Elsewhere in this thread Brian McNally points out that the problem doesn't occur when I use an RHEL6 exec host. Rather than belabor this further I will just try to switch my application to run on RHEL6 instead.
Michael On Tue, Feb 28, 2012 at 11:29 AM, Michael Hoffman <[email protected]> wrote: >> So I think may be you can wrap around the command with a shell script, >> and see if you can reproduce it with something like: >> >> $ qsub -b y -sync y wrapper >> >> And wrapper is a simple shell script: >> >> #!/bin/sh >> >> echo %s >> exit $? > > $ cat > wrapper > #!/bin/sh > > echo %s > exit $? > > $ chmod a+x wrapper > > $ qsub -b y -sync y ./wrapper > Your job 1629725 ("wrapper") has been submitted > Job 1629725 exited with exit code 0. > > $ cat ~/output/*/wrapper.1629725* > %s > > That works fine. It really appears to be the occurrence of "%s" in the > submission command line that breaks things. A non-sensical example: > > $ qsub -b y -sync y -l testing=TRUE ./wrapper %s > Your job 1629726 ("wrapper") has been submitted > Job 1629726 exited because of signal SIGSEGV > > Do you have any suggestions on how to diagnose what has gone wrong > with our environment? > > Many thanks, > Michael Hoffman -- Michael Hoffman, PhD Senior Fellow Department of Genome Sciences University of Washington 3720 15th Ave NE Seattle, WA 98195-5065 [email protected] _______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
