wiki.chpc.ac.za

User Tools

Site Tools

You are here: start » quick » pbspro
Trace: • loadshedding • walltime • new • root • moab • connect • lustre • logo • compiling_octave_3_2_4 • pbspro
quick:pbspro

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
quick:pbspro [2015/04/02 11:38]
ischeepers [Example PBS job submission scripts] 		quick:pbspro [2025/09/22 17:47] (current)
kevin [Example 4 : empty job]
Line 1: Line 1:
 ====== Job submission using PBSPro ====== ====== Job submission using PBSPro ======
  
-As of January 2014, CHPC has a new scheduler - PBSPro.+The CHPC system uses the PBSPro scheduler.
  
-Main user documentation can be found at http://resources.altair.com/pbs/documentation/support/PBSProUserGuide12.1.pdf+Main user documentation can be found at [[https://www.altair.com/pbs-works-documentation/|Altair]]
  
 But here is a summary of how to submit jobs at CHPC. But here is a summary of how to submit jobs at CHPC.
Line 18: Line 18:
 ====How to submit a batch job==== ====How to submit a batch job====
  
-Use the qsub command to submit a script file. PBSPro job file allows for specification of options at the beginning of the file prefaced by the #PBS delimiter followed by PBS commands. Commands will be executed in the order given in the script.+Use the ''qsub'' command to submit a script file to the scheduler. PBS-Pro jobs can be controlled using command line switches, //or// directives in the job submission file prefaced by the ''#PBS'' keyword. 
  
-A script can include just about anything which you could do during a terminal session such as set environment variables, change directories, and move files. See the qsub man page for a complete list of options.+See the ''qsub'' man page for a complete list of options. 
  
-''qsub myscriptname''+The first line of a script file may specify the interpreter, eg., bash is selected by using ''#!/bin/bash'' as the first line of the job script file.
  
-The job number will be returned if the job has been accepted by the scheduler. Use this +A script can include just about anything which you could do during a terminal session such as set environment variables, change directories, and move files. 
-number to check on the progress of a specific job:+
  
-''qstat -f <jobnum>''+> Job script files can be named to anything allowed for a regular file, but it is strongly recommended that you use ''.pbs'' as the file extension to make it easy to distinguish job scripts from other shell scripts (''.sh'').
  
-To inspect jobs that have already finished: 
  
-''qstat -x -f <jobnum>''+====How to start an interactive job on a compute node====
  
-=====PBS-Pro Job Exit Codes =====+''qsub -I -V''
  
-====Job Exit Codes Between 0 and 128 (or 256)==== +to start an interactive session, retaining all environment variables.
-This is the exit value of the top process in the job, typically the shell. This may be the exit value of the last command executed in the shell or the .logout script if the user has such a script (csh).+
  
-====Job Exit Codes >= 128 (or 256)==== 
-This means the job was killed with a signal. The signal is given by X modulo 128 (or 256). For example an exit value of 137 means the job's top process was killed with signal 9 (137 mod 128 = 9). 
-The exit status values greater than 128 (or 256) indicate which signal killed the job. Depend- ing on the system, values greater than 128 (or on some systems 256; see man wait(2) for more information), are the value of the signal that killed the job. (i.e., 265 mod 256 = 9 or 271 mod 256 = 15.) 
  
-To interpret (or “decode”) the signal contained in the exit status value, subtract the base value from the exit status. For example, if a job had an exit status of 143, that indicates the job was killed via a SIGTERM (e.g. 143 - 128 = 15, signal 15 is SIGTERM). See the kill(1) manual page for a mapping of signal numbers to signal name on your operating system.+====How to monitor job submissions to the queue====
  
-====Terminate a job by signal==== +The job number will be returned if the job has been accepted by the scheduler. Use this 
-Send a signal to the job+number to check on the progress of a specific job:
  
-'' qsig -s SIGTERM  <jobnum>  ''+''qstat -f <jobnum>''
  
-=====Example PBS job submission scripts=====+To inspect jobs that have already finished:
  
-See [[howto:PBS-Pro Job submission examples]].+''qstat -x -f <jobnum>''
  
 +or
  
-Job submission scripts are shell scripts with PBS directives in comments. +''qstat -xf <jobnum>''
  
-**PLEASE NOTE:**  +====How to monitor job arrays====
-  - The first line should always have the hash-bang and shell path, eg. **#!/bin/bash** or **#!/bin/sh** +
-  - The output and error file paths (''#PBS -o'' and ''#PBS -e'' directives) should be files in your scratch directory, or else the job will be killed automatically. (There should be a symbolic link called "scratch" in your home directory that points to the workspace on the Lustre file system.) +
-  - Job names (''#PBS -N '') longer than 15 characters don't work +
-  +
-On a cluster system, run an executable on 3 nodes using all 8 cores per node with 8 MPI processes for 5 hours:+
  
