Unmounting File Systems (hang)

[booman]

good question. I went down the list and "umount" all of them. Some were obviousl "in use" like any thing on sda1 cause that is the OS.
I would have to go through the process and do one at a time to narrow it down to a specific mount.

I tried the live CD but wasn't totally sure what to do. I tried fsck but got the same error for any drive. I tried it on /dev/sdb0, sdb0, sda1, /dev/sda1
and it didn't seem to work.

I also applied the rest of the updates to see if it fixed something. nope

[booman]

I did a little searching about the white screen after login (with a black mouse pointer) and found a few posts.
One was reccomended to look in the Xorg.0.log and look for errors.
I didn't find any EE errors but there were a few WW warnings about the video driver (I think).
Its weird because the white screen acts like a window in full screen but nothing is there, just a pointer. I can't minimize, resize or even switch to a different desktop.

[booman]

I tried glxinfo:
direct rendering: No

Maybe the update screwed with my ATI drivers.

I tried to run Enemy Territory and it tried to start, but I get video errors...

You are using software Mesa (no hardware acceleration)!
Driver DLL used: libGL.so.1
If this is intentional, add
"+set r_allowSoftwareGL 1"
to the command line when starting the game.
************************************************** *********
...WARNING: could not set the given mode (4)
Initializing OpenGL display
...setting mode 3: 640 480
Using XFree86-VidModeExtension Version 2.2
XF86DGA Mouse (Version 2.0) initialized
XFree86-VidModeExtension Activated at 640x480
Received signal 11, exiting...

[Kaitain]

OK, the ATi drivers installation is botched, but is unlikely to be related to devices not unmounting successfully.

You would need to go through each device individually to record what mounts or unmounts. You may find something gets recorded in /var/log/messages or another logfile before / gets unmounted, which may lead you to identify which device isn't successfully unmounting.

When you boot up a livecd, they can be a little bit random with their device naming. If you are booting using something like the Ubuntu livecd, the easiest way to check for errors is using gparted.

Otherwise click one of the partition's icons and examine the output of the mount command.

If not, then trawl either the output of dmesg or the contents of /var/log/messages for something like:

May 27 23:24:22 localhost kernel: scsi 0:0:0:0: Direct-Access     ATA      TOSHIBA MK3256GS LH01 PQ: 0 ANSI: 5
May 27 23:24:22 localhost kernel: sd 0:0:0:0: [sda] 625142448 512-byte logical blocks: (320 GB/298 GiB)
May 27 23:24:22 localhost kernel: sd 0:0:0:0: Attached scsi generic sg0 type 0
May 27 23:24:22 localhost kernel: sd 0:0:0:0: [sda] Write Protect is off
May 27 23:24:22 localhost kernel: sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
May 27 23:24:22 localhost kernel: sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
May 27 23:24:22 localhost kernel: sda:

However I suspect that your problems will go away naturally after upgrading to a more modern distribution, though that does involve a bit of reconfiguration.

Did you install the ATi drivers from your distribution's repository or by using ATi's installer? Did you build a custom kernel?

[booman]

After all my messing around last night, it got worse.
I tried a repair with the Fedora 8 disc but after that it would automatically do a fsck and "FAILED" each time.

So I broke down and reinstalled Fedora 8. I know, old distro but I know it pretty well and works good.

I got Fedora 8 reinstalled just fine.
The cool thing is my /home and /backup are on separate partitions, so all of my data is secure.

Now I have to install updates, create my shares & SMB users, disable SELinux and configure my backup programs.

Thanks for your help guys....

[booman]

I bet you thought everything was working fine at this point.

so did I....

Actually I only have 2 problems and are that same as before the reinstall.
I found a post where someone has the same hardware and same symptoms.
Shutdown Failed
Its interesting because he has an Intel 1000 ethernet adapter and an Asus Deluxe board too.
He stated that the onboard SATA controller and Ethernet Adapter share the same IRQ.
I think its IRQ 23 on my machine because it shows on the shutdown process, even though it still hangs on filesystem.
Let me know what you think.
I already tried adding "noirqdebug" at the end of grum.conf but it still happens.

[booman]

Here is another post I found.
He had messed up the umountnfs.sh, umountfs, umountroot, halt and reboot in an /etc/rc0.d and /etc/rc6.d file.
I'm afraid to mess with these because I don't want to make symptoms worse.
Let me know if you think this may fix my problem
/etc/rc0.d

[booman]

oh the other problem
Grub keeps wanting to boot to sda(1,0) instead of sda(0,0)
So I have to keep editing the grub during startup to make it boot to the correct hard drive.
I even manually edited the grub.conf but it still happens

[Gizmo]

Originally Posted by booman

oh the other problem
Grub keeps wanting to boot to sda(1,0) instead of sda(0,0)

Sounds to me like you have two drives, and you have grub installed on both. However, on your boot drive, you ran something like setup(hd1,0), so your boot drive thinks it is supposed to load the stage 2 grub from the other drive.

