Last night I had a Kernel panic on one of my linux boxes.

Last Output from the BOINC client seems to be ok:
<pre>...
02r9_100028558 - PH 1 TS 005998 - 05/04/1811 23:00 - H:M:S=0004:32:20 AVG= 2.72 DLT= 2.00
02r9_100028558 - PH 1 TS 005999 - 05/04/1811 23:30 - H:M:S=0004:32:21 AVG= 2.72 DLT= 1.00</pre>

stderr.txt is empty.

In the panic message was mentioned that it was a problem with PREEMPT. While trying to preempt the model-process from the kernel, the kernel panic occured.

Kernel: 2.6.7-gentoo-r11
CPU: AMD Athlon(tm) XP 2600+ (no overclocking)
RAM: 1GB DDR400

Do you need some other informations?

_____________________________________________________________
Phantasie ist wichtiger als Wissen, denn Wissen ist begrenzt. (Albert Einstein)

cpdn and boinc are not at fault for your kernel panic other than just happening to be running and loading your system.

They are "user space" programs and as such they are completely at the mercy of the kernel.

Note that the preempt features in 2.6 are still being developed. There's lots about this on linux.kernel.

See if you get better stability with preempt turned off or with a lower compiler optimisation for your system.

Check your system logs for what else was happening!


Good luck,
Martin

Mandrake 10.0.1, kernel 2.6.3-7mdk

For sure, BOINC is running as a 'user space' program.
But in the end, the kernel panic occured, while some other process (I don't know which process) wanted to preempt the model.
If you look at it that way, you are right, when you say, that this isn't a problem of the climate model.

But I look at it a bit in another way: I'm running about 5 linux boxes with a 2.6 kernel since December 2003 (in the meantime, there are about 10 linux boxes...). I compiled all kernels with preempting enabled. Max uptime for the machines are around 100 days. Until now, I have NEVER had a kernel panic before!
I installed the BOINC client yesterday on one of these machines (uptime 25 days with kernel 2.6.7-gentoo-r11). The model was running for 4,5 hours when the kernel panic occured.

Nothing strange is mentioned in the system logs (when there is a panic, there will be written nothing to the logs anymore...).

After all, this is a beta-test - so I think, that all errors should be reported...

_____________________________________________________________
Phantasie ist wichtiger als Wissen, denn Wissen ist begrenzt. (Albert Einstein)

&gt; But in the end, the kernel panic occured, while some other process (I don't know which process) wanted to preempt the model.

Linux has preemtive multitasking, which means that the kernel shedules the cputime for every process. If time is over a process is preempted (if it is in user mode) and has to wait until the sheduler gives him the next time slice.
The new feature "preemptive kernel" in the kernel 2.6 is that it is "allowed to preempt a process even if it is in kernel mode executing a system call".

I would try to switch "preemptive kernel" off. If it still crashes, it must be a hardware error.