From owner-freebsd-current@FreeBSD.ORG Sun Aug 24 12:40:58 2003 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0578416A4BF for ; Sun, 24 Aug 2003 12:40:58 -0700 (PDT) Received: from lakemtao07.cox.net (lakemtao07.cox.net [68.1.17.114]) by mx1.FreeBSD.org (Postfix) with ESMTP id 13B5443FEA for ; Sun, 24 Aug 2003 12:40:57 -0700 (PDT) (envelope-from mezz7@cox.net) Received: from mezz.mezzweb.com ([68.103.32.11]) by lakemtao07.cox.net (InterMail vM.5.01.06.04 201-253-122-130-104-20030726) with ESMTP id <20030824194055.VYYN18874.lakemtao07.cox.net@mezz.mezzweb.com>; Sun, 24 Aug 2003 15:40:55 -0400 Date: Sun, 24 Aug 2003 14:39:59 -0500 To: Thomas Gutzler Content-Type: text/plain; charset=utf-8; format=flowed References: <20030824192611.GA14668@brain.hadiko.de> From: Jeremy Messenger MIME-Version: 1.0 Message-ID: In-Reply-To: <20030824192611.GA14668@brain.hadiko.de> User-Agent: Opera7.11/Linux M2 build 406 cc: FreeBSD Current Subject: Re: recover superblock X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 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: Sun, 24 Aug 2003 19:40:58 -0000 On Sun, 24 Aug 2003 21:26:11 +0200, Thomas Gutzler wrote: > Hi, > > I just lost all of my filesystems on my 5.1 box. > I was running mnogosearch's indexer on my website while several errors > occured. I guess there were 320 of these errors, cause the last message > was "last message repeated 320 times". Unfortunately /var is lost as > well, so I don't have more error messages. > > kernel: swap_pager: indefinite wait buffer: device: ad0s1b, blkno: 23536, > size: 4096 > kernel: ad0: WRITE command timeout tag=0 serv=0 - resetting > kernel: ata0: resetting devices .. > kernel: done > kernel: bad block 4190320178290426368, ino 402986 > kernel: pid 42 (syncer), uid 0 inumber 402986 on /usr: bad block > > I decided to reboot into singleuser and run fsck. I noticed, that the > buffers couldn't be flushed before restarting and the system couldn't > boot. It said: "not ufs" > > I plugged the harddisk into another computer and tried to run fsck. > It couldn't find a superblock, used an alternative one and coredumped > while trying to repair the first error. Before dumping, it said > something like "Couldnt allocate 3900000000 bytes for inoinfo". I can > only remember the value of around 3.9G. > This happened on all filesystems of ad0. I could mount the > filesystem but not access it, but df said: Used: -2.8T (on a 40G disk). > I surrendered on that and ran newfs. > > I don't know why, but fsck works on ad1. > fsck said something like "softupdate inconsistency" while trying to > repair the filesystem which obviously didn't completely work. I have lost over 1,000 to 2,000 emails like three times when the Nvidia driver crashed and had to do the power off and on to get it boot to let the fsck background do the job. Yesterday, I CVSup'ed and did the buildworld to update my -CURRENT. I got panic and reboot; first time I see the same message above, 'softupdate inconsistency' and I had to ran fsck by manual. After that, I discovered that my /var/, /compat/linux/ and very little stuff in my home directory have been destroyed. I can't produce the panic anymore so can't really give any help that much now; it was only panic at once for unknown reason. Just want to add in here that you aren't only person in here. Cheers, Mezz > # fsck /dev/ad1s1e > ** /dev/ad1s1e > Cannot find file system superblock > > LOOK FOR ALTERNATE SUPERBLOCKS? [yn] y > > USING ALTERNATE SUPERBLOCK AT 32 > ** Last Mounted on > ** Phase 1 - Check Blocks and Sizes > ** Phase 2 - Check Pathnames > ** Phase 3 - Check Connectivity > ** Phase 4 - Check Reference Counts > ** Phase 5 - Check Cyl groups > 19093 files, 42115174 used, 33631703 free (6471 frags, 4203154 blocks, > 0.0% fragmentation) > > # mount /dev/ad1s1e /mnt/ > mount: /dev/ad1s1e on /mnt: incorrect super block > > same thing using fsck -b 32 or any other alternative > > Google somewhere advised to use tunefs, but: > # tunefs -A /dev/ad1s1e > tunefs: /dev/ad1s1e: could not read superblock to fill out disk > > Even trying to copy an alternate superblock using dd faied. > > Do you have any ideas, how I could recover the data on this disk ? > > Cheers, > Tom -- bsdforums.org 's moderator, mezz.