Board index Linux

HELP: can't umount filesystem on shutdown!

HELP: can't umount filesystem on shutdown!

Post by Ben Log » Fri, 06 Oct 2000 11:10:02



I really need some help here...my system won't shut down properly any more.

I have two drives in my RedHat 6.2 box.  One of them had linux on it and the
other windoze.  I was running out of space in linux, so I wiped windows off
the other disk (/dev/hda1).  Then I created an ext2 filesystem on it and
mounted it in /usr/hda1.
Then I moved /usr/lib to /usr/hda1/lib and made a symlink so that the
libraries could still be found in /usr/lib.

Everything went fine, until I tried to shut my system down.  Then I got the
following error messages (the first message isn't an error, I included it
for context):

Turning off quotas  [ok]
Unmounting file systems umount2: Device or resource busy
umount: /usr/hda1: device is busy
umount2: Device or resource busy
umount: /usr: device is busy

No process references; use -v for the complete list
No automatic removal.  Please use umount /usr/hda1
INIT: no more processes left in this runlevel

I have another system with which I did the same thing, except that I mounted
the new partition (/dev/hdc1 in this case instead of /dev/hda1) directly as
/usr/lib--no symlinks that way.  It has the same problem now when I try to
shut it down.

Here's my /etc/fstab file:
/dev/hdb7               /                       ext2    defaults        1 1
/dev/hdb1               /boot                   ext2    defaults        1 2
/dev/hdb6               /home                   ext2    defaults        1 2
/dev/cdrom              /mnt/cdrom              iso9660 noauto,user,ro  0 0
/dev/hdb5               /usr                    ext2    defaults        1 2
/dev/fd0                /mnt/floppy             auto    noauto,user     0 0
/dev/hda1               /usr/hda1               ext2    defaults        1 2
none                    /proc                   proc    defaults        0 0
none                    /dev/pts                devpts  gid=5,mode=620  0 0
/dev/hdb8               swap                    swap    defaults        0 0

Any help would be greatly appreciated.
Ben Logan

 
 
 
Top

HELP: can't umount filesystem on shutdown!

Post by Giacomo Catenazz » Sun, 31 Dec 1899 09:00:00



> I really need some help here...my system won't shut down properly any more.

> I have two drives in my RedHat 6.2 box.  One of them had linux on it and the
> other windoze.  I was running out of space in linux, so I wiped windows off
> the other disk (/dev/hda1).  Then I created an ext2 filesystem on it and
> mounted it in /usr/hda1.
> Then I moved /usr/lib to /usr/hda1/lib and made a symlink so that the
> libraries could still be found in /usr/lib.

A problem is here! (I used to made symlink from /usr/lib to /sys/3.
A problem is the symlink behaviour (and it is not as bash works!)

In your /usr/hda1 (== /usr/lib) there are surelly some symlinks that
start
with ../.
E.g. Normally /usr/lib/X11 points to /usr/X11R6/lib/X11. In your
/usr/lib
this symlink is probably X11 -> ../X11/lib/X11. Thus when you point
/usr/lib
to /usr/hda1/lib the X directory become /usr/hda1/lib/../X11/lib/X11 (
== /usr/hda1/X11/lib/X11) which is different to /usr/X11/lib/X11).

This cause a lot of problems.

Quote:

> Everything went fine, until I tried to shut my system down.  Then I got the
> following error messages (the first message isn't an error, I included it
> for context):

> Turning off quotas  [ok]
> Unmounting file systems umount2: Device or resource busy
> umount: /usr/hda1: device is busy
> umount2: Device or resource busy
> umount: /usr: device is busy

> No process references; use -v for the complete list
> No automatic removal.  Please use umount /usr/hda1
> INIT: no more processes left in this runlevel

> I have another system with which I did the same thing, except that I mounted
> the new partition (/dev/hdc1 in this case instead of /dev/hda1) directly as
> /usr/lib--no symlinks that way.  It has the same problem now when I try to
> shut it down.

But this should happen! Probably is an RH error: there is a program in
shout-down procedure that uses the /usr/lib. (Any critical program
should useonly the library
in /lib).

Thus:   1- Don't use symlink!
        2- Update the RH or move some over dirs. (/var /usr/doc /usr/share,...)

        giacomo

- Show quoted text -

Quote:

> Here's my /etc/fstab file:
> /dev/hdb7               /                       ext2    defaults        1 1
> /dev/hdb1               /boot                   ext2    defaults        1 2
> /dev/hdb6               /home                   ext2    defaults        1 2
> /dev/cdrom              /mnt/cdrom              iso9660 noauto,user,ro  0 0
> /dev/hdb5               /usr                    ext2    defaults        1 2
> /dev/fd0                /mnt/floppy             auto    noauto,user     0 0
> /dev/hda1               /usr/hda1               ext2    defaults        1 2
> none                    /proc                   proc    defaults        0 0
> none                    /dev/pts                devpts  gid=5,mode=620  0 0
> /dev/hdb8               swap                    swap    defaults        0 0

> Any help would be greatly appreciated.
> Ben Logan


 
 
 
Top

1. 'shutdown' unable to umount partition, causing disk corruption

Please respond via e-mail, I don't read comp.os.linux.help.  If this is
indeed a bug I'll send it to whoever needs to know.

I'm having a problem with e2fs under kernel 0.99pl12 (I haven't been
able to install pl13 or 14, but that's another story).  Sometimes, at
shutdown time, shutdown(1) is unable to umount my /local (/dev/hda2)
e2fs partition, claiming the device is busy.  Consequently fsck(1) is
checking the entire partition every time I boot up, and it often finds
dangling i-nodes, which makes me break out in a cold sweat.  I've
already lost a filesystem recently due to corruption (yes I have
backups but it's still creepy).  I didn't have this problem under
0.99pl9, which I was using up until recently.

To try and find why the device was busy, I tried running most of the
programs I usually run, one at a time (as a user), then umounting and
re-mounting /local (as root, from a different console).  It worked
every time -- I was not able to find a case where it couldn't unmount.
Then I went back to my user activities for a few hours, then tried
umounting and BLAM, it can't umount /local.  My /local partition
contains everything that didn't come with SLS 1.03 -- my home directory
and various binaries and sources like gnuplot, xloadimage, rc, etc.
Running ps -aux shows there are no rogue processes around that may have
getpwd(3) in /local, and the shell I'm running in is under root in the
root directory.

Does anyone have any ideas why this is happening?  Have I found some
weird kernel/FS bug?  For the life of me I can't think of anything that
would leave the file system busy after all processes accessing it are
dead.  (Except maybe sockets used by term(1) or screen(1), even after
all remnants of both are gone?)

If I don't get a response in a day or two, I'll post to
comp.os.linux.development.  (Consider this a threat.  :-)

--
"The money spent on the Mars Observer was not sent to Mars. A few tons of metal
were.  The engineering lessons and experience from building the spacecraft and
instruments will remain on Earth to benefit future missions if we choose to

2. IOMEGA Ditto Tape Backup

3. "shutdown -r now" doesn't umount fs

4. Pan newsreader program - can't post

5. can't shutdown with 'shutdown'

6. FS: AIX Companion Book & X Users Guide

7. I can't umount a NFS filesystem

8. help with xterm and rlogin on command line

9. Can't umount a NFS filesystem

10. Can't umount a JFS filesystem

11. HELP! Unable to umount filesystem

12. shutdown doesn't unmount the filesystem

13. reboot/shutdown/halt don't unmount filesystems


All times are UTC
Board index