cluster:119
Warning: Undefined array key -1 in /usr/share/dokuwiki/inc/html.php on line 1458
Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
cluster:119 [2013/08/21 11:03] hmeij |
cluster:119 [2021/06/17 15:32] (current) hmeij07 |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| \\ | \\ | ||
| **[[cluster: | **[[cluster: | ||
| - | |||
| - | Jobs need to be submitted to the scheduler on host sharptail itself for now and will be dispatched to nodes n33-n37 in queue mwgpu. | ||
| - | --- // | ||
| ==== Submitting GPU Jobs ==== | ==== Submitting GPU Jobs ==== | ||
| + | |||
| + | Please plenty of time between multiple GPU job submissions. | ||
| + | |||
| + | Jobs need to be submitted to the scheduler via cottontail to queues mwgpu, amber128, exx96. | ||
| + | |||
| + | This page is old, the gpu resource '' | ||
| + | --- // | ||
| + | |||
| + | **Articles** | ||
| + | |||
| + | * [[http:// | ||
| + | * [[http:// | ||
| + | |||
| Line 44: | Line 54: | ||
| </ | </ | ||
| - | With '' | + | With '' |
| < | < | ||
| Line 58: | Line 68: | ||
| 3 Tesla K20m 21 C 0 % | 3 Tesla K20m 21 C 0 % | ||
| ==================================================== | ==================================================== | ||
| + | |||
| + | [hmeij@sharptail sharptail]$ ssh n33 gpu-free | ||
| + | 1,3,0 | ||
| + | |||
| + | |||
| </ | </ | ||
| Line 116: | Line 131: | ||
| #!/bin/bash | #!/bin/bash | ||
| # submit via 'bsub < run.gpu' | # submit via 'bsub < run.gpu' | ||
| - | rm -f mdout.[0-9]* auout.[0-9]* | + | rm -f mdout.[0-9]* auout.[0-9]* apoa1out.[0-9]* |
| #BSUB -e err | #BSUB -e err | ||
| #BSUB -o out | #BSUB -o out | ||
| Line 122: | Line 137: | ||
| #BSUB -J test | #BSUB -J test | ||
| - | ## leave sufficient time between job submissions (30-60 secs) | + | # from greentail we need to set up the module env |
| - | ## the number of GPUs allocated matches | + | export PATH=/ |
| - | ## always reserve GPU (gpu=1), setting this to 0 is a cpu job only | + | / |
| - | ## reserve 6144 MB (5 GB + 20%) memory per GPU | + | / |
| - | ## run all processes (1< | + | / |
| + | / | ||
| + | / | ||
| + | / | ||
| + | export LD_LIBRARY_PATH=/ | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| - | #BSUB -n 1 | ||
| - | #BSUB -R " | ||
| - | |||
| - | # unique job scratch dirs | ||
| - | MYSANSCRATCH=/ | ||
| - | MYLOCALSCRATCH=/ | ||
| - | export MYSANSCRATCH MYLOCALSCRATCH | ||
| - | cd $MYSANSCRATCH | ||
| - | |||
| - | # AMBER | ||
| - | # stage the data | ||
| - | cp ~/ | ||
| - | # feed the wrapper | ||
| - | lava.mvapich2.wrapper pmemd.cuda.MPI \ | ||
| - | -O -o mdout.$LSB_JOBID -inf mdinfo.1K10 -x mdcrd.1K10 -r restrt.1K10 -ref inpcrd | ||
| - | # save results | ||
| - | cp mdout.[0-9]* ~/ | ||
| - | |||
| - | # LAMMPS | ||
| - | # GPUIDX=1 use allocated GPU(s), GPUIDX=0 cpu run only (view header au.inp) | ||
| - | export GPUIDX=1 | ||
