From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 14 Sep 2001 19:47:22 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 14 Sep 2001 19:47:11 -0400 Received: from mx2.port.ru ([194.67.57.12]:61444 "EHLO smtp2.port.ru") by vger.kernel.org with ESMTP id ; Fri, 14 Sep 2001 19:46:58 -0400 From: Samium Gromoff <_deepfire@mail.ru> Message-Id: <200109150404.f8F44wV02060@vegae.deep.net> Subject: Re: tail corruptions: saga continues To: vs@namesys.com (Vladimir V. Saveliev) Date: Sat, 15 Sep 2001 04:04:01 +0000 (UTC) Cc: linux-kernel@vger.kernel.org In-Reply-To: <3BA080E2.39742E30@namesys.com> from "Vladimir V. Saveliev" at Sep 13, 2001 01:48:18 PM X-Mailer: ELM [version 2.5 PL6] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org " Vladimir V. Saveliev wrote:" > > Hi > > > Subject: tail corruptions: saga continues > > Date: Thu, 13 Sep 2001 00:33:23 +0000 (UTC) > > From: Samium Gromoff <_deepfire@mail.ru> > > To: reiser@namesys.com > > CC: linux-kernel@vger.kernel.org > > > > 2.4.7/3.x.0k-pre8 pair in _very_rare_ circumsistances > > You should get 3.x.0k-pre9. Ouch! Actually that was pre9... sorry for a mistake. > > > > > gives tails corruptions with --rebuild-tree. > > > Actually i found only one file filled with "." (ofcourse <4k) > > As a detail, on sudden reboots i have log tails filled with "." > > but it seems to be unrelated. > > It may look like this one file "." fill was caused by sudden > > inability to sync while corruption raised, but in fact at the moment > > when fs was corrupted i havent accessed this file for ages... (.c file) > > > > So, am I correct that before running reiserfsck --rebuild-tree the file > contained proper data and did not after it? Yep > Or the file appeared corrupted after sudden reboot/mount and it was not > touched before that sudden crash? > > Why did you decide to run --rebuild-tree? Did reiserfsck --check recommend you > to do that? To repair fs with some blocks set to zero. Zeroified blks are unrelated though... > > Thanks, > vs > > >