Hi all,

I could use some help getting DMRAA to run.  We already have a galaxy instance with pbs running, but I'd like to connect to Torque using dmraa because we eventually intend to port over to a different resource manager.

I've followed these instructions:
http://apps.man.poznan.pl/trac/pbs-drmaa/wiki#Introduction
http://wiki.galaxyproject.org/Admin/Config/Performance/Cluster

Got dmraa installed and exported a path in my shell config: (PREFIX is long so I omitted it)
DRMAA_LIBRARY_PATH=PREFIX/drmaa/lib/libdrmaa.so.1.0.12

in my universe_wsgi.ini:
start_job_runners = drmaa
default_cluster_job_runner = drmaa://-q tstgalxy -V/
addValue = drmaa:///

Here's my pbs_drmaa.conf:
wait_thread: 0,
pool_delay: 5,
keep_completed: 60,
user_state_dir: PREFIX/%s/jobs,

I tried something simple like the add value tool.  When I run it, I get the following error state:

perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
	LANGUAGE = (unset),
	LC_ALL = (unset),
	LC_CTYPE = "en_US",
	LANG = "C"
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").

The output file is actually correct, Galaxy just thinks there's an error. I get 2 files in user_state_dir with a job number and .started (empty) and .exitcode (0). Sounds like an env issue?  If I run the perl script generated in the .sh file, it works fine. Output as expected. If I use local:/// it works fine with no error.

I couldn't find any documentation on how to actually link dmraa to torque.  Am I missing something major?
Thanks in advance,

Carrie Ganote