Hi All,
I am using torque and maui set up on a custom cluster.
But when i run a qsub job, it exits and completes the job with no log files.
qstat -f spews this error:
Unable to copy file:
/var/spool/torque/spool/132.myhost.OU to /clstr/work/myjob.pbs.132
I have checked my torque cluster setup and I have bidirectional access -
to and fro between masternode and cluster nodes without a password.
The same user account with identical uid/gid is also on all the compute
nodes.
I am running jobs on a a nfs mounted partition (/clstr) visible across all
compute nodes. pbsnodes -l shows all the compute nodes in free state.
looks like this is an issue with torque unable to write to
/var/spool/torque/spool. This dir has 2777 permission and I could manually
copy to this dir.
Am I missing something in the configuration ? please let me know.
Regards,
KM