Hi,
we are using pbspro V14.0.2 on linux systems.
I would believe that if I restart pbs_server and pbs_sched by running this command,
/etc/init.d/pbs restart
that would not affect running jobs on compute nodes. however, the fact is that after I restarted pbs server and scheduler, all running subjobs of a job array were requeued and re-ran. non array jobs kept running without interruption.
is this a bug with v14.0.1? has it been identified and eliminated in later versions?
Thanks,
Sue