-<code> +To see the array job status on the queue, postfix the 
-#!/bin/bash +jobnumber with empty square brackets:
-#PBS -e scratch/error.out +
-#PBS -o scratch/outputfile.txt +
-#PBS -j oe +
-#PBS -l walltime=5:00:00 +
-#PBS -l select=3:ncpus=8:mpiprocs=8 +
-#PBS -V+
  
-source /etc/profile.d/modules.sh +''qstat -J <jobnum>[]''
-module add openmpi/openmpi-1.6.5_gcc-4.7.2+
  
-cd $HOME/scratch 
  
-mpirun -np 24  /opt/gridware/bioinformatics/bin/hello+To look at one specific job in the array, add ''-t'' to the qstat arguments:
  
-</code>+''qstat -xft <jobnum>[array-index]''
  
-===How to specify number of nodes and cores for a job=== 
  
-This is done using the ''#PBS -l select'' directive in the job submission script.  +eg
-Cluster partitions differ in how many cpu's per node there are: +
-  * Harpertown: 8 cores per node +
-  * Nehalem: 8 cores per node +
-  * Westmere: 12 cores per node +
-  * Dell: 12 cores per node +
-  +
-To run a 24-core job on the Dell partition, for instance, only 2 nodes are needed:+
  
-''#PBS -l select=2:ncpus=12:mpiprocs=12:jobtype=dell:group=nodetype +''qstat -xft 967890[2]''
-''+
  
-===Number of cluster nodes in a job=== 
  
-The number of cores can be obtained by: 
  
-''NP=`cat $PBS_NODEFILE | uniq | wc -l` +=====PBS-Pro Job Exit Codes =====
-''+
  
 +====Job Exit Codes Between 0 and 128 (or 256)====
 +This is the exit value of the top process in the job, typically the shell. This may be the exit value of the last command executed in the shell or the .logout script if the user has such a script (csh).
  
-===Number of cores (cpu's) in a job===+====Job Exit Codes >= 128 (or 256)==== 
 +This means the job was killed by the PBS scheduler by sending it a signal. 
  
-To determine the number of cores, include the following line in the submission script:+The **modulo** or **remainder** operator [[http://en.wikipedia.org/wiki/Modulo_operation]] is used to interpret (or “decode”) PBS job exit values. 
  
-''set NP = `cat $PBS_NODEFILE | wc -l`'' 
  
 +The signal is given by X modulo 128 (or 256). An exit value of 137 means the job's top process was killed with signal 9 because ''137 mod 128 = 9''. 
  
-===Determining the machinefile (list of nodes)===+Depending on the system, a value greater than 128 or 256, see ''man wait(2)'', is the signal that was sent by PBS-Pro to kill the job. (i.e., 265 mod 256 = 9 or 271 mod 256 = 15.) 
  
-The mpirun command can use the PBS variable ''$PBS_NODEFILE'' that contains the  +Exit code 271 means that the job exceeded the limit requested at submission, eg. walltime or cpu usage, and was killed.
-file name with the list of nodes:+
  
 +See ''man kill'' for a mapping of signal numbers to signal name on your operating system.
 +====Terminate a job by signal====
 +Send a signal to the job
  
-''mpirun -np $NP -machinefile $PBS_NODEFILE ~/bin/myprogram +'' qsig -s SIGTERM  <jobnum>  ''
-''+
  
