From: Zlatko Calusic Subject: Re: e2fsck not fixing deleted inode referenced errors? Date: Wed, 01 Oct 2014 08:44:00 +0200 Message-ID: <542BA2B0.4040604@bitsync.net> References: <542AEED4.5050303@bitsync.net> <20140930183012.GA9942@birch.djwong.org> <542AF9B8.2090800@bitsync.net> <20140930195408.GD17142@thunk.org> <542B1220.8020208@bitsync.net> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Cc: "Darrick J. Wong" , linux-ext4@vger.kernel.org To: Theodore Ts'o Return-path: Received: from bitsync.net ([80.83.126.10]:34571 "EHLO bitsync.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750868AbaJAGoD (ORCPT ); Wed, 1 Oct 2014 02:44:03 -0400 In-Reply-To: <542B1220.8020208@bitsync.net> Sender: linux-ext4-owner@vger.kernel.org List-ID: On 30.09.2014 22:27, Zlatko Calusic wrote: > > Unless you have other ideas, I will run badblocks. Although, as ext4 fs > is on /dev/md2, I think I should run it on /dev/md2 only? Do you really > mean to run it on /dev/sda2, /dev/sdb2 - underlying devices? I'm not > sure how MD would cope with it. > > But, I'm pretty sure that it will come out clean. The md check I did > last night would surely detected bad blocks if there were any. Or not? So, I ran badblocks on both underlying devices, and both came up clean, as I suspected would happen. Then I ran debugfs -w /dev/md2 and unlinked 7 inodes that were problematic. Afterwards e2fsck found those inodes and connected them to lost+found. Second e2fsck run finished clean. And at this time I see no errors on the fs. This is definitely the case where e2fsck can't see or repair fs errors that are real. Regards, -- Zlatko