Automounting no longer works
Warren
never bothered to learn the details of how. Following a power failure,
however, I can now only access removable media by manually mounting
and automounting (e.g. mount /media/cdrom/ as root).
fstab reads:
[root@localhost warren]# cat /etc/fstab
/dev/sda1 / xfs relatime 1 1
/dev/sda6 /home xfs relatime 1 2
/dev/cdrom /media/cdrom auto
umask=0022,users,iocharset=utf8,noauto,ro,exec 0 0
/dev/fd0 /media/floppy auto
umask=0022,users,iocharset=utf8,noauto,exec,flush 0 0
none /proc proc defaults 0 0
none /tmp tmpfs defaults 0 0
/dev/sda5 swap swap defaults 0 0
I found haldaemon to be stopped, so I started it. Inserting a CD in
the
drive, it was detected and KDE asked me what I wanted done. I asked
for a Konqueror file manager window, which produced the error message
"Feature only available with HAL", which is interesting given that HAL
remains running.
haldaemon no longer starts at boot as it did before, BTW, despite
being configured in MCC to start at boot. I can start it manually
without complaint.
Programs that access the CD burner, such as Brasero, no longer offer
burning as an option: now all they offer is to create a CD image to
disk.
I usually Google and RTFM before calling for help here, but with this
issue I can't even think of any keywords to search for save "Feature
only available with HAL", which got a lot of hits but I didn't see
anything obviously helpful for me.
I do notice a recent thread in this NG, "Impossible to read CD after
an update of my 2008.0", which may well be the same as my problem and
which suggests that a recent update, rather than a power failure
induced coruption, may be to blame -- but no solution there as yet.
Any ideas on what I should be looking at to fix this issue?
P.S.: I apologize for posting via Google Groups. At the moment both my
primary and secondary newsservers are down.
--
Warren Post
Santa Rosa de Copán, Honduras
http://srcopan.vze.com/
ray
You'll probably find more information if you look for 'hot plug' than
automount, because what you describe is not automount. Auotomount is a
method of setting up remote filesystems so that they automatically mount
when referenced and then unmount after a period of inactivity.
David W. Hodgins
> haldaemon no longer starts at boot as it did before, BTW, despite
> being configured in MCC to start at boot. I can start it manually
> without complaint.
Sounds like the power failure may have corrupted some files, in which
case it may be easier to reinstall. Assuming you don't want to do that,
the first step is to figure out why haldaemon is failing to start, at
bootup. What does /var/log/prcsys.log show?
Regards, Dave Hodgins
--
Change nomail.afraid.org to ody.ca to reply by email.
(nomail.afraid.org has been set up specifically for
use in usenet. Feel free to use it yourself.)
Warren
wrote:
> Sounds like the power failure may have corrupted some files, in which
> case it may be easier to reinstall.
You are probably right, and I may end up doing that, but in the
meantime, if I can learn something, so much the better.
> the first step is to figure out why haldaemon is failing to start, at
> bootup. What does /var/log/prcsys.log show?
[root@localhost warren]# less /var/log/prcsys.log | grep haldaemon
WARNING! Loop detected anacron -> haldaemon
-> Ignoring requires haldaemon
WARNING! Loop detected haldaemon -> messagebus
...or do you want the full ungreped output?
David W. Hodgins
>
> [root@localhost warren]# less /var/log/prcsys.log | grep haldaemon
> ...or do you want the full ungreped output?
Snipping out the part from shorewall shouldn't leave more than about 50 lines.
Warren
> Snipping out the part from shorewall shouldn't leave more than about 50 lines.
