I’m running PBS pro community 20.0.1 on Rocky Linux 8.6 (RHEL-equivalent).
The system have a login node with PBS server on it, but does not run job itself.
A computing node receives job from the login node.
The problem is, the software works by “resubmit” itself in iterations - the subsequent jobs are submitted from the computing node, not the login node.
But I have such error:
[mike@forest ~]$ qsub test.sh 2993.forest [mike@forest ~]$ ssh node01 Last login: Fri Aug 26 14:52:24 2022 from 192.168.1.1 [mike@node01 ~]$ qsub test.sh qsub: Bad UID for job execution
I’ve setup the password-less SSH and home folder sharing between the 2 machines.
The qmgr output follows:
create queue workq set queue workq queue_type = Execution set queue workq enabled = True set queue workq started = True # # Set server attributes. # set server scheduling = True set server acl_host_enable = False set server default_queue = workq set server log_events = 511 set server mail_from = adm set server query_other_jobs = True set server resources_default.ncpus = 1 set server default_chunk.ncpus = 1 set server scheduler_iteration = 600 set server resv_enable = True set server node_fail_requeue = 310 set server max_array_size = 10000 set server pbs_license_min = 0 set server pbs_license_max = 2147483647 set server pbs_license_linger_time = 31536000 set server eligible_time_enable = False set server max_concurrent_provision = 5 set server max_job_sequence_id = 9999999
There’s a allow_node_submit server property in Torque to do this, but I don’t think the PBS pro has this. Any ideas?