From: "kaefert@gmail.com" Subject: Re: e2fsck extremly slow after: EXT4-fs.. ext4_check_descriptors: Checksum for group .. failed Date: Fri, 9 Nov 2012 07:05:02 +0100 Message-ID: References: <20121109000156.GQ19977@thunk.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Cc: linux-ext4@vger.kernel.org To: "Theodore Ts'o" Return-path: Received: from mail-ea0-f174.google.com ([209.85.215.174]:63880 "EHLO mail-ea0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750771Ab2KIGFX (ORCPT ); Fri, 9 Nov 2012 01:05:23 -0500 Received: by mail-ea0-f174.google.com with SMTP id c13so1332432eaa.19 for ; Thu, 08 Nov 2012 22:05:22 -0800 (PST) In-Reply-To: <20121109000156.GQ19977@thunk.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: 2012/11/9 Theodore Ts'o > > Can you please run e2fsck from the command line, and capture the > output (i.e., using "script"). I really need the e2fsck output to > understand what is going on. The strace output is really not helpful. > > In general, you may be better off simply not trusting gparted to run > e2fsck and resize2fs for you. If there are no problems I'm sure it's > fine, but it's really hard to debug things if you insist on letting > gparted swallon all of the useful debugging output.... > > - Ted Hi there Ted! Thanks for the answer! Of course I understand that when you want to debug something you really gotta run it from the console, It's just when I started to run this, I did not think anything would go wrong, you never think it hits you ;) After e2fsck failed for the first time (where I don't know why, since gparted crashed after I tried to save the details), I started to run e2fsck manually, and since the -p option made him cancel the run I started a third run in interactive mode. I've posted the console output of this second and third run as an update to my question at serverfault.com (see http://serverfault.com/questions/446074/e2fsck-extremely-slow-although-enough-memory-exists - start reading at "UPDATE4") The 3rd run is still running (since about 20 hours now) and showing the same pattern as the first run that failed after 78 hours. Thanks for looking at this, Thomas K.