Skip to content

A simple HTCondor setup for running `ipole` on the Open Science Grid

Notifications You must be signed in to change notification settings

avjoshi21/ipole-osg

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

70 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ipole-osg

A simple setup for running a large number of ipole jobs on the Open Science Grid (OSG) using HTCondor.

Usage

This git repository is designed to be cloned to OSG connect as a run directory for both submitting a large OSG job and gathering the output. To start,

ssh osg
cd /public/<user>/
git clone git@github.com:bhpire/ipole-osg.git run-01
cd run-01

All the scripts are placed inside bin/. The empty directories dat/, log/, and out/ will be used to stage GRMHD data (as input to ipole), OSG run logs, and ipole output.

Before submitting a job, one needs to copy a static linked ipole binary to bin/ and populate dat/ with GRMHD data.

cp ~/src/ipole/ipole bin
scp bh:sim-lib/sample/dump_*.h5 dat

Edit to bin/pargen may be made to generate the necessary parameter sets. Then, simply submit an OSG job by

bin/submit

bin/submit is a standard Condor submission script starts with a hashbang directive to use the system condor_submit. It uses bin/pargen to generate a list of parameter sets based on all the GRMHD input data in dat/ and the parameter listed inside bin/pargen. These parameter sets are then passed to bin/wrapper on the worker machines as command line arguments. bin/wrapper will automatically generate a ipole parameter file based on the arguments, and start ipole with it.

All Condor logs, stdout, and stderr will be sent to log/. The parameter files and spectrum output will be saved to out/. Their file names are transform according to the rules described in bin/submit.

About

A simple HTCondor setup for running `ipole` on the Open Science Grid

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Shell 69.9%
  • Python 30.1%