[Wien] cluster problem

[Michael Gurnett]

The search continues. I checked both the speed of the hard drives and for
the nfs and both seem ok (the extreme test of writing 4 Gb took 1m43 to hd
and 2m18 to the nfs). Smaller files less than 50mb take less than a second.
I tracked down an error that wasn't being written in the error files but was
found in the log. It appears that case.energy_1 was not being written to
completion and I am therefore getting an error in fermi_tetra due to the
file ending prematurely. If someone is running a pc cluster I would love
some info or recomendations ie optimum way to set up nfs any tweaks that may
be required and so on

Home someone can help

best wishes


Michael