Hi,
After running a certain application on my data for 2 hours, my
machine suddenly goes from feeling normal to not reacting to
anything: ctrl-alt-del doesnt work, mouse pointer does not
move, ping returns nothing. I then reran in console mode as
root, hoping there would be something to look at at on the
screen, but I missed it, the monitor went to sleep (and then
gets no signal, cant wake it up). Then press reset button,
and look in /var/log/messages - nothing unusual. Then I
did kern* to /var/log/kernel in /etc/syslog.conf, but again
no errors. It looks like the program I run stops when it
crawls over 512 mb of ram (there is 768 Mb installed), and I
got the source code of it (but no program should be able to
die a system, right?) What could it be. When I get sleep, we
will try to swap the motherboard. Hopefully its hardware, but
then why does the program stop at the same point, shouldnt it
then be more random. If a new motherboard makes no difference,
then I dont know how to find the trouble, and will feel a bit
desparate then, because of deadlines. The trouble machine is
part of a "production environment", and I just installed it
from scratch (formatting all disks) using mandrake60-2.iso.
I guess my question is, how does one get the maximum log
information out of the system, in order to get clues?
Second, now that I have re-installed 4 more machines due to
that file-corruption deal, have there been any serious
trouble reports for the current Mandrake? or could I go
ahead and propagate it to the next 10-20 machines?
Niels L, [EMAIL PROTECTED]
