Just installed queue-1.4b on three machines in a cluster. The spool dir is shared via no_squash_root NFS, and the configure command is
./configure --enable-root --sharedstatedir=/home/scratch/nonlocal/com
Whenever queue decides to run commands on a host different than the local machine, I receive emails from the daemons on the local and non-local machine stating things like:
Can't creat(/home/scratch/nonlocal/com/spool/queue/wait/ofm961141216) output fi\le: File exists
and
Can't creat(/home/scratch/nonlocal/com/spool/queue/wait/ofm961141219) output fi\le: File exists
and more messages like
Job 'cfm961141216' in 'wait' queue_b on bedevere.bioc.cam.ac.uk has completed s\uccessfully.
Total CPU used: 25.3 sec.
Can't stat output file wait/ofm961141216
Can't unlink(wait/ofm961141216): No such file or directory
Of course, I never receieve my command outputs. The commands are run using the test.sh script included below