CRAB3 tutorial

crab3_setup

0) You can get the complete tutorial about CRAB3 
   https://twiki.cern.ch/twiki/bin/view/CMSPublic/WorkBookCRAB3Tutorial (UPDATED TWIKI PAGE)
   For impatient users, a cheat-sheet is available 
   https://twiki.cern.ch/twiki/bin/viewauth/CMS/CRAB3CheatSheet


1) login to lxplus SLC6 machine by doing 
    ssh -Y [email protected]

2) source LCG GRID UI by 
    source /afs/cern.ch/cms/LCG/LCG-2/UI/cms_ui_env.sh

echo $SCRAM_ARCH  --->  slc6_amd64_gcc472

export SCRAM_ARCH=slc6_amd64_gcc481

3) setup the CRAB3 environment by
    source /cvmfs/cms.cern.ch/crab3/crab.sh

4) request the proxy 
    voms-proxy-init -voms cms -valid 192:00

5) check the info about the proxy
    voms-proxy-info -all

6) setup your working dir and install CMSSW and setup CMSSW environmentlike the following
    cmsrel CMSSW_7_0_5
    cd CMSSW_7_0_5/src
    cmsenv 

7) before submitting the jobs to grid, if you are interested to run cmssw code locally, then do it with cmsRun, 
   but please do everything except the CRAB3 setup step...otherwise it'll give an error.

8) In crab2, we usually put all necessary parameters in crab.cfg..but here(in crab3) instead you've to put all the  
   parameters in crabConfig.py(you can change the name accordingly)
   
   crab checkusername
   
   crab checkwrite --site=T2_IN_TIFR

9) submit the crab jobs(you don't have to create the jobs and submit them like in crab2)
   crab submit -c crabConfig.py --dryrun

10) check the status of jobs
    crab status -d crab_projects/crab_tutorial_MC_analysis_test1

    If you want a detail report, then add -l(long) to crab status command

11) resubmit the failed jobs
    crab resubmit -d crab3_projects/crab_tutorial_mc_analysis/
    
12) kill the jobs 
    crab kill -d crab3_projects/crab_tutorial_mc_analysis/