[Linux] Buffer I/O error on device hda1, logical block 0

Sat, 19 Jan 2008 22:46:48 -0600

------=_Part_6854_19716545.1200804408816
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

I am doing pretty heavy read/write work on my ubuntu box, and I noticed a
problem when I tried to delete some directories (rm -rf .../) either as a
regular user or as root and got an error saying that it was a read only file
system. Checking with ls -a showed the files were rw for the user. So I
rebooted, and I am now running into a cyclical problem.

When I run the usual boot, the boot process finds error in the file system
and forces a check. At 13.6% (consistently) I am getting I/O problems,
followed by the message: Buffer I/O error on device hda1, logical block 0
(or other blocks, around 10 messages and blocks, altogether), followed by
the instruction to run fsck manually.

I do, get some 10 error messages about I/O problems (right now I can't even
get there, so from memory it had to do with short reads or I/O problem with
the nodes). I then have the choice - ignore error (default is y, if I answer
n it exits fsck), followed by the option to force rewrite (again, y is the
default).

When fsck is done, if I run it again, it does not find any problems. I press
Ctrl-D as directed by the OS, it reboots, then finds the same problems and
we start again.

So now the questions:

1. Does that mean that the drive is dying or dead, or is it something that
can be fixed.
2. If the latter - how do I break the catch 22 and fix it?
3. Would running badblocks at any stage of the boot process help me get over
the problem?
4. Would a simple re-install of ubuntu resolve the issues?
5. Could it be a software fault (all these ftp, bit-torrent, http streams
going in and out of the system at once)?

Thanks,

Z.

-- 
Check out my web site - www.words2u.net

------=_Part_6854_19716545.1200804408816
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

I am doing pretty heavy read/write work on my ubuntu box, and I noticed a problem when I tried to delete some directories (rm -rf .../) either as a regular user or as root and got an error saying that it was a read only file system. Checking with ls -a showed the files were rw for the user. So I rebooted, and I am now running into a cyclical problem.
 When I run the usual boot, the boot process finds error in the file system and forces a check. At 13.6% (consistently) I am getting I/O problems, followed by the message: Buffer I/O error on device hda1, logical block 0 (or other blocks, around 10 messages and blocks, altogether), followed by the instruction to run fsck manually.
 I do, get some 10 error messages about I/O problems (right now I can&#39;t even get there, so from memory it had to do with short reads or I/O problem with the nodes). I then have the choice - ignore error (default is y, if I answer n it exits fsck), followed by the option to force rewrite (again, y is the default). 
 When fsck is done, if I run it again, it does not find any problems. I press Ctrl-D as directed by the OS, it reboots, then finds the same problems and we start again. So now the questions: 1. Does that mean that the drive is dying or dead, or is it something that can be fixed.
 2. If the latter - how do I break the catch 22 and fix it? 3. Would running badblocks at any stage of the boot process help me get over the problem? 4. Would a simple re-install of ubuntu resolve the issues? 
5. Could it be a software fault (all these ftp, bit-torrent, http streams going in and out of the system at once)? Thanks, Z. -- Check out my web site - <a href="http://www.words2u.net">
www.words2u.net</a>

------=_Part_6854_19716545.1200804408816--