Hello,
I am currently setting up a Warewulf cluster to use LAM and SGE. I
have confirmed that LAM works fine on its own. I run lamboot, see that
the nodes are part of the universe, and when I run my job, it gets
distributed properly.
However, when I try to run the job through SGE, the job gets qued and
run, but the output file consists of the following:
Now, as far as I understand, the lamstart.sh script, which SGE should be
calling before processing the job, should take care of this, but
apparently it doesn;t.
I have also tried (in vain) to load lamboot manually before queuing the
job, and that didn't work either.
Obviously, I'm very new to the LAM/SGE environment. Any ideas where I
went wrong?
Thanks,
Jason
--
======================
Jason Somers
Network Administrator
Red Barn Technology Group. Inc.
1235 Front Street - Suite 3
Binghamton, NY 13905
(607) 772-1888 x222
|