Here you go:
[root@localhost warren]# cat /var/log/prcsys.log
Activating Compat-mode for non LSB-script nessusd
Activating Compat-mode for non LSB-script webmin
WARNING! Loop detected anacron -> haldaemon
-> Ignoring requires haldaemon
WARNING! Loop detected avahi-daemon -> messagebus
-> Ignoring requires messagebus
WARNING! Loop detected avahi-daemon -> network-up
-> Ignoring requires network-up
WARNING! Loop detected consolekit -> messagebus
-> Ignoring requires messagebus
WARNING! Loop detected cups -> network-up
-> Ignoring requires network-up
WARNING! Loop detected dm -> xfs
-> Ignoring requires xfs
WARNING! Loop detected dm -> network-up
-> Ignoring requires network-up
WARNING! Loop detected haldaemon -> messagebus
-> Ignoring requires messagebus
WARNING! Loop detected httpd -> network-up
-> Ignoring requires network-up
WARNING! Loop detected keytable -> netfs
-> Ignoring requires netfs
WARNING! Loop detected ksysguard -> network-up
-> Ignoring requires network-up
WARNING! Loop detected lisa -> network-up
-> Ignoring requires network-up
Checking if partitions have enough free diskspace:
[ OK ]
Starting atd:
[ OK ]
Starting numlock:
[ OK ]
Starting system logger:
[ OK ]
Starting kernel logger:
[ OK ]
Starting resolvconf:
[ OK ]
Starting acpi daemon:
[ OK ]
ALSA driver (version 1.0.14) is already running.Doing alsactl to
restore mixer
settings:
[ OK ]
Starting crond:
[ OK ]
Loading keymap: es-latin1
[ OK ]
The BackSpace key sends: ^?
[ OK ]
Starting ksysguardd:Starting HAL daemon:
[FAILED]
Starting Avahi daemon:
[FAILED]
Starting ConsoleKit:
[ OK ]
Starting display manager:
[ OK ]
Starting lisa:
[ OK ]
error: libhal_ctx_set_dbus_connection:
org.freedesktop.DBus.Error.FileNotFound: Failed to connect to socket /
var/run/dbus/system_bus_socket: No such file or directory
Starting anacron:
[ OK ]
Loopback device ('lo', 127.0.0.1) needed by CUPS, starting it ...
Adding loopback device to routing table ...
Starting CUPS printing system:
[ OK ]
Starting httpd: httpd: Could not reliably determine the server's fully
qualified domain name, using 127.0.0.1 for ServerName
[ OK ]
Bringing up loopback interface:
[ OK ]
Bringing up interface eth0:
[ OK ]
Waiting for network to be up
[ OK ]
Mounting other filesystems:
[ OK ]
Starting X Font Server:
[ OK ]
Starting system message bus:
[ OK ]
Syncing time for ntpd:
[FAILED]
Starting ntpd:
[FAILED]
Compiling...
Initializing...
** snip shorewall stuff **
Starting mandi daemon: mandi_daemon_add_watch(): READABLE
unable to open white list file
[ OK ]
Total time: 52 seconds
Feel free to comment on anything messed up you see, not just what is
directly relevant to my immediate problem. I'm trying to use this as a
learning opportunity.
David W. Hodgins
> Starting ksysguardd:Starting HAL daemon:
> [FAILED]
> Starting Avahi daemon:
> [FAILED]
> error: libhal_ctx_set_dbus_connection:
> org.freedesktop.DBus.Error.FileNotFound: Failed to connect to socket /
> var/run/dbus/system_bus_socket: No such file or directory
> Starting httpd: httpd: Could not reliably determine the server's fully
> qualified domain name, using 127.0.0.1 for ServerName
> Syncing time for ntpd:
> [FAILED]
> Starting ntpd:
> [FAILED]
Please confirm that the following directory does exist ...
ll -d /var/run/dbus
drwxr-xr-x 2 root root 30 2008-08-02 13:08 /var/run/dbus/
Unless you normally use shared printers etc, disable avahi with
chkconfig avahi-daemon off
Looks to me like something has messed up the parallel services startup.
Try re-booting with the kernel paramater "nopinit", and then post the
prcsys.log again.
Warren
wrote:
> Please confirm that the following directory does exist ...
> ll -d /var/run/dbus
Yes, it does:
[warren@localhost ~]$ ll -d /var/run/dbus
drwxr-xr-x 2 root root 30 2008-08-03 16:20 /var/run/dbus/
> Unless you normally use shared printers etc, disable avahi with
> chkconfig avahi-daemon off
Done; thanks for the tip.
> Try re-booting with the kernel paramater "nopinit", and then post the
> prcsys.log again.
