You are not logged in.
Pages: 1
on 50% of the times when I shutdown or reboot my computer, I get this message thrown at me just before the computer powers off or reboots:
mount: / is busy
Power Off
and at next boot it says something like:
root system was not shutdown cleanly.
recovery-mode. read/write access will enabled during recovery
and it sucks lots of my time there and it happens lots of times. I simply use poweroff/reboot commands then why do I get this message so often. this problem started with the 1st release of Don't Panic and it is still here.
Offline
That happens to me too, but the recovery thing only takes 1 or 2 seconds, IIRC. At least not like a disk check every 38 reboots/1xx days.
Offline
I have a feeling some background daemon/app doesn't shut down properly. Try killing different processes and shutting down - you should be able to narrow things down
Offline
I had seen this issue when not exiting a chroot etc.. Anyway: the initscripts shouldn't just poweroff or reboot when a device could not be unmounted.
Offline
... Anyway: the initscripts shouldn't just poweroff or reboot when a device could not be unmounted.
I agree.
I had a few of those unclean shutdowns too, it was usually after a system upgrade, especially after I did the upgrade when the system was already running for some time. I don't think I ever experienced this when I did pacman -Syu right after system startup.
Offline
Perhaps this is a hit or miss thing? I once upgraded two separate machines, and one of them had this strange issue, while the other continued to reboot and function normally.
Offline
I think the only times where I've seen this, is after a glibc upgrade ("reloading init").
1000
Offline
That happens to me too, but the recovery thing only takes 1 or 2 seconds, IIRC. At least not like a disk check every 38 reboots/1xx days.
1-2 seconds ?
you are lucky then . for me, it eats my time, not to mention the unnecesssary frustration.
Offline
I think the only times where I've seen this, is after a glibc upgrade ("reloading init").
it SURELY happens onoce when you try to reboot/poweroff after upgrade <pacman -Syu> but to me it also happens at other times. you can say in case of 15% of shutdowns.
Offline
ex.: /dev/hda3 / ext3 defaults 0 0
Try to change to sda3
http://ispconfig.lt - ISPConfig 3 based hosting. Coming Soon!
Offline
I don't have any /dev/hd* in /etc/fstab and still had a few of those can't remount issues.
Perhaps we should first figure out what is going on just before unclean reboot:
--- rc.shutdown 2007-12-10 23:39:44.000000000 +0100
+++ rc.shutdown.new 2007-12-10 23:38:54.000000000 +0100
@@ -111,6 +111,14 @@
stat_busy "Remounting Root Filesystem Read-only"
/bin/mount -n -o remount,ro /
+if [ $? -ne 0 ]; then
+ /bin/ps aux >>"/root/ps-aux.$(/bin/date +%Y-%m-%d)" 2>&1
+ /sbin/killall5 -15 >> "/root/sigterm.$(/bin/date +%Y-%m-%d)" 2>&1
+ /bin/sleep 5
+ /sbin/killall5 -9 >> "/root/sigkill.$(/bin/date +%Y-%m-%d)" 2>&1
+ /bin/sleep 1
+ /bin/mount -n -o remount,ro /
+fi
stat_done
# Power off or reboot
Then after a failed read-only remount of / filesystem one would know which processes were running and if killall5 gave any errors. Maybe sleep time should be increased to a few minutes so data would more likely be written to disk and not just cached.
Offline
it happened to me a lot of time that mount: / is busy ... poweroff.
generally it got no fsck at reboot (or i didn't notice)
but sure it is rather disturbing to see that the system is being shutdown in such unsafely way...
i am getting used to it and still not have lost any data ... but ...
is it related to the problem about daemon rc files?? :
when you launch manually a /etc/rc.d/ files it is not cleanly shutdown when the pc shutdown
that's an old bug still not closed ?
or when you manually shutdown a daemon with /etc/rc.d file it is still shutdown at shutdown time while it should not ?
i have not tested that idea
Last edited by solstice (2007-12-15 12:23:34)
Offline
Pages: 1