07-09-04 04:56 PM
In article <e10cccdf.0407081836.7283c96@posting.google.com>, furufuru@ccsr.u-tokyo.ac.jp (Ry
o Furue) writes:
> Hello all,
>
> On a Solaris 9 machine, I'm having an I/O error:
>
> $ cd /some/directory
> $ echo * | (fgrep -s ' msg00175.html '; echo $?)
> 0 #--> file does exist.
> $ ls msg00175.html
> msg00175.html: I/O error #--> cannot stat
> $ truss ls msg00175.html
> [...]
> lstat64("msg00175.html", 0xFFBFFB50) Err#5 EIO
> [...]
> $
>
> There's no related error message either in /var/log/syslog or
> /var/adm/messages . The message "I/O error" isn't very helpful;
> the message doesn't tell you what's wrong with the file.
>
> I'm wondering how I can obtain more information about the
> (apparently) broken file beyond "I/O error". The only fix
> may be just to delete the file, but
> 1) I'm wondering if it's possible to recover the file, and
> 2) I'm curious.
>
> Thank you,
> Ryo
If it is a local disk ('df -F' tells it),
try to umount it and run fsck.
(otherwise 'reboot -- -s' then run fsck).
If it is on the network, umount/mount it.
--
Michael Tosch
IT Specialist
HP Managed Services Germany
Phone: +49 2407 575 313
Mail: michael.tosch:hp.com
[ Post a follow-up to this message ]
|