From owner-freebsd-questions@FreeBSD.ORG Fri Feb 15 12:44:22 2008 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 521F616A481 for ; Fri, 15 Feb 2008 12:44:22 +0000 (UTC) (envelope-from bc@default.co.yu) Received: from smtp1.default.co.yu (bv.default.co.yu [87.237.201.138]) by mx1.freebsd.org (Postfix) with SMTP id 8EEBF13C45A for ; Fri, 15 Feb 2008 12:44:20 +0000 (UTC) (envelope-from bc@default.co.yu) Received: (qmail 21976 invoked by uid 89); 15 Feb 2008 12:44:19 -0000 Received: from unknown (HELO ?10.5.7.109?) (bc@default.co.yu@213.198.226.187) by smtp1.default.co.yu with SMTP; 15 Feb 2008 12:44:19 -0000 Message-ID: <47B58923.90603@default.co.yu> Date: Fri, 15 Feb 2008 13:44:19 +0100 From: Bogdan Culibrk Organization: Default Inc. User-Agent: Thunderbird 2.0.0.9 (Windows/20071031) MIME-Version: 1.0 To: Wojciech Puchar References: <1202313434.5415.209.camel@serafim> <20080206174915.O4330@wojtek.tensor.gdynia.pl> In-Reply-To: <20080206174915.O4330@wojtek.tensor.gdynia.pl> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-questions@freebsd.org Subject: Re: Some ZFS experience X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 15 Feb 2008 12:44:22 -0000 Hi and sorry for resurrecting this old topic Wojciech Puchar wrote: >> >> What happened: >> 1. Base system running on gmirror volume consisting of 4 mirrors booted >> up normally in degraded mode with 3 of 4 drives online. >> 2. /data powered by raidz1 zfs was showing 3 drives, 2 online and 1 >> faulted. zfs list was showing that volume is unavailable due lack of >> spares. > > you told about having raidz over 4 drives. > so while it reports 2 online and 1 faulted, not 3 online and 1 faulted? As I mentioned in first email I did have 4 drives, but new host wasnt able to accept all 4 drives so I gave a shot with 3 drives expecting that it should work. The day after I got new cablings and connected all 4 drives and all worked like charm. Point is that raidz1 refused to work in somewhat degraded mode with 3 drives. It showed 1 out of 3 drive as faulted :/ IIRC, I didnt do: # zpool export tank before migration. It might be source of the problem, but I'm not sure. So lesson learned: always export your ZFS volumes before doing anything like that :) > >> What was wrong when 3 drives were connected? Wasnt supposed to raidz1 >> survive lack of 1 drive? Or I did something wrong there? > > yes it should work normally. in case of raid-z - with just a bit > slower speed according to ZFS theory. >