  
 +===== Examples: PBS-Pro job submission scripts =====
  
-On a non-cluster system, run an executable on 24 cores for 5 hours +[[howto:PBS-Pro_job_submission_examples]]
- +
-<code> +
-#!/bin/bash +
-#PBS -j oe +
-#PBS -l walltime=5:00:00 +
-#PBS -l ncpus=24 +
-#PBS -V +
- +
-cd $HOME/scratch5 +
-</code> +
- +
-===Using Modules=== +
- +
-To use the module system, add the following line to your submission script after +
-the #PBS lines: +
- +
-''source /etc/profile.d/modules.sh +
-'' +
- +
-Then you will be able to include your required module to set up an environment: +
- +
-''module add openmpi/openmpi-1.6.5_gcc-4.7.2 +
-'' +
- +
-To list modules, the MODULEPATH variables normally are set by default and can be viewed  +
-from the command line using: +
- +
-''module avail +
-'' +
- +
-For Bioinformatics users, add the additional modules in your .profile as follows: +
- +
- +
-''export MODULEPATH=/opt/gridware/bioinformatics/modules:$MODULEPATH +
-'' +
- +
-===How long can a job run=== +
- +
-There is no limit to the amount of time that a PBS job can run. If no time limit is specified, a default of 12 hours is assigned. Remember to check point long running jobs. +
- +
-To specify a time, include the following line in the submission script: +
- +
-''#PBS -l walltime=hhhh:mm:ss'' +
- +
-===Exclusive use of cluster nodes=== +
- +
-On cluster systems, if a job does not request all the cores on a node, it is possible another job will share the same node. To prevent this, request exclusive use. +
- +
-''#PBS -l place=excl'' +
- +
-===How to export interactive session environment to job=== +
- +
-To include the shell environment in the batch job, either use the -V option with qsub or as a PBS directive +
- +
-''qsub -V'' +
-         +
-''#PBS -V'' +
- +
-===Hyper-Threading=== +
- +
-On the CHPC cluster, Hyperthreading is disabled, except on the MIC nodes. +
- +
-===Suppose I want to do a parameter study. How do I submit all these jobs with a different value of parameter?=== +
- +
-Use PBS Job Arrays. You can specify how many jobs to run by adding the directive +
- +
-#PBS -J <range> +
- +
-where range is X-Y:Z. So 2-10:2 indicates all the even jobs from 2 to 10, i.e., 2,4,6,8 and 10. +
- +
-PBS defines two environment variables: +
- +
-PBS_ARRAY_INDEX     job array index +
- +
-PBS_ARRAY_ID        job array id +
- +
-These variables are also defined as attributes: +
- +
-array_index +
- +
-array_id +
- +
-Based on the job array index, different input files can be used for each job in the job array. Below is a job script example that submits two jobs, each using one processor. +
- +
-NOTE: The PBS directives specify the resources that will be used by EACH individual job, NOT all the jobs together. +
- +
-<code> +
-#!/bin/sh +
-#PBS -V +
-#PBS -l select=1:ncpus=1:mpiprocs=1,walltime=48:00 +
-#PBS -N Job_Array_Test +
-#PBS -j oe -o ja.^array_index^.pbs +
-#PBS -J 1-2 +
- +
-source /etc/profile.d/modules.sh +
-module add openmpi/openmpi-1.6.5_gcc-4.7.2 +
-unset echo +
- +
-cd $PBS_O_WORKDIR              +
-echo $PBS_ARRAY_INDEX                +
-echo $PBS_ARRAY_ID $PBS_ARRAY_INDEX >> ja.$PBS_ARRAY_INDEX.out +
-echo ' '        >> ja.$PBS_ARRAY_INDEX.out +
-/opt/gridware/bioinformatics/bin/pi < pi.inp >> ja.$PBS_ARRAY_INDEX.out +
-                 +
-exit +
-</code> +
- +
- +
-===A job submitted to the queue is not running=== +
- +
-Look at the job status with qstat. The next to last line of the output is a comment describing the job status. Possible outputs include: +
- +
-''qstat -f yourjobid'' +
- +
-  *  Job run at date at time on hostname +
-        Job is running OK  +
- +
-  *  Not Running: No available resources on nodes +
-        Job requires more memory,cores, or processors than currently available, or you have requested more resources than what physically exists on the system.  +
- +
-  *  Job held, too many failed attempts to run +
-        Delete and resubmit job +
- +
-If this persists, contact User support  +
- +
-  *  Not Running, Draining system to allow starving job to run +
-        Job will not run because the resources are reserved for other jobs.  +
- +
-===Queued Jobs stuck in an error state (E)=== +
- +
-Try deleting the job with qdel +
- +
-''qdel -Wforce yourjobid'' +
- +
-The following must be **plain text** files, as created by unix programs like ''nano'' or ''vi''. If you wish to create them on your desktop and copy them over, ensure they contain no formatting and have unix line endings. Microsoft Word would be a bad choice. +
- +
- +
- +
-===== Examples 1: Gaussian jobs on 1 node or more than 1 node ===== +
  
 +===== Examples: Gaussian jobs =====
  
  
Line 273: Line 109:
  
  
-===== Example 2 (a) :  Amber job on GPU nodes (NB: For other executables e.g. python and sender, see example 2(b)) =====+===== Examples: Amber jobs =====
  
-<code> 
-#!/bin/sh 
-#PBS -N eric.job 
-#PBS -l select=1:ncpus=12:mpiprocs=12 
-#PBS -l walltime=1:00:00 
-#PBS -q kepla_k20 
-#PBS -m be 
-#PBS -o /export/home/username/amber/amber12/eric.out 
-#PBS -e /export/home/username/amber/amber12/eric.err 
-#PBS -M email@mail.com 
  
-cd $PBS_O_WORKDIR+[[howto:amber|Amber job submission scripts]]
  