That parameter hangs the boot at starting HAL daemon. For what it's
worth, here's the prcsys.log from the successive boot (without
nopinit):
[root@localhost warren]# cat /var/log/prcsys.log
Activating Compat-mode for non LSB-script nessusd
Activating Compat-mode for non LSB-script webmin
WARNING! Loop detected consolekit -> messagebus
-> Ignoring requires messagebus
WARNING! Loop detected haldaemon -> messagebus
-> Ignoring requires messagebus
WARNING! Loop detected mandi -> messagebus
-> Ignoring requires messagebus
Checking if partitions have enough free diskspace:
[ OK ]
Starting numlock:
[ OK ]
Starting atd:
[ OK ]
Starting system logger:
[ OK ]
Starting kernel logger:
[ OK ]
Starting resolvconf:
[ OK ]
Starting ConsoleKit:
[ OK ]
Starting acpi daemon:
[ OK ]
ALSA driver (version 1.0.14) is already running.Doing alsactl to
restore mixer
settings:
[ OK ]
Bringing up loopback interface:
[ OK ]
Bringing up interface eth0:
[ OK ]
Waiting for network to be up
[ OK ]
Mounting other filesystems:
[ OK ]
Loading keymap: es-latin1
[ OK ]
The BackSpace key sends: ^?
[ OK ]
Starting X Font Server:
[ OK ]
Starting display manager:
[ OK ]
Starting ksysguardd:Adding loopback device to routing table ...
Starting CUPS printing system:
[ OK ]
Compiling...
Initializing...
** shorewall stuff snipped **
Starting mandi daemon: Failed to connect to the D-BUS daemon: Failed
to connect to socket /var/run/dbus/system_bus_socket: No such file or
directory
process 6091: arguments to dbus_connection_send() were incorrect,
assertion "connection != NULL" failed in file dbus-connection.c line
2971.
This is normally a bug in some application using the D-Bus library.
process 6091: arguments to dbus_connection_flush() were incorrect,
assertion "connection != NULL" failed in file dbus-connection.c line
3264.
This is normally a bug in some application using the D-Bus library.
[FAILED]
Starting HAL daemon:
[FAILED]
Starting crond:
[ OK ]
error: libhal_ctx_set_dbus_connection:
org.freedesktop.DBus.Error.FileNotFound: Failed to connect to socket /
var/run/dbus/system_bus_socket: No such file or directory
Starting anacron:
[ OK ]
Starting lisa:
[ OK ]
Starting httpd: httpd: Could not reliably determine the server's fully
qualified domain name, using 127.0.0.1 for ServerName
[ OK ]
Starting system message bus:
[ OK ]
Total time: 289 seconds
David W. Hodgins
> That parameter hangs the boot at starting HAL daemon. For what it's
> worth, here's the prcsys.log from the successive boot (without
> nopinit):
How long did you wait? Try it again, wait at least a couple of minutes,
then reboot from a cd/dvd, and copy the prcsys.log file, and post that.
>
> [root@localhost warren]# cat /var/log/prcsys.log
> Activating Compat-mode for non LSB-script nessusd
> Activating Compat-mode for non LSB-script webmin
> WARNING! Loop detected consolekit -> messagebus
> -> Ignoring requires messagebus
> WARNING! Loop detected haldaemon -> messagebus
> -> Ignoring requires messagebus
> WARNING! Loop detected mandi -> messagebus
> -> Ignoring requires messagebus
haldaemon is failing to start (without nopinit), because messagebus is
requried, but due to the above, messagebus is not being started, til far
to late in the boot process.
That is a long time to start. Mine is 28 seconds.
In my copy of 2008.0, which I haven't updated since making 2008.1, my primary
system, /etc/init.d/messagebus has
### BEGIN INIT INFO
# Provides: messagebus dbus
# Default-Start: 2 3 4 5
# Short-Description: The D-BUS systemwide message bus
# Description: This is a daemon which broadcasts notifications of system events
# and other messages. See http://www.freedesktop.org/software/dbus/
### END INIT INFO
There are no Required-Start lines, so there shouldn't be anything for pinit to
detect as a loop. What does yours have?
The 2008.1 version of the same file has
# Required-Start: $syslog $local_fs
Any chance you've installed some updates from a 2008.1 mirror?
