From owner-freebsd-current@FreeBSD.ORG Tue Oct 2 21:22:02 2007 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 363D416A417; Tue, 2 Oct 2007 21:22:02 +0000 (UTC) (envelope-from cb@severious.net) Received: from ion.gank.org (ion.gank.org [69.55.238.164]) by mx1.freebsd.org (Postfix) with ESMTP id 181C013C465; Tue, 2 Oct 2007 21:22:02 +0000 (UTC) (envelope-from cb@severious.net) Received: by ion.gank.org (Postfix, from userid 1001) id E3D2811221; Tue, 2 Oct 2007 16:22:01 -0500 (CDT) Date: Tue, 2 Oct 2007 16:22:00 -0500 From: Craig Boston To: Pawel Jakub Dawidek Message-ID: <20071002212200.GC33227@nowhere> Mail-Followup-To: Craig Boston , Pawel Jakub Dawidek , Brooks Talley , freebsd-current References: <3939DF4C-D688-40B2-9A9A-1400E528AB07@online.no> <30482584.83141191352060006.JavaMail.root@zmail.illuminati.org> <20071002210544.GB5375@garage.freebsd.pl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20071002210544.GB5375@garage.freebsd.pl> User-Agent: Mutt/1.4.2.3i Cc: freebsd-current , Brooks Talley Subject: Re: ZFS corrupting data, even just sitting idle X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 02 Oct 2007 21:22:02 -0000 On Tue, Oct 02, 2007 at 11:05:44PM +0200, Pawel Jakub Dawidek wrote: > For me it's just quite interesting to observe how people live for years > with h/w corrupting their data and only putting ZFS there reveals the > problems. I wonder how many strange UFS-related problems/panics reported > over the years are due to buggy hardware. I suspect some of it could be due to ZFS making multi-disk (>3) setups more attractive. For example: A controller / drive combination may work perfectly fine with UFS on a single disk for years with no corruption, but add 1 or 2 more and it stresses out the power supply and causes strange things. I recall reading about bugs in cheap ATA controllers that were only tickled when trying to send simultaneous commands to multiple drives, for instance. IIRC a few of them showed up when gmirror started to become popular. Craig