-pwd 
-echo "My job starts here" 
-date 
-source /etc/profile.d/modules.sh 
-module add intel-XE/c8000 cuda/5.0-c8000 mvapich2/c8000 amber/k20 
-module add amber/k20 
  
-exe=pmemd.cuda.MPI 
-nproc=`cat $PBS_NODEFILE|wc -l` 
-time mpirun -np $nproc -machinefile $exe -O -i mdin -o mdout -p prmtop -c inpcrd 
-echo "My job ends here" 
-date 
-</code> 
  
-===== Example 2 (b) :  Amber job on normal nodes (Please use this for python,sander and other executables that do not work on GPU=====+===== Example: Quantum Espresso =====
  
-<code> 
-#!/bin/sh 
-#PBS -N eric.job 
-#PBS -l select=1:ncpus=12:mpiprocs=12 
-#PBS -l walltime=1:00:00 
-#PBS -q workq 
-#PBS -m abe 
-#PBS -o /export/home/username/amber/amber12/eric.out 
-#PBS -e /export/home/username/amber/amber12/eric.err 
-#PBS -M email@mail.com 
  
-cd $PBS_O_WORKDIR+[[howto:quantum_espresso|Quantum Espresso]]
  
-pwd 
- 
-source /etc/profile.d/modules.sh 
-module add intel-XE/13.0 openmpi/openmpi-1.6.5-intel amber/12 
- 
-echo "My job starts here" 
-date 
- 
-exe=sander.MPI 
- 
-nproc=`cat $PBS_NODEFILE|wc -l` 
-time mpirun -np $nproc -machinefile $exe -O -i mdin -o mdout -p prmtop -c inpcrd 
-echo "My job ends here" 
-date 
-</code> 
- 
-===== Example 3 : Quantum Espresso job ===== 
- 
-<code> 
-#!/bin/sh 
-###These lines are for PBSpro 
-#PBS -N QEspresso 
-#PBS -l select=2:ncpus=12:mpiprocs=12:jobtype=dell,place=free:group=nodetype:excl 
-#PBS -l walltime=24:00:00 
-#PBS -q workq 
-#PBS -m be 
-#PBS -o /export/home/username/scratch/run-directory/stdout 
-#PBS -e /export/home/username/scratch/run-directory/stderr 
-#PBS -V 
-##### Environment Settings 
-source /etc/profile.d/modules.sh 
-module add intel-XE/13.0 
-module add QEspresso/5.0.2 
-##### Running commands 
-NP=`cat $PBS_NODEFILE | wc -l` 
-cd $HOME/scratch5 
-mpirun -f $PBS_NODEFILE -r ssh -ib -env I_MPI_DEBUG 2 -np $NP pw.x -npool 3  -inp inputfile.inp  
-</code> 
  
-===== Example 4 :  empty job =====+===== Example 4 : empty job =====
  
 <code> <code>
Line 375: Line 141:
 #<hosts> = number of hosts #<hosts> = number of hosts
 #<architecture> = architecture where the jobs will run (nehalem, westmere, dell) #<architecture> = architecture where the jobs will run (nehalem, westmere, dell)
-#<queue> = submission queue ( workq , priorityq , specialq , intel_mic , spark , kepla_k20 , accelrys)+#<queue> = submission queue ( serial, smp, normal, ...,  accelrys)
 #<executable> = Binary to be run #<executable> = Binary to be run
  
Line 382: Line 148:
 </code> </code>
  
-//Note that the M9000 sparc queue is called 'spark'// 
  
-===== MPI example =====+ 
 +===== Example 5 : MPI example =====
  
 Syntax Syntax
Line 394: Line 160:
 ... ...
 </code> </code>
- 
 ===== Further examples ===== ===== Further examples =====
   * Various [[howto:bioinformatics?&#basic_examples|bioinformatics examples]] illustrate some further possibilities. Specifically the blast examples illustrate job arrays and job dependencies.   * Various [[howto:bioinformatics?&#basic_examples|bioinformatics examples]] illustrate some further possibilities. Specifically the blast examples illustrate job arrays and job dependencies.
   * This [[howto:gpu_gromacs#pbs_job_script|gromacs example]] shows an interesting case where mpi, openmp and the gpus are used in a single job.   * This [[howto:gpu_gromacs#pbs_job_script|gromacs example]] shows an interesting case where mpi, openmp and the gpus are used in a single job.
/app/dokuwiki/data/attic/quick/pbspro.1427967538.txt.gz · Last modified: 2021/12/09 16:42 (external edit)

Page Tools

Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Noncommercial-Share Alike 3.0 Unported
CC Attribution-Noncommercial-Share Alike 3.0 Unported Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki