSI bus when I removed one of the drives. I then wrote to >> the RAID array. I got: Brad> Vinum does not (currently) support any kind of hot-swap Brad> capability. See and Brad> . Actually, I don't think this is a vinum problem. It appears that the SCSI layer locks hard when a drive is missing, but I could be wrong. >> 2. It's not obvious how to replace a dead drive with vinum. I have >> tried several times to perform this action without success. Vinum >> attach says it should do it. It refuses. Brad> Assuming that the system has been safely shutdown and is Brad> physically powered off, I believe that the following process Brad> will work: Brad> 1. Remove the old drive, noting it's physical position in the Brad> SCSI chain, it's SCSI id, it's termination status, etc... Brad> 2. Replace the drive with an *identical* new one (same model Brad> number, same number of heads and platters, same firmware Brad> revision, etc...), and make sure that the new drive sits in the Brad> same physical position in the SCSI chain, has the same SCSI id, Brad> same termination status, etc... Brad> 3. Reboot the machine. Brad> 4. Bring up vinum (but don't attempt to mount the affected Brad> filesystem) and init the drive. Brad> 5. Use the "rebuildparity" command to start the recovery Brad> process. Brad> 6. When complete, try mounting the filesystem. Brad> Of course, I haven't actually tested this procedure (and I Brad> hope I never have to), so YMMV and use at your own risk. Problem is that we have to program for this case... if it doesn't work, then there's literally no point in vinum supporting RAID-5. The big problem is that the barrier-to-entry in this experiement is a few thousand $$, and people likely only get to work on it in the few weeks between equipment arriving and equipment going into production. Dave. -- ============================================================================ |David Gilbert, Velocet Communications. | Two things can only be | |Mail: dgilbert@velocet.net | equal if and only if they | |http://www.velocet.net/~dgilbert | are precisely opposite. | =========================================================GLO================ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message