TACC / launcher

A simple utility for executing multiple sequential or multi-threaded applications in a single multi-node batch job
MIT License
63 stars 33 forks source link

Allow launcher to launch multi-core jobs such as openMP jobs. #6

Closed armando-fandango closed 8 years ago

armando-fandango commented 8 years ago

Currently launcher only schedules one job on a core. How can I make launcher launch a job at multiple cores such as openmp jobs.

lwilson commented 8 years ago

You can change the total number of processes per node by setting LAUNCHER_PPN. So if you have a 24-core node and want to run 3 OpenMP applications with 8 threads each, just do the following:

export OMP_NUM_THREADS=8 export LAUNCHER_PPN=3 $LAUNCHER_DIR/paramrun

Also, if you explicitly export a value for LAUNCHER_PPN, it will honor your settings even if you have scheduler integration enabled.

zhenzuo2 commented 5 years ago

I tried to add

export LAUNCHER_PPN=1

But I checked the log it still runs multiple jobs together. Can I limit the number of jobs running at the same time?

zhenzuo2 commented 5 years ago

!/bin/bash

#

Simple SLURM script for submitting multiple serial

jobs (e.g. parametric studies) using a script wrapper

to launch the jobs.

#

To use, build the launcher executable and your

serial application(s) and place them in your WORKDIR

directory. Then, edit the CONTROL_FILE to specify

each executable per process.

-------------------------------------------------------

-------------------------------------------------------

#

<------ Setup Parameters ------>

#

SBATCH -J Parametric

SBATCH -n 96

SBATCH -N 2

SBATCH -p hadoop

SBATCH -o Parametric.o%j

SBATCH -t 48:00:00

<------ Account String ----->

<--- (Use this ONLY if you have MULTIPLE accounts) --->

SBATCH -A

------------------------------------------------------

module load launcher export LAUNCHER_RMI=SLURM export LAUNCHER_JOB_FILE=/work/05978/zzuo/mysharedirectory/eqtl/code/job_chr_1.sh export LAUNCHER_WORKDIR=/work/05978/zzuo/mysharedirectory/eqtl/processed/ export LAUNCHER_SCHED=interleaved export LAUNCHER_PPN=1 $LAUNCHER_DIR/paramrun

/opt/apps/launcher/launcher-3.1/paramrun: line 171: [: -eq: unary operator expected Launcher: Setup complete.

------------- SUMMARY --------------- Number of hosts: 2 Working directory: /work/05978/zzuo/mysharedirectory/eqtl/processed/ Processes per host: 48 Total processes: 96 Total jobs: 50 Scheduling method: interleaved /opt/apps/launcher/launcher-3.1/paramrun: line 211: [: -eq: unary operator expected


Launcher: Starting parallel tasks... Launcher: Task 1 running job 2 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WGS-$Launcher: Task 7 running job 8 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WGS-$Launcher: Task 6 running job 7 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WGS-$Launcher: Task 5 running job 6 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WGS-$Launcher: Task 2 running job 3 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WGS-$Launcher: Task 19 running job 20 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WG$Launcher: Task 13 running job 14 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WG$Launcher: Task 8 running job 9 on c252-110.wrangler.tacc.utexas.edu (module load biocontainers; module load fastqtl/ctr-v2.184dfsg-5-deb_cv1; fastQTL -V merge2.WGS-$L