On Fri, 5 Oct 2012, ron at ron-l-j.com wrote:

> The first thing I would try is booting into a shell and running fsck on
> your boot disk.
> Reboot the machine and press e in the grub screen for edit.
> Go to the kernel line and press e to edit that line.
> Then add to the kernel line
>
> init=/bin/bash
>
> when you boot you go straight to the shell.
> then do a file system check on your boot drive
> fsck -t ext3 /dev/sdax
> t is for file system type in this case ext3
> and the x in sdax is your partition number.
> sda is for sata drives, if you have an ide drive use hdax
>
> Most of the time its a file system error and was very common when I was
> doing raid arrays in my advanced linux classes. Bad super block is also
> common.
> If you are still having troubles run fsck from a live cd/usb. But you will
> have to run the cd in live mode, mount the offending drive, and chroot
> into the drive.
> As a last resort you can update-grub from the live cd after you have
> chroot'ed into you installed environment.
> Let me know what you encounter.


Thanks, Ron.  (I'm cc'ing you in case you are doing digests only and want 
to see this before the digest comes in.)

Did you see that I was able to boot up to the window manager login prompt 
using an earlier version of the kernel?  Can I just take it from there?

My impression right now is that the problem was caused by my failure to 
reboot for weeks after installing some packages that required rebooting. 
Then I installed 400 more packages before rebooting.  Is that possibly the 
cause of my troubles?  Could it be that I just need to fix the packages 
and reboot?

The other issue is that I have a RAID1, so mirrored drives, and I think 
that means I don't want to fsck them one at a time.  df shows this:

$ sudo df -HT
Filesystem     Type      Size  Used Avail Use% Mounted on
/dev/md1       ext4      3.0T  220G  2.6T   8% /
udev           devtmpfs  8.4G   13k  8.4G   1% /dev
tmpfs          tmpfs     3.4G  832k  3.4G   1% /run
none           tmpfs     5.3M     0  5.3M   0% /run/lock
none           tmpfs     8.4G  148k  8.4G   1% /run/shm

Does that mean I would fsck /dev/md1?

Thanks again.  I really appreciate your taking the time to reply.

(One thing I've learned -- I should run package-manager daily and try to 
reboot soon when it is required.  I often have a lot of stuff running that 
I don't want to kill, so reboots are a hassle.)

Mike