Sent from my iPad
On Apr 3, 2012, at 15:38, juliosergio <julio...@gmail.com> wrote:
> I'm trying to install rocks in a cluster composed of one frontend and 10
> nodes.
>
> After I, supposedly, installed the system in the Frontend, I noticed that
> when booting, it couldn't start something called "sgemaster". The message
> displayed was:
>
> Starting sgemaster.rossby starting sge_qmaster
> sge_qmaster starting problem
> sge_qmaster didn't start
>
> [FAILED]
>
>
>
>
> Could I proceed with the installation of the system in the other nodes or
> the mentioned problem is
> something crtical?
>
> Do you have any comments on this?
>
> Thanks,
>
> --Sergio.
>
>
> --
> View this message in context: http://discussion-on-rocks-clusters.16692.n6.nabble.com/sgemaster-sge-qmaster-didn-t-start-tp4683086p4683086.html
> Sent from the Discussion on Rocks Clusters mailing list archive at Nabble.com.
On Tue, Apr 3, 2012 at 12:38 PM, juliosergio <julio...@gmail.com> wrote:
> I'm trying to install rocks in a cluster composed of one frontend and 10
> nodes.
>
> After I, supposedly, installed the system in the Frontend, I noticed that
> when booting, it couldn't start something called "sgemaster". The message
> displayed was:
>
> Starting sgemaster.rossby starting sge_qmaster
> sge_qmaster starting problem
> sge_qmaster didn't start
>
> [FAILED]
>
>
>
>
> Could I proceed with the installation of the system in the other nodes or
> the mentioned problem is
> something crtical?
>
> Do you have any comments on this?
you should look into:
/opt/gridengine/default/spool/qmaster/messages
for some hints of what might be the problem.
And yes it should work out of the box, and if you want to use the
scheduler to submit jobs yes it is very critical.
Sincerely,
Luca