[booman]

I think you are right too.
The /backup never had a grub installed on it, but /home may.
I'm not sure. Is there a way find all the grub files on each partition?


Also I found the "hang" is on a mount called sysfs
if I unmount sysfs then it shuts down just fine

[Gizmo]

Can you give me the output of

ls /etc/rc0.d

[booman]

says command not found

I typed it exactly. the 0 is a zero right?
I even went to that directory and there is a folder called rc0.d

[Gizmo]

Originally Posted by booman

I think you are right too.
The /backup never had a grub installed on it, but /home may.
I'm not sure. Is there a way find all the grub files on each partition?

The grub boot sector is just that, it's the boot sector of the drive. There is no file to look at.

You can look at the output of the dd command and see if there is something that looks like a boot block there:

dd bs=512 count=1 if=/dev/sda of=boot.sect
hexdump boot.sect

Use /dev/sdb for the second drive....

[Gizmo]

Originally Posted by booman

says command not found

?!?!

That means it can't find ls ('el es'). That's part of the core system for most versions of Linux/Unix. Either you typed something wrong, or you've got something very messed up.

[booman]

Originally Posted by Gizmo

The grub boot sector is just that, it's the boot sector of the drive. There is no file to look at.

You can look at the output of the dd command and see if there is something that looks like a boot block there:

dd bs=512 count=1 if=/dev/sda of=boot.sect
hexdump boot.sect

Use /dev/sdb for the second drive....

That didn't work either.
it didn't like "hexdump"

I tried it without "hexdump" and the last "boot.sect"
output:
1+0 records in
1+0 records out
512 bytes (512 B) copied, 0.0165665 s, 30.9 kB/s

[booman]

Originally Posted by Gizmo

?!?!

That means it can't find ls ('el es'). That's part of the core system for most versions of Linux/Unix. Either you typed something wrong, or you've got something very messed up.

I don't think Fedora is totally messed up because I will still boot and run fine. I just have to manually set the boot drive for grub and then manually umount sysfs

If reinstalling is a failsafe way to fix it. I can do that.
All of my files are backed up

[booman]

ok, I'm stupid (again)
I was typing "is" instead of "ls", which I use all the time.
I even got new glasses last night.

Ok it worked but there is a bunch of files.... which one do you want to look at?
K74rsyslog
K95firstboot

[Gizmo]

Originally Posted by booman

Ok it worked but there is a bunch of files.... which one do you want to look at?

I need to see the whole list. rc0.d is the list of services that get shut down to transition to run-level 0, which reminds me; I actually need to see the listing for rc6.d, which is the run-level that you transition to in order to shut down.

Originally Posted by booman

That didn't work either.
it didn't like "hexdump"

try using 'xxd' instead of 'hexdump' then.

[booman]

K01dnsmasq K25sshd K74lm_sensors K88auditd
K01smartd K30sendmail K74nscd K88pcscd
K01smolt
K35nmb
K74ntpd
K89dund
K02avahi-daemon
K35smb
K74rsyslog
K89netplugd
K02haldaemon
K50bluetooth
K75fuse
K89pand
K02NetworkManager
K50netconsole
K75netfs
K89rdisc
K03yum-updatesd
K50snmpd
K75udev-post
K90network
K05anacron
K50snmptrapd
K84btseed
K91capi
K05atd
K56acpid
K84bttrack
K91isdn
K05saslauthd
K60crond
K85mdmonitor
K92ip6tables
K10ConsoleKit
K69rpcsvcgssd
K85rpcgssd
K92iptables
K10cups
K72autofs
K85rpcidmapd
K95firstboot
K10psacct
K72wpa_supplicant
K86nfslock
K95kudzu
K15gpm
K73messagebus
K87irqbalance
K99cpuspeed
K15httpd
K73setroubleshoot
K87multipathd
K99microcode_ctl
K20nfs
K73winbind
K87restorecond
S00killall
K24irda
K73ypbind
K87rpcbind
S01halt

[booman]

by the way, I fixed the booting problem.
I changed the sda(1,0) to sda(0,0) in the grub.conf
and it booted fine the last time.
We'll see if it sticks