From owner-freebsd-questions@FreeBSD.ORG Fri Jun 11 22:23:42 2010 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 EE7A61065677 for ; Fri, 11 Jun 2010 22:23:42 +0000 (UTC) (envelope-from kgc@corp.sonic.net) Received: from a.mail.sonic.net (a.mail.sonic.net [64.142.16.245]) by mx1.freebsd.org (Postfix) with ESMTP id B18FA8FC14 for ; Fri, 11 Jun 2010 22:23:42 +0000 (UTC) Received: from meno.noc.sonic.net (meno.noc.sonic.net [64.142.18.2]) by a.mail.sonic.net (8.13.8.Beta0-Sonic/8.13.7) with ESMTP id o5BLZqqa023889 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO) for ; Fri, 11 Jun 2010 14:35:52 -0700 Received: from meno.noc.sonic.net (localhost [127.0.0.1]) by meno.noc.sonic.net (8.14.3/8.14.3) with ESMTP id o5BLZqR6010797 for ; Fri, 11 Jun 2010 14:35:52 -0700 (PDT) (envelope-from kgc@corp.sonic.net) Received: (from kgc@localhost) by meno.noc.sonic.net (8.14.3/8.14.3/Submit) id o5BLZqVf010796 for freebsd-questions@freebsd.org; Fri, 11 Jun 2010 14:35:52 -0700 (PDT) (envelope-from kgc@corp.sonic.net) Date: Fri, 11 Jun 2010 14:35:52 -0700 From: Kelsey Cummings To: freebsd-questions@freebsd.org Message-ID: <20100611213552.GB3154@corp.sonic.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline X-PGP-Key: http://sonic.net/~kgc/gpgkey.txt User-Agent: Mutt/1.5.19 (2009-01-05) Subject: Areca RAID Failure? 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, 11 Jun 2010 22:23:43 -0000 This isn't specifically freebsd related but I'm fishing to see if anyone has observered similar behavior from and Areca raid controller before. We're already in touch with their support... Last night a disk failed on a 7 disk raid-6 array on a ARC-1220 with 1TB WD REmumble disks. This is certainly normal enough, except that rather than taking the normal ~30 hours to rebuild array after a failure it appears to have added the spare disk directly into the array and started servicing reads from it without rebuilding it from parity! This obviously seriously scrambled the filesystem on it and sorta defeats the whole point of H/W raid in the first place. XFS recovered well enough, but most files are large enough to span a stripe and are corrupted for it. It's currently running a check and is finding lots of errors, I am optomistic that it's check routine might rebuild the data from parity but am glad this occured on a log archiving volume so it isn't a great loss and we don't have to restore from backups anyway. Anyone else seen such amazing examples of FAIL from Areca's? We've got 50 or so 3wares in production and in the past 8 years have only seen one 3ware tank - it destroyed the filesystem on it's way but also complained on the way out and wouldn't initialize since it failed its internal diags. Performance issues or not, at least they do their job. -K