Hi Aage,

thanks for your first hint.
We just did it. We changed from a SIEMENS Cluster with RedHat 6.2
(2x2cpu XEON 750, RAID 5 70GB) to a "simple" SIEMENS dual cpu
maschine with raid 5 disks and Red Hat 6.2.

The same effects occured after some days :-((

We are still thinking about replacing these production machines with
our simple development server, which never had this problems and
errors.

Unfortunately we don't have any idea how to reproduce these errors
with our development machines and tools. They only happen in the
production environment.

Do you have any more experiences or ideas??

Robbi

PS: DB size 1 GB, biggest table has about 4 mio. records