Run Jobs on Grid

If you have already built your analysis package and tested it locally, you can also submit the jobs to the grid and specify BNL/SLAC destination in prun option –destSE.

Setup of Panda Env

First you need set up the panda env:

setupATLAS     # if this command is already defined
lsetup panda   # Panda env setup

Job Submission to the Grid

Then submit the jobs using the command prun

asetupTag=21.2.81,AnalysisBase
dset=data18_13TeV.00348885.physics_Main.deriv.DAOD_EXOT12.f937_m1972_p3553_tid14278917_00
prun --exec "bin/Exam_JetsPlot %IN" --bexec "build-it.sh" \
     --athenaTag=$asetupTag --cmtConfig $CMTCONFIG \         # specify the env needed for the job
     --inDS $dset --nFiles 2 --nFilesPerJob 1 \                                # input
     --outDS user.yesw.Tutorial-BNLT3.test5 --destSE BNL-OSG2_SCRATCHDISK   # output

Please change the above output dataset name and/ output RSE if needed.

Where the script build-it.sh is used to rebuild the analysis package on the grid machines. The script read:

rm -f CMakeCache.txt
cmake src
make

For prun usage, please run “prun -h” or visit the wiki page

Job Status

You can check the job status via command pbook or visit the panda job webpage.