HPC/Submitting and Managing Jobs: Difference between revisions

From CNM Wiki
< HPC
Jump to navigation Jump to search
Line 132: Line 132:
so you can omit the <code>-machinefile</code> and <code>-np</code> options if you wish.  (May not actually work. --[[User:Stern|stern]])
so you can omit the <code>-machinefile</code> and <code>-np</code> options if you wish.  (May not actually work. --[[User:Stern|stern]])
-->
-->
=== Requesting fewer cores per node ===
An application may require fewer cores than the whole node offers, on Carbon <code>ppn&lt;8</code>, for example, due to the following reasons:
* the code is not parallelized
* the code is parallelized for e.g. 2 or 4 cores only
* the application runs multi-threaded
* the node's memory is not sufficient to run 8 processes.
; Shared node access:
: This is the default.  A job requests fewer cores than the node has; free cores can be allocated to other jobs.  Simply specify <code>ppn</code> less than 8:
#PBS -l nodes=__:ppn=4
; Exclusive node access:
: This may be needed when a job requires nearly the whole memory of a node, or nearly all of its I/O bandwidth.
  #PBS -l nodes=__:ppn=4 '''-l naccesspolicy=SINGLEJOB'''
; Multithreading (OpenMP):
#PBS -l nodes=__:ppn=1
...
cd $PBS_O_WORKDIR
export OMP_NUM_THREADS=`uniq -c $PBS_NODEFILE | awk '{print $1; exit}'`
...
; Hybrid OpenMP + MPI:
This is still a very new field and subject to ongoing research.  Usage is the same as the preceding case (OpenMP only), plus you must export <code>OMP_NUM_THREADS</code> to all MPI satellite nodes:
#PBS -l nodes=__:ppn=1
...
cd $PBS_O_WORKDIR
export OMP_NUM_THREADS=`uniq -c $PBS_NODEFILE | awk '{print $1; exit}'`
...
mpiexec -x OMP_NUM_THREADS …    # OpenMPI-specific
For either case the following holds:
* The <code>$PBS_NODEFILE</code> seen by the job script will always match <code>ppn</code>.
* Forr accounting, the job will be ''billed'' by the number of cores blocked from use by other users, i.e., the actual ppn for shared nodes, and ppn=8 for naccesspolicy=SINGLEJOB.


=== Policies ===
=== Policies ===

Revision as of 22:25, November 5, 2010

Directories and Environment

First read: directory configuration.

Applications

We use the environment-modules package to manage user applications. This is similar to places like NERSC or PNNL. The basic CNM-specific user environment is configured automatically in /etc/profile.d/cnm.{sh,csh}.

For now the only applications are the Development tools.

Admin note: The master copy of these files resides in mgmt{01,02}:/opt/teamhpc/node-skel/etc/profile.d and is distributed by ~root/bin/skeldistrib.

Submitting jobs to Moab/Torque

 qsub [-A accountname] [options] jobfile

For details on options:

 man qsub
 qsub --help

More details are at the Torque Manual, in particular the qsub man page.

The single main queue is batch and need not be specified. All job routing decisions are handled by the scheduler. In particular, short jobs are accommodated by a daily reserved node and by backfill scheduling, i.e. "waving forward" while a big job waits for full resources to become available.

Debug queue

For testing job processing and your job environment, use qsub -q debug or #PBS -q debug. The queue accepts jobs under the following conditions

nodes <= 2
ppn <= 4
walltime <= 1:00:00

Querying jobs

Use the command qstat (from PBS) or showq (from Moab):

qstat [-u $USER]
showq [-u $USER]
regular output
qstat -a
showq -n
alternate format (showing names)
qstat -f [jobnum]
full information
checkjob [-v] jobnum
get extended jobs status information – useful to diagnose problems with "stuck" jobs.

