[GE users] Help: OpenMPI Integration Problem

Lee Amy openlinuxsource at gmail.com
Fri May 23 06:47:15 BST 2008


    [ The following text is in the "ISO-8859-1" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some special characters may be displayed incorrectly. ]

Hello,

I make an OpenMPI tight integration parallel environment. I'm sure that I've
enabled gridengine support for OpenMPI 1.2.6. However, when I submit job
script the qstat shows the task is runing, but the program which is in
script dosen't run any more.

Now I will paste the relative files.

Parallel Environment
######
pe_name           openmpi
slots             20
user_lists        NONE
xuser_lists       NONE
start_proc_args   /bin/true
stop_proc_args    /bin/true
allocation_rule   $round_robin
control_slaves    FALSE
job_is_first_task TRUE
urgency_slots     min
######

Job Script
######
#!/bin/bash

/usr/local/openmpi/bin/mpirun --mca pls_agent_rsh rsh -np $NSLOTS
/usr/local/clustalw-mpi/clustalw-mpi -infile=/usr/local/test/2000
######

Error Output
######
error: executing task of job 5 failed:
[gnode4:15774] ERROR: A daemon on node gnode1 failed to start as expected.
[gnode4:15774] ERROR: There may be more information available from
[gnode4:15774] ERROR: the 'qstat -t' command on the Grid Engine tasks.
[gnode4:15774] ERROR: If the problem persists, please restart the
[gnode4:15774] ERROR: Grid Engine PE job
[gnode4:15774] ERROR: The daemon exited unexpectedly with status 1.
error: executing task of job 5 failed:
[gnode4:15774] ERROR: A daemon on node gnode2 failed to start as expected.
[gnode4:15774] ERROR: There may be more information available from
[gnode4:15774] ERROR: the 'qstat -t' command on the Grid Engine tasks.
[gnode4:15774] ERROR: If the problem persists, please restart the
[gnode4:15774] ERROR: Grid Engine PE job
[gnode4:15774] ERROR: The daemon exited unexpectedly with status 1.
error: executing task of job 5 failed:
[gnode4:15774] ERROR: A daemon on node gnode3 failed to start as expected.
[gnode4:15774] ERROR: There may be more information available from
[gnode4:15774] ERROR: the 'qstat -t' command on the Grid Engine tasks.
[gnode4:15774] ERROR: If the problem persists, please restart the
[gnode4:15774] ERROR: Grid Engine PE job
[gnode4:15774] ERROR: The daemon exited unexpectedly with status 1.
######

Shall I configure anything else?

Thank you very much~

Regards,

Amy Lee



More information about the gridengine-users mailing list