| - | # stage the data | ||
| - | cp ~/ | ||
| - | # feed the wrapper | ||
| - | lava.mvapich2.wrapper lmp_nVidia \ | ||
| - | -c off -var GPUIDX $GPUIDX -in au.inp -l auout.$LSB_JOBID | ||
| - | # save results | ||
| - | cp auout.[0-9]* ~/ | ||
| - | |||
| - | </ | ||
| - | |||
| - | |||
| - | ==== lava.mvampich2.wrapper ==== | ||
| - | |||
| - | < | ||
| - | |||
| - | #!/bin/bash | ||
| - | # submit via 'bsub < run.gpu' | ||
| - | rm -f mdout.[0-9]* auout.[0-9]* apoa1out.[0-9]* | ||
| - | #BSUB -e err | ||
| - | #BSUB -o out | ||
| - | #BSUB -q mwgpu | ||
| - | #BSUB -J test | ||
| ## leave sufficient time between job submissions (30-60 secs) | ## leave sufficient time between job submissions (30-60 secs) | ||
| Line 217: | Line 198: | ||
| # save results | # save results | ||
| cp apoa1out.$LSB_JOBID ~/ | cp apoa1out.$LSB_JOBID ~/ | ||
| + | |||
| + | |||
| + | </ | ||
| + | |||
| + | |||
| + | ==== gromacs.sub ==== | ||
| + | |||
| + | < | ||
| + | |||
| + | #!/bin/bash | ||
| + | |||
| + | rm -rf gromacs.out gromacs.err \#* *.log | ||
| + | |||
| + | # from greentail we need to recreate module env | ||
| + | export PATH=/ | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | export LD_LIBRARY_PATH=/ | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | |||
| + | #BSUB -o gromacs.out | ||
| + | #BSUB -e gromacs.err | ||
| + | #BSUB -N | ||
| + | #BSUB -J 325monolayer | ||
| + | |||
| + | # read / | ||
| + | . / | ||
| + | export VMDDIR=/ | ||
| + | |||
| + | ## CPU RUN: queue mw256, n<=28, must run on one node (thread_mpi) | ||
| + | ##BSUB -q mw256 | ||
| + | ##BSUB -n 2 | ||
| + | ##BSUB -R " | ||
| + | #export PATH=/ | ||
| + | #. / | ||
| + | #mdrun -nt 2 -s 325topol.tpr -c 325monolayer.gro -e 325ener.edr -o 325traj.trr -x 325traj.xtc | ||
| + | |||
| + | ## GPU RUN: gpu (1-4), queue mwgpu, n (1-4, matches gpu count), must run on one node | ||
| + | ##BSUB -q mwgpu | ||
| + | ##BSUB -n 1 | ||
| + | ##BSUB -R " | ||
| + | ## signal GMXRC is a gpu run with: 1=thread_mpi | ||
| + | #export GMXRC=1 | ||
| + | #export PATH=/ | ||
| + | #. / | ||
| + | # | ||
| + | # | ||
| + | |||
| + | # GPU RUN: gpu (1-4), queue mwgpu, n (1-4, matches gpu count), must run on one node | ||
| + | #BSUB -q mwgpu | ||
| + | #BSUB -n 1 | ||
| + | #BSUB -R " | ||
| + | # signal GMXRC is a gpu run with: 2=mvapich2 | ||
| + | export GMXRC=2 | ||
| + | export PATH=/ | ||
| + | . / | ||
| + | lava.mvapich2.wrapper mdrun_mpi \ | ||
| + | -testverlet -s 325topol.tpr -c 325monolayer.gro -e 325ener.edr -o 325traj.trr -x 325traj.xtc | ||
| + | |||
| + | |||
| + | </ | ||
| + | |||
| + | ==== matlab.sub ==== | ||
| + | |||
| + | < | ||
| + | |||
| + | #!/bin/bash | ||
| + | |||
| + | rm -rf out err *.out | ||
| + | |||
| + | # from greentail we need to recreate module env | ||
| + | export PATH=/ | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | export PATH=/ | ||
