[GE users] [sge6] Jobs submitted succesfully won't run.

Sean Dilda agrajag at dragaera.net
Mon Jul 26 14:38:20 BST 2004

On Sat, 2004-07-24 at 10:21, Wee Yeh Tan wrote:
> On Sat, 24 Jul 2004 21:59:53 +0800, Wee Yeh Tan <weeyeh at gmail.com> wrote:
> > I'm gonna give this a shot.
> *Sic* The issue was resolved by rebooting the system (most likely
> would have worked if I restarted qmaster & schedd).  Somehow it didn't
> cross me to try to restart the services even after 4 days of trying.

There is a bug in SGE6 (supposedly fixed in CVS) where sge_schedd will
stop talking to sge_qmaster, and thus cause no jobs to get scheduled,
for any reason.  If this problem shows up again, you can run 'qconf
-sss'.  If it gives the hostname for your master node, then its not that
problem.  But if it says something about having a scheduler host not
being defined, you can then kill the running sge_schedd process and just
start a new one and everything should work.  Or you can use the
sgemaster script to stop all the master processes, then restart them.


To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net

More information about the gridengine-users mailing list