[GE users] controlling jobs on failed nodes

templedf dan.templeton at sun.com
Fri Aug 14 16:36:41 BST 2009


For reschedule_unknown to work, your jobs need to be rerunnable.  To be 
rerunnable, they either need to be submitted with "-r y" or they need to 
be running in a queue with "rerun TRUE" (and they need to not have been 
submitted with "-r n").

Daniel

snosov wrote:
> Thank you for the information. I set that up. The only confusion that 
> I still have is that the man page says that for reschedule_unknown to 
> work, the queue need to be 'rerunnable' and the users need to submit 
> jobs with "-r y". Is it necessary to have both: the "rerun" for a 
> queue and "-r y" for a job? Aren't these the same, only the former 
> being for the whole queue and the latter for just a particular job?
>
> Thank you,
> Serge.
>

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=212267

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].



More information about the gridengine-users mailing list