Re: amber-developers: latest tarballs for AmberTools and Amber10

From: Lachele Foley <lfoley.ccrc.uga.edu>
Date: Mon, 31 Mar 2008 13:10:12 -0400

I figured out the main problem with the parallel run tests. Again, I should have thought of this already. This is something I know: I set it up. It isn't your fault, but you might be able to help.

We put our common-area executables in a directory that is not writable by the compute nodes. This is important to us since we're likely to have homemade programs running about, and wish to minimize unintended overwrites. So... while $AMBERHOME is visible to all the nodes, any test that requires a node writing to $AMBERHOME/anything will fail for us, and the diff will be against an empty file. Programs that merely send output to the scheduler will work.

I just tried copying the test and src directories elsewhere, but that didn't work. I'm not sure why yet. There is certainly some variable(s) somewhere that I can tweak. If not, and even if so, then put it on my list of requests to have an $OUTPUT_DIRECTORY defined so that test output can be written to an alternate directory.


:-) Lachele
--
B. Lachele Foley, PhD '92,'02
Assistant Research Scientist
Complex Carbohydrate Research Center, UGA
706-542-0263
lfoley.ccrc.uga.edu
Received on Fri Apr 18 2008 - 21:15:36 PDT
Custom Search