From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751121AbWBHWFW (ORCPT ); Wed, 8 Feb 2006 17:05:22 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751144AbWBHWFV (ORCPT ); Wed, 8 Feb 2006 17:05:21 -0500 Received: from zproxy.gmail.com ([64.233.162.200]:502 "EHLO zproxy.gmail.com") by vger.kernel.org with ESMTP id S1751138AbWBHWFU convert rfc822-to-8bit (ORCPT ); Wed, 8 Feb 2006 17:05:20 -0500 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:mime-version:content-type:content-transfer-encoding:content-disposition; b=M0u2O38zmcz3jz6DFEWfIQ8bvsN50VQYPDVMK1uzOfaWAzlOkZklfjiAbLZ9++1cq9xPMIqfyA/J1mL79Vd/E4W+RYYumFwtQRbm8rBP1fDLuw5EkdEMOtZEn2i15UEO06SR2DbQ7YSHG02uMGAqk8TMnQBYTiYStiuV9CCZ1YE= Message-ID: Date: Wed, 8 Feb 2006 15:05:19 -0700 From: David Wilk To: linux-kernel@vger.kernel.org Subject: data retention after mkfs.ext3 MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org We deploy a linux system by rsync'ing a filesystem off of a squashfs FS on a CD. The deployment is automated and always involves an mkfs.ext3 prior to any file copy. We have had some reports that tend to indicate that some of the data from a previous install (of the same method) is corrupting at least some of the files on a new install. Old settings and random behavior has been seen, and breaking the RAID1 logical drive and rebuilding prior to install has always fixed the issue. It is my impression that mkfs.ext3 will make any existing data on a block device invisiable to the new ext3 FS. Is it possible that if the old FS is similar enough to the new one that the old data could somehow influence the new files written to a new ext3 FS? I know this *should* be impossible, but is it truly impossible?