| + | export LD_LIBRARY_PATH=/ | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | / | ||
| + | |||
| + | #BSUB -o out | ||
| + | #BSUB -e err | ||
| + | #BSUB -N | ||
| + | #BSUB -J test | ||
| + | |||
| + | # GPU RUN: (1-4), queue mwgpu, n (1-4, matches gpu count), must run on one node | ||
| + | #BSUB -q mwgpu | ||
| + | #BSUB -n 1 | ||
| + | #BSUB -R " | ||
| + | # signal MATGPU is a gpu run | ||
| + | export MATGPU=1 | ||
| + | lava.mvapich2.wrapper matlab -nodisplay | ||
| + | |||
| + | |||
| + | </ | ||
| + | |||
| + | ==== lava.mvampich2.wrapper ==== | ||
| + | |||
| + | < | ||
| + | |||
| + | #!/bin/sh | ||
| + | |||
| + | # This is a copy of lava.openmpi.wrapper which came with lava OCS kit | ||
| + | # Trying to make it work with mvapich2 | ||
| + | # -hmeij 13aug2013 | ||
| + | |||
| + | # | ||
| + | # Copyright (c) 2007 Platform Computing | ||
| + | # | ||
| + | # This script is a wrapper for openmpi mpirun | ||
| + | # it generates the machine file based on the hosts | ||
| + | # given to it by Lava. | ||
| + | # | ||
| + | |||
| + | # RLIMIT_MEMLOCK problem with libibverbs -hmeij | ||
| + | ulimit -l unlimited | ||
| + | |||
| + | |||
| + | usage() { | ||
| + | cat << | ||
| + | USAGE: | ||
| + | This command is a wrapper for mpirun (openmpi). | ||
| + | only be run within Lava using bsub e.g. | ||
| + | bsub -n # "$0 -np # {my mpi command and args}" | ||
| + | |||
| + | The wrapper will automatically generate the | ||
| + | machinefile used by mpirun. | ||
| + | |||
| + | NOTE: The list of hosts cannot exceed 4KBytes. | ||
| + | USEEOF | ||
| + | } | ||
| + | |||
| + | if [ x" | ||
| + | usage | ||
| + | exit -1 | ||
| + | fi | ||
| + | |||
| + | MYARGS=$* | ||
| + | WORKDIR=`dirname ${LSB_JOBFILENAME}` | ||
| + | MACHFILE=${WORKDIR}/ | ||
| + | ARGLIST=${WORKDIR}/ | ||
| + | |||
| + | # Check if mpirun is in the PATH -hmeij | ||
| + | T=`which --skip-alias mpirun_rsh` | ||
| + | #T=`which mpirun_rsh` | ||
| + | if [ $? -ne 0 ]; then | ||
| + | echo " | ||
| + | exit -2 | ||
| + | fi | ||
| + | |||
| + | echo " | ||
| + | #T=`grep -- -machinefile ${ARGLIST} |wc -l` | ||
| + | T=`grep -- -hostfile ${ARGLIST} |wc -l` | ||
| + | if [ $T -gt 0 ]; then | ||
| + | echo " | ||
| + | echo " | ||
| + | exit -3 | ||
| + | fi | ||
| + | |||
| + | # Make the open-mpi machine file | ||
| + | echo " | ||
| + | tr '\/ ' ' | ||
| + | |||
| + | MPIRUN=`which --skip-alias mpirun_rsh` | ||
| + | # | ||
| + | #echo " | ||
| + | |||
| + | # sanity checks number of processes 1-4 | ||
| + | np=`wc -l ${MACHFILE} | awk ' | ||
| + | if [ $np -lt 1 -o $np -gt 4 ]; then | ||
| + | echo " | ||
| + | echo " | ||
| + | exit -4 | ||
| + | fi | ||
| + | |||
| + | # sanity check single node | ||
| + | nh=`cat ${MACHFILE} | sort -u | wc -l` | ||
| + | if [ $nh -ne 1 ]; then | ||
| + | echo " | ||
| + | exit -5 | ||
