Re: [tlug] freeze on AMD64 Dual Opteron server

[Evan Monroig <evan.monroig@example.com>]

Tapio Peltonen <tappel@example.com> writes:

> On Thu, Oct 26, 2006 at 03:11:21PM +0900, Evan Monroig wrote:
>
>> But now I would like to send the data to another computer using rsync,
>> and after about 20 seconds it invariably hangs.
>> 
>> There is no way to do anything.  The ssh connection is cut, the computer
>> is doesn't appear on the network (I configured a static IP address).
>> I tried connecting a screen and keyboard to it while doing the rsync,
>> but after failure the computer is completely hung.  Frozen screen,
>> keyboard not responding (numlock, capslock..).
>
> This sounds like a hardware failure (or interoperability problem). I
> would try the following:
>
> 1. Check if it responds to the Magic SysRq (I would bet it won't, but
> you never know). If it does, you at least get a bit more information for
> troubleshooting.

One more new thing to me.  Thanks !  According to this page [1], I tried
the following commands one after another, without effect (I didn't
connect the screen, because that's too bothersome, and I wanted to try
other things as well).

ALT-SysRq-s (sync filesystems)
ALT-SysRq-u (remount filesystems readonly)
ALT-SysRq-b (reboot)

> 2. Try disabling hd dma, or at least try different dma settings. This is
> the most probable cause for this kind of behaviour.

I tried the following settings, in vain:

1) disable dma
sudo hdparm -d 0 /dev/hda

> 3. Run a full memory check with memtest86.

Before doing this, I did other tests, and found the error cause (see
next messages)

Thanks for your help !

Evan


[1] http://snafu.freedom.org/linux2.2/docs/sysrq.txt