[GE users] qdel does not delete the job

Srikanth srikanth at locuz.com
Fri Mar 24 15:49:58 GMT 2006


I am using MPICH 1.2.7-1 which comes default along with ROCKS cluster 4.1.

Even I tried submitting the job with environment variable
"MPICH_PROCESS_GROUP=no", but it could not kill slave processes on the
compute nodes.

Please suggest how to come out of this issue.


Sri

-----Original Message-----
From: Rayson Ho [mailto:rayrayson at gmail.com] 
Sent: Tuesday, March 21, 2006 6:26 PM
To: users at gridengine.sunsource.net
Subject: Re: [GE users] qdel does not delete the job

Which MPI implementation are you using??

You need to use the "tight integration" for the PE:
http://gridengine.sunsource.net/howto/howto.html#Tight%20Integration%20of%20
Parallel%20Libraries

Rayson


On 3/21/06, Srikanth <srikanth at locuz.com> wrote:
> I am facing typical problem in my Rocks Cluster.
>
> I have installed and using the SUN grid roll on the ROCKS Cluster,
>
> When I use 'qdel' to delete the running job on the cluster it displays
that
> the job is deleted and even in 'qstat -f' output the job is cleared from
the
> queue but the instances of the jobs are running on all the nodes of the
> cluster.
>
> It was unable to kill the all job processes and unable to clear 'ipcs".
>
> Kindly help me resolving the issue.
>
>
> Regards,
>
> Sri

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





---------------------------------------------------------------------
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