On 22 October 2012 15:31, Guillermo Marco Puche <[email protected]> wrote: > Hello, > > I'm using a program where I can specify the number of threads I want to > use. I'm able to launch multiple instances of that tool in separate nodes. > For example: job_process_00 in compute-0-0, job_process_01 in compute-1 > etc.. each job is calling that program which splits up in 8 threads > (each of my nodes has 8 CPUs). > > When i setup 16 threads i can't split 8 threads per node. So I would > like to split them between 2 compute nodes. > > Currently I've 4 compute nodes and i would like to speed up the process > setting 16 threads of my program splitting between more than one compute > node. At this moment I'm stuck using only 1 compute node per process > with 8 threads.
Assuming you are using SGE the normal way to do this is to submit the job to a suitable parallel environment. Grid Engine will point the environment variable PE_HOSTFILE at a list of nodes to use. Either the parallel environment,your job script or a combination of the two need to convert this to a format that your code will understand. Ideally you also want the processes of the job to be launched via grid engine so it can monitor and control them properly. This generally involves making your code use qrsh -inherit to launch processes on other nodes. William > > Thank you ! > > Best regards, > Guillermo. > _______________________________________________ > users mailing list > [email protected] > https://gridengine.org/mailman/listinfo/users > > _______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
