I just wonder if there is a way to make a W job back to Q and reschedule it，I know that we can use qrun to force the W job immediately rerun if resources is satisfied or preempt if it is enabled，but has no effect if it cant run
restart server would do the trick，but dont make sense if only for this
Any suggestion？
Could you please share the reason for the “W” state of the job ?
Please try qrerun command, otherwise, the job has to be deleted (forcefully if required) and resubmitted again.
e.g. if you add a new user and forget to configure passwdless ssh, it’s simple to fix. 
but rescue ‘W’ jobs is annoyed
Thank you , after correcting the user credentials , try to qrerun the job ( qrerun ) . Otherwise, delete and resubmit .