Bug#490984: marked as done (linux-image-2.6-686: / is busy causes EXT3-fs: INFO: recovery required on readonly filesystem)
Your message dated Thu, 17 Jul 2008 09:33:04 +0200
with message-id <20080717073304.GC5691@stro.at>
and subject line Re: Bug#490984: linux-image-2.6-686: / is busy causes EXT3-fs: INFO: recovery required on readonly filesystem
has caused the Debian Bug report #490984,
regarding linux-image-2.6-686: / is busy causes EXT3-fs: INFO: recovery required on readonly filesystem
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 this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@bugs.debian.org
immediately.)
--
490984: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=490984
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems
--- Begin Message ---
Package: linux-image-2.6-686
Version: 2.6.25+14
Severity: normal
Thanks for maintaining debian's kernel packages.
They're a remarkable technology in more ways than
one.
Here's how I duplicate the bug.
1.) $ shutdown -r now
2.) ctl-alt-F7
3.) look quickly for something like
"mount: / is busy"
4.) After the system has rebooted, if the root
filesystem is EXT3, look for
EXT3-fs: INFO: recovery required on readonly filesystem.
in /var/log/messages.
If the root filesystem is EXT2, I get a long,
slow fsck recovery.
It seems to me that this can be partially
explained by the shutdown command running a
script named
/etc/init.d/umountroot
which calls the "mount" command to remount the
root file system as read only.
It fails, and complains that
"/ is busy"
Since the root file system wasn't cleanly
unmounted, it'll be recovered during the next
boot.
I investigated other processes interfering with
the root file system, without result.
However, it seems that inserting
"cat /proc/mounts"
just before the (re)mount command in
/etc/init.d/umountroot
usually allows mount to complete.
As an aside, the words "mount" and "remount" occur
hundreds of times in the changelog for 2.6.26.
Frankly, I'm a little worried about data
corrutpion.
Thanks,
Kingsley
-- System Information:
Debian Release: lenny/sid
APT prefers unstable
APT policy: (990, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Shell: /bin/sh linked to /bin/bash
Kernel: Linux 2.6.25-2-686
Locale: LANG=en_US, LC_CTYPE=en_US (charmap=ISO-8859-1)
Versions of packages linux-image-2.6-686 depends on:
ii linux-image-2.6.25-2-686 2.6.25-6 Linux 2.6.25 image on PPro/Celeron
linux-image-2.6-686 recommends no packages.
-- no debconf information
--- End Message ---
--- Begin Message ---
On Wed, 16 Jul 2008, Kingsley G. Morse Jr. wrote:
> Hi Maks,
>
> Thank you for replying quickly.
>
> On 07/16/08 10:47, maximilian attems wrote:
> [...]
> > hmm how is the bug you reported related to the kernel?
> [...]
>
> To be fair, I'm not sure that the kernel is to
> blame.
closing as bug is unreproducible and certainly not a kernel bug.
> It could be data corruption (I'm recovering from a
> bad CPU chip), incompatible package versions, or
> even something else I suppose.
>
> However, I suspect the kernel because the
> changelog at kernel.org for the next version,
> 2.6.26, mentions the words "mount" and "remount"
> hundreds of times, which leads me to suspect bugs
> and active development.
please read the kernelnewbies page about 2.6.26
it will tell you a lot about ro bind mounts.
nothing to be worried about.
best regards
--
maks
--- End Message ---
Reply to: