From: Herbert Poetzl (herbert_at_13thfloor.at)
Date: Thu 15 May 2003 - 06:47:58 BST
On Wed, May 14, 2003 at 06:29:15PM +0100, David M. Wilson wrote:
> On Wed, May 14, 2003 at 06:31:23PM +0200, Thomas Hug wrote:
> > hi john
> > On Wed, 14 May 2003 09:24, John Goerzen wrote:
> > > I have upgraded to 2.4.20ctx-17, hoping that this would solve our
> > > problems. Alas, it has not. Our server went down again today, this
> > > time with absolutely no console messages whatsoever (our console is
> > > now on the serial port and logged -- and nothing there.) Magic SysRq
> > > was useless.
> > >
> > > Does anyone have any ideas? Is there any more recent patch that we
> > > should be using?
> > are you sure that your hardware is ok? i'm just coming from the server
> > room where i bootet a machine with some virtual memory blabal failure.
> > the same machine as i booted last friday with the same screen output.
> > another server had similar problems. as i changed cpu/ram/mainboard, the
> > problems were gone...
> RAM/CPU is my first choice when random crashing occurs, followed by
> motherboard then PSU. Hope that helps.
I agree, but recently I also expierienced very strange
stability issues with inadequate chipset cooling on AMD
systems, and missing PS/2 mices *smile* so it could be
how to test?:
- memory: usually memtest86
- cpu: 1day kernel kompiling (on SMP with -jN)
- chipset/etc cooling (get a hairdryer ;)
if it is one system, simply replace motherboard
(with different type/cpu) and memory and see
if it is stable (usually the simples and fastest
way to stabilize) then, if you want to, hut the
issue down ...