[AMBER-Developers] incompatible processor requirements in parallel tests

From: Lachele Foley <lfoley.ccrc.uga.edu>
Date: Thu, 18 Feb 2010 11:55:51 -0500

Even though I now have CVS access, I don't feel qualified to fix this myself (except maybe to add some documentation).

Some parallel tests require np=2, some require np>2, at least one requires np=1. But, these skipped tests don't show up in "TEST_FAILURES" and aren't documented. So, unless the user runs the stdout from the tests into a file and checks it (like I do), they never know. I think this could be fixed with some documentation and perhaps a segregation of parallel tests (make test.parallel.np2, etc.). If there is documentation, I didn't find it in the places I expected it.

Examples follow:

When I run the parallel tests with np=4, I get complaints on some tests like this:
cd tgtmd/change_target.ntr && ./Run.tgtmd
SANDER: Targeted MD with changing target and restraints
 DO_PARALLEL set to mpirun -np 4
 too many processors for this test, exiting (Max = 2)

...but if I set np=2, I get (again, for example):
make[1]: Leaving directory `/programs/amber11_20100216/test'
cd neb/neb_gb && ./Run.neb_classical
 This test case requires a least 8 mpi threads.
 The number of mpi threads must also be a multiple of 8 and not more than 24.
 Not running test, exiting.....
cd neb/neb_gb_large_system && ./Run.neb_ls_classical
 This test case requires a least 32 mpi threads.
 The number of mpi threads must also be a multiple of 32 and not more than 128.
 Not running test, exiting.....
cd ncsu && ./run-parallel.sh
>>>>>>> doing 'abmd_ANALYSIS'
diffing save/mdout with mdout

And there's this one:
  Perturb to nothing test
  This test needs to be run by hand, using sander.MPI on 1 CPU

:-) Lachele
B. Lachele Foley, PhD '92,'02
Assistant Research Scientist
Complex Carbohydrate Research Center, UGA
AMBER-Developers mailing list
Received on Thu Feb 18 2010 - 09:00:03 PST
Custom Search