I got in touch with Thomas Kuhr, and he added libtcl8.3.so to the cdfsoft2 release directory on GridKA, and ran MCProd_v6_1_4mc_q_strip.tar.gz on one of their SL5 machines.
./mcProduction/scripts/run1segment -v 1 -J 1 -b cdfpewk -d tewk9t -f local -x USE_TIKI=1 2>&1 | tee job.log
It ran successfully, and I am currently trying to see if the results look the same as from a run at Fermilab. I suspect the random number generator will make them different, though.
Update: The results are different, and it looks very much like a random number generator was set differently between the two runs. They even create different numbers of files: one creates 1 and the other 2, though the total size is similar. The internals have the same shape, in everything I have checked so far (about 40 histograms), but differ on the order of a few events here and there.
Modified 06-August-2009 at 14:41
|Previous notes||Next notes||Main slide directory|