Your message dated Mon, 1 Oct 2007 12:28:38 -0600
with message-id <[EMAIL PROTECTED]>
and subject line Bug#440828: umount fails to remove entry in /etc/mtab when
unmounting
has caused the attached Bug report to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what I am
talking about this indicates a serious mail system misconfiguration
somewhere. Please contact me immediately.)
Debian bug tracking system administrator
(administrator, Debian Bugs database)
--- Begin Message ---
Package: mount
Version: 2.12r-19
We have applications that will automount 25-40 mounts all at once.
When the automount timer expires it will umount most of these all at
once. Very often, some of them will unmount but umount fails to
remove the entry from /etc/mtab. The following error appears in
daemon.log:
Aug 22 18:16:37 linus121 automount[8030]: >> Cannot create link /etc/mtab~
Aug 22 18:16:37 linus121 automount[8030]: >> Perhaps there is a stale lock file?
The next time automount tries to mount the stale entry, it fails to
mount stating that it is already mounted.
To fix this issue, I modified the following code in util-linux-2.12r
source code:
~/util-linux-2.12r/mount/fstab.c in routine lock_mtab:
/* Limit the number of iterations - maybe there
still is some old /etc/mtab~ */
if (tries++ > 3) {
if (tries > 5)
die (EX_FILEIO, _("Cannot create link %s
\n"
"Perhaps there is a stale lock file?
\n"),
MOUNTED_LOCK);
sleep(1);
}
What I did was increase the number of tries above to > 5 on the first
loop and > 10 on the second one that sleeps. Basically, with so many
umounts going on at once, the code was not trying enough times to gain
the lock on mtab before it gave up. By increasing the number of tries
before giving up, umount had enough time to obtain the lock and update
all the mtab entries. Please consider upping these values in the main
package if you don't see any harm in doing so. Thanks.
Rich Wohlstadter
Sys. Admin
Genome Sequencing Center
Washington University of Saint Louis
--- End Message ---
--- Begin Message ---
Version: 2.13
This was fixed in 2.13
lamont
--- End Message ---
