[RAS] nebka

[Christian Zimmermann]

On Tue, 12 Feb 2008, Thomas Krichel wrote:

>
>  Argh!!!
>
> nebka:~#
> [1]+  Done                    nohup rsync -qa /home/adrepec/ adrepec  (wd: /vol/opt)
> (wd now: ~)
> nebka:~# cd /home/adrepec/
> nebka:/home/adrepec# ls -l
> Bus error
>
>  The disk has gone bad again.
>
>  We should not be working with it. We need to transfer everything
>  to the /vol disk.
>
> nebka:~# shutdown -h now
> -bash: /sbin/shutdown: Input/output error
>

This just confirms that this machine is not capable of handling any 
instruction involving large directory trees. If you have to use rsync, 
split it in pieces...

I went to reboot the machine. The screen was showing the same ext3 errors 
we had with earlier crashes.

HOWEVER, the machine is not booting now. It hangs with:

aacraid: Host adapter reset request. SCSI hang?

A rapid googling of this reveals that we have either to modify BIOS specs, 
recompile the raid driver, make sure we have kernel version 2.4 (I think 
we have), or other steps that are beyond my competence. I warrant Tim can 
do this, but he is not in today. I hope he can do it tomorrow. I will be 
teaching 8:00-1:45 tomorrow.

Thomas, please note that Tim is a very busy person, and he has helped us a 
lot. But we should not abuse his goodwill. So think hard before launching 
big operations on this machine.

>
>
>  Cheers,
>
>  Thomas Krichel                    http://openlib.org/home/krichel
>                                RePEc:per:1965-06-05:thomas_krichel
>                                               skype: thomaskrichel
>