The file is part of the dbus rpm package. The version I have on 2008.0 is
dbus-1.1.20-5mdv2008.1
Warren
wrote:
> On Sun, 03 Aug 2008 21:07:19 -0400, Warren <war...@copaninvest.com> wrote:
> > That parameter hangs the boot at starting HAL daemon.
Sorry, my bad. I tried again, this time with more patience. It booted
after 5 minutes. Oddly, prcsys.log file now looks like a shutdown log,
not a startup log:
[root@localhost warren]# cat /var/log/prcsys.log
Activating Compat-mode for non LSB-script webmin
Activating Compat-mode for non LSB-script oki4daemon
WARNING! Loop detected network-up -> oki4daemon
-> Ignoring requires oki4daemon
Stopping ConsoleKit:
[FAILED]
Stopping lisa:
[ OK ]
Disabling numlocks on ttys:
[ OK ]
Shutting down X Font Server:
[ OK ]
Stopping display manager:
[ OK ]
Shutting down httpd:
[ OK ]
Stopping resolvconf:
[ OK ]
Stopping crond:
[ OK ]
Stopping acpi daemon:
[ OK ]
Shutting down kernel logger:
[ OK ]
Shutting down system logger:
[ OK ]
Stopping system message bus:
[ OK ]
Stopping anacron:
[ OK ]
Stopping atd:
[ OK ]
Stopping CUPS printing system:
[ OK ]
Doing alsactl to store mixer settings...
[ OK ]
Shutting down ALSA sound driver (version 1.0.14):
[ OK ]
Stopping Shorewall...
WARNING: DISABLE_IPV6=Yes in shorewall.conf but this system does
not appear to have ip6tables
done.
Shutting down interface eth0: Failed to open connection to system
message bus: Failed to connect to socket /var/run/dbus/
system_bus_socket: Connection refused
[ OK ]
Shutting down loopback interface: Failed to open connection to system
message bus: Failed to connect to socket /var/run/dbus/
system_bus_socket: Connection refused
[ OK ]
Total time: 6 seconds
I'm not sure what to make of that.
> In my copy of 2008.0, which I haven't updated since making 2008.1, my primary
> system, /etc/init.d/messagebus has
> ### BEGIN INIT INFO
> # Provides: messagebus dbus
> # Default-Start: 2 3 4 5
> # Short-Description: The D-BUS systemwide message bus
> # Description: This is a daemon which broadcasts notifications of system events
> # and other messages. Seehttp://www.freedesktop.org/software/dbus/
> ### END INIT INFO
>
> There are no Required-Start lines, so there shouldn't be anything for pinit to
> detect as a loop. What does yours have?
Mine is almost word for word with yours, adding only:
# Should-Start: $network
No Required-Start lines in mine, either.
> The 2008.1 version of the same file has
> # Required-Start: $syslog $local_fs
>
> Any chance you've installed some updates from a 2008.1 mirror?
I doubt it. I update using urpmi, and I've just checked my sources:
all 2008.0.
> The file is part of the dbus rpm package. The version I have on 2008.0 is
> dbus-1.1.20-5mdv2008.1
Hm, mine is dbus-1.0.2-10.2mdv2008.0, and urpmi thinks this is up to
date:
[root@localhost warren]# urpmi dbus
Package dbus-1.0.2-10.2mdv2008.0.i586 is already installed
Packages dbus-1.0.2-10mdv2008.0.i586, dbus-1.0.2-10mdv2008.0.i586 can
not be installed
And by the way... thank you!
David W. Hodgins
> Sorry, my bad. I tried again, this time with more patience. It booted
> after 5 minutes. Oddly, prcsys.log file now looks like a shutdown log,
> not a startup log:
Hopefully, it won't take so long next time. Keep the nopinit for now.
I haven't tested, but I think prcsys.log may only get used when you don't
use nopinit. When you do use nopinit, the messages should be in
/var/log/syslog, and /var/log/messages. A lot more trimming required,
before posting those! :)
> Hm, mine is dbus-1.0.2-10.2mdv2008.0, and urpmi thinks this is up to
> date:
Sorry, my mistake. I had done a chroot to 2008.0 in one konsole, but ran
the "rpm -q -i dbus" command in the wrong window, so it showed the 2008.1
version. Checking the mirror I used for 2008.0, you have the latest version.
