Linux reboots itself during operations that require large volumes of
database writes to the disks, such as ORACLE imports and data loading.
We have a adaptec 2940 ultrawide scsi host adaptor with 5 * 4GB disks.
Therefore, this may be related to bug number 2938; but we have a slightly
different hardware configuration (including a pentium processor).
We have checked for bad memory using memtest86 and there's no problem
Before we recompile the kernel, could you give us any further information
or a suitable resolution for this problem. The reporter of bug 2938 found
his own resolution to the problem of Linux 'spontaneously' rebooting, but
he says he's not 'absolutley certain' of the cause of the problem and
seems to have changed several settings in the kernel.
Julia Wixler & Norman Stevens, PMPL
Re-compiled the kernel with APM and watchdog timer removed, and using 386
instead of 586 optimisations. This has not resolved the problem!!! The machine
is still consistently crashing during heavy disk usage.
This might be totally left field... but I've had similar problems (with M$
systems) that turned out to be Power Supply Problems...
Assigned to dledford
has this been resolved with newer kernels/releases