How to submit an SGE job to the cloud using the Bright CMSUB command

    

This article shows how to submit a simple OGS (SGE) job to the cloud using the Bright CMSUB command.

Let's get started.



Here's the job script we'll use. It's a standard OGS job script with some added #CMSUB directives. These #CMSUB directives are ignored by the shell, and by OGS, but they provide a convenient way to pass arguments to the Bright CMSUB command.

#!/bin/sh

#$ -N ogs-job.sh
#$ -j yes
#$ -cwd
#$ -S /bin/bash

#CMSUB --input=/home/rstober/ogs/cloud/test.txt
#CMSUB --output=/home/rstober/ogs/cloud/analysis.output
#CMSUB --regions=us-west-1
#CMSUB --arguments="-q cloud"
#CMSUB --wlmanager=sge

date >> test.txt
echo "This job is running on node: `hostname`"

echo
echo "In the directory: `pwd`"

echo
echo "As the user: `whoami`"

echo "this is a file created by the job" > analysis.output

echo
echo "Here is a listing of the current directory:"
ls -l

echo
echo "This is the input file upload by cmsub, and modified by this job:"
echo
cat test.txt

sleep 30

Submit the job using the Bright cmsub command.

[rstober@atom-head1 cloud]$ cmsub ogs-job.sh
  Upload job id: 75
    User job id: 76
Download job id: 77

The qstat command shows the current status of the available SGE queues and the jobs associated with the queues. When called without arguments, qstat displays a list of running and pending jobs without queue status information.

[rstober@atom-head1 ogs]$ qstat
job-ID  prior   name       user         state submit/start at     queue                          slots ja-task-ID
-----------------------------------------------------------------------------------------------------------------
     75 0.00000 ogs-job.sh rstober      qw    01/21/2013 10:44:05                                    1
     76 0.00000 ogs-job.sh rstober      hqw   01/21/2013 10:44:05                                    1
     77 0.00000 ogs-job.sh rstober      hqw   01/21/2013 10:44:05                                    1

The qstat -f option causes summary information on all queues to be displayed along with the queued job list.

[rstober@atom-head1 ogs]$ qstat -f
queuename                      qtype resv/used/tot. load_avg arch          states
---------------------------------------------------------------------------------
all.q@atom01.cm.cluster        BIP   0/0/4          0.00     linux-x64
---------------------------------------------------------------------------------
all.q@atom02.cm.cluster        BIP   0/0/4          0.00     linux-x64
---------------------------------------------------------------------------------
all.q@atom03.cm.cluster        BIP   0/0/4          0.00     linux-x64
---------------------------------------------------------------------------------
all.q@atom04.cm.cluster        BIP   0/0/4          -NA-     -NA-          au
---------------------------------------------------------------------------------
cloud@cnode1.cm.cluster        BI    0/1/4          0.09     linux-x64
     76 0.55500 ogs-job.sh rstober      r     01/21/2013 10:44:31     1
---------------------------------------------------------------------------------
cloudtransfers@master.cm.clust BI    0/0/1          0.28     linux-x64

############################################################################
 - PENDING JOBS - PENDING JOBS - PENDING JOBS - PENDING JOBS - PENDING JOBS
############################################################################
     77 0.00000 ogs-job.sh rstober      hqw   01/21/2013 10:44:05     1

qstat -j provides detailed information about the specified job.

[rstober@atom-head1 ogs]$ qstat -j 76
==============================================================
job_number:                 76
exec_file:                  job_scripts/76
submission_time:            Mon Jan 21 10:44:05 2013
owner:                      rstober
uid:                        1001
group:                      rstober
gid:                        1001
sge_o_home:                 /home/rstober
sge_o_log_name:             rstober
sge_o_path:                 /cm/shared/apps/gcc/4.7.0/bin:/usr/lib64/qt-3.3/bin:/usr/local/bin:/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/sbin:/sbin:/usr/sbin:/cm/shared/apps/openmpi/gcc/64/1.4.5/bin:/cm/local/apps/cluster-tools/bin:/cm/shared/apps/slurm/2.3.4/bin:/cm/shared/apps/slurm/2.3.4/sbin:/home/rstober/bin:/cm/shared/apps/sge/2011.11/bin/linux-x64
sge_o_shell:                /bin/bash
sge_o_workdir:              /home/rstober/ogs
sge_o_host:                 atom-head1
account:                    sge
cwd:                        /home/rstober/ogs
merge:                      y
hard resource_list:         cloudregion=us-west-1
mail_list:                  rstober@master.cm.cluster
notify:                     FALSE
job_name:                   ogs-job.sh
jobshare:                   0
hard_queue_list:            cloud
shell_list:                 NONE:/bin/bash
env_list:
script_file:                /tmp/tmpd62nmm.job
jid_predecessor_list (req):  75
jid_successor_list:          77
usage    1:                 cpu=00:00:00, mem=0.00000 GBs, io=0.00000, vmem=N/A, maxvmem=N/A
scheduling info:            queue instance "all.q@atom04.cm.cluster" dropped because it is temporarily not available

We're done.

Bright OpenStack