Removing jobs

 qdel jobnumber

Example job file

  • sample job file for Infiniband interconnect (recommended):
#!/bin/bash

##  Basics: Number of nodes, processors per node (ppn), and walltime (hhh:mm:ss)
#PBS -l nodes=5:ppn=8
#PBS -l walltime=0:10:00
#PBS -N job_name
#PBS -A account

## File names for stdout and stderr.  If not set here, the defaults
## are <JOBNAME>.o<JOBNUM> and <JOBNAME>.e<JOBNUM>
#PBS -o job.out
#PBS -e job.err

## send mail at begin, end, abort, or never (b, e, a, n)
#PBS -m ea

# change into the directory where qsub will be executed
cd $PBS_O_WORKDIR

# count allocated cores
NPROCS=`wc -l < $PBS_NODEFILE`

# start MPI job over default interconnect
mpirun -machinefile $PBS_NODEFILE -np $NPROCS \
        programname
  • If your program reads from files or takes options and/or arguments, use and adjust one of the following forms
mpirun -machinefile $PBS_NODEFILE -np $NPROCS \
       programname  < run.in
mpirun -machinefile $PBS_NODEFILE -np $NPROCS \
       programname  -options arguments < run.in
mpirun -machinefile $PBS_NODEFILE -np $NPROCS \
       programname < run.in > run.out 2> run.err
In this form, anything after programname is optional. If you use specific redirections for stdout or stderr as shown (>, 2>), the job-global files job.out, job.err declared earlier will remain empty or only contain output from your shell startup files (which should really be silent), and the rest of your job script.
  • Infiniband (OpenIB) is the default (and fast) interconnect mechanism for MPI jobs. This is configured through the environment variable $OMPI_MCA_btl.
  • To select ethernet transport (e.g. for embarrasingly parallel jobs), specify an -mca option:
mpirun -machinefile $PBS_NODEFILE -np $NPROCS \
	-mca btl self,tcp \
        programname

The account parameter

The parameter for option -A account is in most cases the CNM proposal, specified as follows:

cnm123
(3 digits) for proposals below 1000
cnm01234
(5 digits, 0-padded) for proposals from 1000 onwards.
user
for a limited personal startup allocation
staff
for discretionary access by staff.

You can check your account balance in hours as follows:

mybalance -h
gbalance -u $USER -h

Using OpenMP

For hybrid MPI/OpenMP operation under PBS (which is what happens when linking the MKL with OpenMP), two adjustments are necessary:

  1. The environment variable OMP_NUM_THREADS needs to be set to the number of available cores per node, i.e., the ppn parameter. By default, this variable is set to 1 to select single-threading of OpenMP-compiled applications.
  2. The machinefile needs to be thinned out in the job file, to have each node listed only once.

Example

#!/bin/bash
#PBS -l nodes=nnn:ppn=8
...
MACHINEFILE=$PBS_NODEFILE
...
if [ multithreaded ]            # insert specific condition
then
    sort -u $MACHINEFILE > machinefile
    MACHINEFILE=machinefile
    export OMP_NUM_THREADS=8
fi
...
NPROC=`wc -l < $MACHINEFILE`
...

Hybrid MPI+OpenMP is currently unsupported and may well be less efficient than compiling and running with MPI-only communication.

Requesting fewer cores per node

An application may require fewer cores than the whole node offers, on Carbon ppn<8, for example, due to the following reasons:

  • the code is not parallelized
  • the code is parallelized for e.g. 2 or 4 cores only
  • the application runs multi-threaded
  • the node's memory is not sufficient to run 8 processes.
Shared node access
This is the default. A job requests fewer cores than the node has; free cores can be allocated to other jobs. Simply specify ppn less than 8:
#PBS -l nodes=__:ppn=4
Exclusive node access
This may be needed when a job requires nearly the whole memory of a node, or nearly all of its I/O bandwidth.
 #PBS -l nodes=__:ppn=4 -l naccesspolicy=SINGLEJOB
Multithreading (OpenMP)
#PBS -l nodes=__:ppn=1
...
cd $PBS_O_WORKDIR
export OMP_NUM_THREADS=`uniq -c $PBS_NODEFILE | awk '{print $1; exit}'`
...
Hybrid OpenMP + MPI

This is still a very new field and subject to ongoing research. Usage is the same as the preceding case (OpenMP only), plus you must export OMP_NUM_THREADS to all MPI satellite nodes:

#PBS -l nodes=__:ppn=1
...
cd $PBS_O_WORKDIR
export OMP_NUM_THREADS=`uniq -c $PBS_NODEFILE | awk '{print $1; exit}'`
...
mpiexec -x OMP_NUM_THREADS …    # OpenMPI-specific

For either case the following holds:

  • The $PBS_NODEFILE seen by the job script will always match ppn.
  • Forr accounting, the job will be billed by the number of cores blocked from use by other users, i.e., the actual ppn for shared nodes, and ppn=8 for naccesspolicy=SINGLEJOB.

Policies

  • Direct user access to nodes is only possible while a job is running for that user. This is governed by the torque-pam package.