| + | fi | ||
| + | |||
| + | # one host, one to four gpus | ||
| + | gpunp=`cat ${MACHFILE} | wc -l | awk ' | ||
| + | gpuhost=`cat ${MACHFILE} | sort -u | tr -d ' | ||
| + | gpuid=( $(for i in `ssh $gpuhost gpu-free | sed "s/,/ /g"`; do echo $i; done | shuf | head -$gpunp) ) | ||
| + | if [ $gpunp -eq 1 ]; then | ||
| + | CUDA_VISIBLE_DEVICES=$gpuid | ||
| + | echo "GPU allocation instance $gpuhost: | ||
| + | else | ||
| + | gpuids=`echo ${gpuid[@]} | sed "s/ /,/g"` | ||
| + | CUDA_VISIBLE_DEVICES=" | ||
| + | echo "GPU allocation instance $gpuhost: | ||
| + | fi | ||
| + | # namd ignores this | ||
| + | export CUDA_VISIBLE_DEVICES | ||
| + | #debug# setid=`ssh $gpuhost echo $CUDA_VISIBLE_DEVICES | tr ' | ||
| + | #debug# echo " | ||
| + | |||
| + | |||
| + | if [ -n " | ||
| + | # gromacs needs them from base 0, so gpu 2,3 is string 01 | ||
| + | if [ ${# | ||
| + | gmxrc_gpus=" | ||
| + | elif [ ${# | ||
| + | gmxrc_gpus=" | ||
| + | elif [ ${# | ||
| + | gmxrc_gpus=" | ||
| + | elif [ ${# | ||
| + | gmxrc_gpus=" | ||
| + | fi | ||
| + | |||
| + | if [ $GMXRC -eq 1 ]; then | ||
| + | newargs=`echo ${MYARGS} | sed " | ||
| + | echo " | ||
| + | $newargs | ||
| + | elif [ $GMXRC -eq 2 ]; then | ||
| + | newargs=`echo ${MYARGS} | sed " | ||
| + | echo " | ||
| + | ${MPIRUN} -ssh -hostfile ${MACHFILE} -np $gpunp $newargs | ||
| + | fi | ||
| + | |||
| + | elif [ -n " | ||
| + | echo " | ||
| + | ${MYARGS} | ||
| + | elif [ -n " | ||
| + | cat ${MACHFILE}.lst | tr '\/ ' ' | ||
| + | echo " | ||
| + | charmrun $NAMD_DIR/ | ||
| + | else | ||
| + | echo " | ||
| + | ${MPIRUN} -ssh -hostfile ${MACHFILE} -np $gpunp ${MYARGS} | ||
| + | fi | ||
| + | |||
| + | exit $? | ||
| + | |||
| + | |||
| + | </ | ||
| + | |||
| + | |||
| + | ===== elim code ===== | ||
| + | |||
| + | < | ||
| + | |||
| + | # | ||
| + | |||
| + | while (1) { | ||
| + | |||
| + | $gpu = 0; | ||
| + | $log = ''; | ||
| + | if (-e "/ | ||
| + | $tmp = `/ | ||
| + | @tmp = split(/ | ||
| + | foreach $i (0..$#tmp) { | ||
| + | ($a, | ||
| + | if ( $f == 0 ) { $gpu = $gpu + 1; } | ||
| + | #print "$a $f $gpu\n"; | ||
| + | $log .= " | ||
| + | } | ||
| + | } | ||
| + | # nr_of_args name1 value1 | ||
| + | $string = "1 gpu $gpu"; | ||
| + | |||
| + | $h = `hostname`; chop($h); | ||
| + | $d = `date +%m/ | ||
| + | foreach $i (' | ||
| + | if ( " | ||
| + | `echo " | ||
| + | } | ||
| + | } | ||
| + | |||
| + | # you need the \n to flush -hmeij | ||
| + | # you also need the space before the line feed -hmeij | ||
| + | print " | ||
| + | # or use | ||
| + | # | ||
| + | |||
| + | # smaller than specified in lsf.shared | ||
| + | sleep 10; | ||
| + | |||
| + | } | ||
| + | |||
| </ | </ | ||
cluster/119.1377097412.txt.gz · Last modified: 2013/08/21 11:03 by hmeij
Page Tools
Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Share Alike 4.0 International
