Problems on start of my system

[Greg 'groggy' Lehey]

[Format recovered--see http://www.lemis.com/email/email-format.html]

Computer output wrapped.

On Saturday,  5 November 2005 at  2:02:49 -0200, alicornio wrote:
Content-Description: Mail message body
> Hi all
>
> I have a problem on start of my system, appear this mesage:
>
> Staring file system checks:
> /dev/ad0s1a: 1304 files, 18242 used, 300053 free (1149 frags, 37363 blocks, 0.4% fragmentation)
> /dev/ad0s1g: DEFER FOR BACKGROUND CHECKING
> /dev/ad0s1f: DEFER FOR BACKGROUND CHECKING
> /dev/ad0s1d: DEFER FOR BACKGROUND CHECKING
> /dev/ad0s1e: FILE SYSTEM CLEAN; SKIPPING CHECKS
> /dev/ad0s1e: clean, 57112 free (40 frags, 7134 bloks, 0,1% fragmentation)
> WARNING: /home was not properly dismounted
> ad0: FAILURE - READ-DMA status=51<READ,DSC,ERROR> ERROR=40<UNCORRECTABLE> LBA=2398527
> mount: /dev/ad0s1: input/output error
> WARNING: /usr was not properly dismounted
> WARNING: /var was not properly dismounted
> mounting /etc/fstab filesystems failed, start up aborted
> ad0: FAILURE - READ-DMA status=51<READ,DSC,ERROR> ERROR=40<UNCORRECTABLE> LBA=2191743
> boot interruped
> enter full pathname of shell or return for /bin/sh:
>
> I type something and the system reboot after this mesages:

If you typed something, you should say what.

But in this case, you have two problems.  Most people replying to this
thread have concentrated on the unclean umount.  The real issue seems
to be the hard errors on the disk:

  <UNCORRECTABLE> LBA=2398527
  <UNCORRECTABLE> LBA=2191743

The question is: which file system?  That's not clear from the output.

There's a very good chance that you need a new disk, but the question
is how to recover the data.  You have a choice of what to do now.  The
safest approach  is:

 - buy a new disk and install your chosen version of FreeBSD on it.
 - put the old disk in the system in some other position (slave or
   second IDE channel, for example).
 - mount the file systems read-only, without running fsck
 - copy across the data that you need.

If you choose to keep the existing disk, you can try this:

   - run fsck -u on each file system in turn:

     fsck -y /
     fsck -y /usr
     fsck -y /var
     fsck -y /home

     If the system panics, you will know which file system was
     affected.  The ones that have already been fsck'ed will stay that
     way; you don't need to repeat that.

  - If it succeeds, you're done.  Otherwise you have at least narrowed
    down the problem to a specific file system.  You can mount it
    read-only and back up the contents.  Then, if it's not the root
    file system (this won't work for the root file system), do
    (assuming that it's /dev/ad0s1g, which you don't appear to have
    :-),

      dd if=/dev/zero of=/dev/ad0s1g bs=128k conv=sync seek=9

    This writes zeroes to the entire partition, with the exception of
    the disk label.  Obviously this destroys all data on the
    partition.  It's possible that the drive will then succeed in
    writing the data and clean up the soft errors.  If this works, you
    can run newfs on the partition and restore the data.

Greg
--
When replying to this message, please copy the original recipients.
If you don't, I may ignore the reply or reply to the original recipients.
For more information, see http://www.lemis.com/questions.html
See complete headers for address and phone numbers.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-questions/attachments/20051106/3d2b832c/attachment.bin