Does having haldaemon starting before kde fix the automount problem? Still
have to figure out what's messed up pinit.
Amrein-Marie Christophe
> On Sun, 03 Aug 2008 23:02:11 -0400, Warren <war...@copaninvest.com> wrote:
>
>> Sorry, my bad. I tried again, this time with more patience. It booted
>> after 5 minutes. Oddly, prcsys.log file now looks like a shutdown log,
>> not a startup log:
>
> Hopefully, it won't take so long next time. Keep the nopinit for now.
>
> I haven't tested, but I think prcsys.log may only get used when you don't
> use nopinit. When you do use nopinit, the messages should be in
> /var/log/syslog, and /var/log/messages. A lot more trimming required,
> before posting those! :)
>
>> Hm, mine is dbus-1.0.2-10.2mdv2008.0, and urpmi thinks this is up to
>> date:
>
> Sorry, my mistake. I had done a chroot to 2008.0 in one konsole, but ran
> the "rpm -q -i dbus" command in the wrong window, so it showed the 2008.1
> version. Checking the mirror I used for 2008.0, you have the latest
> version.
>
> Does having haldaemon starting before kde fix the automount problem?
> Still have to figure out what's messed up pinit.
>
> Regards, Dave Hodgins
>
There was update for initscript a few days (or weeks) ago. If the
initscripts have different position than the previous one, than you have a
mess.
You could try this as root in a terminal to be sure but it's dangerous (it
reset all of them):
# don't forget the "?"
rm -f $(find /etc/rc?.d/ -type l)
cd /etc/init.d/
for i in * ; do chkconfig $i reset ; done
sync
reboot
If it doesn't reboot completely, use the kernel rescue keys to reboot
properly: keep pressing the Print key and press one by one those keys:
R S E I U B
Amrein-Marie Christophe
> You could try this as root in a terminal to be sure but it's dangerous (it
> reset all of them):
>
> # don't forget the "?"
>
> rm -f $(find /etc/rc?.d/ -type l)
>
> cd /etc/init.d/
> for i in * ; do chkconfig $i reset ; done
> sync
> reboot
>
>
> If it doesn't reboot completely, use the kernel rescue keys to reboot
> properly: keep pressing the Print key and press one by one those keys:
> R S E I U B
Tested. Need also:
for i in 2 3 4 5 ; do ln -s ../rc.local /etc/rc${i}.d/S99local ; done
Amrein-Marie Christophe
> Amrein-Marie Christophe wrote:
>
>> You could try this as root in a terminal to be sure but it's dangerous
>> (it reset all of them):
>>
>> # don't forget the "?"
>>
>> rm -f $(find /etc/rc?.d/ -type l)
>>
>> cd /etc/init.d/
>> for i in * ; do chkconfig $i reset ; done
>>
>> for i in 2 3 4 5 ; do ln -s ../rc.local /etc/rc${i}.d/S99local ; done
>> reboot
>>
>>
>> If it doesn't reboot completely, use the kernel rescue keys to reboot
>> properly: keep pressing the Print key and press one by one those keys:
>> R S E I U B
>
My good. It's more complicated than this.
Mandriva initscript policy is not easy. A lot of initscripts are not
enable by default and when you remove all previous one you just get all
messy. They will be added with incorrect start prioritie. My bad.
Forget about all that. Just type as root:
# chkconfig messagebus reset
# chkconfig haldaemon reset
That's all you need to do.
Warren
> Just type as root:
>
> # chkconfig messagebus reset
> # chkconfig haldaemon reset
That plus a reboot did it; thank you! On reflection, simply restarting
messagebus and haldaemon might have saved me a reboot. I'll remember
that for next time. ;-)
And thanks to Dave as well for teaching me some useful
troubleshooting.
Amrein-Marie Christophe
init 1 than init 5 perhaps because I think haldaemon and messagebus restart
would have crashed you PC (KDE, Gnome, ...). Reboot was the faster
solution :)
Evil Bastard
I had a similar problem occur recently. My laptop got disconnected
from the power supply (No battery) and when I rebooted, the hot plug
wasn't working for my USB stick. I had to manually mount it. Well
using the above chckconfig commands fixed it. :)