[RegCNET] MPI run

[Maurice.McHugh at noaa.gov]

Hi Saeed,

This is the default openMPI error message.  Go into the error file that openMPI created for that run.  In your run directory there should be two files generated with names like:

RegCM_2029010100_2029073118_1.o1839  and RegCM_2029010100_2029073118_1.e1839

The first string before the dot ('.') is associated with a specific run name within MPI, and the job number is the number to the right of the dot after the e (error) or o (output).

So go and look in the output directory, and examine the error file associated with that run (plus probably also the output file, to get more info on where and maybe why the run failed).

Keep in mind that depending on some systems (but probably not yours) you may need to submit your job to a scheduling system like 'qsub'.  In that case the mpirun -np 7 ./Bin/regcmMPI regcm.in >regcm.out command will be contained within a qsub script.

Hope this helps,

Maurice



----- Original Message -----
From: Saeed Bayat <saeedbayat7276 at yahoo.com>
Date: Tuesday, June 21, 2011 1:55 pm
Subject: [RegCNET] MPI run
To: "RegCNET at lists.ictp.it" <RegCNET at lists.ictp.it>


> Hi Dear RegCM Users
> 
> I am trying to run RegCM in mpirun mode, But when I set the namelist, 
> make terrain, sst and icbc and go to the run, I got this error:
> 
> MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD 
> with errorcode 1.
> 
> NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
> You may or may not see output from other processes, depending on
> exactly when Open MPI kills them.
> --------------------------------------------------------------------------
> --------------------------------------------------------------------------
> mpirun has exited due to process rank 0 with PID 24514 on
> node saeed exiting without calling "finalize". This may
> have caused other processes in the application to be
> terminated by signals sent by mpirun (as reported here).
> 
> my system is DELL studio, core i 7, 4 G ram and command for run in 
> mpirun is:
> 
> mpirun -np 7 ./Bin/regcmMPI regcm.in >regcm.out
> 
> any help?
> 
> Regards
> 
> Saeed Bayat
> M.A Student of Climatology
> Department of Geography
> Ferdowsi University of Mashhad(Iran)
> _______________________________________________
> RegCNET mailing list
> RegCNET at lists.ictp.it