Old grid page
paw to ROOT for crusties like me.

General athena setup

Computing work book
Physics analysis work book
My "testarea" is /usera/lester/testarea

My grid/atlas related scripts are in ~/scripts (not proj/scripts as everything lese)

From WorkBookSetAccount : 

mkdir -p ~/testarea/17.5.0
cd ~/testarea/17.5.0

source ~/scripts/atlas_setup.sh 17.5.0
asetup
Now athena is in the path, and thousands of evnironment vars are set up pointintg to directories (including in my test area) with 17.5.0 in them,etc.

Panda

First general setup of Panda. (NB, not Gang Fu Panda.) 
mkdir -p ~/testarea/17.5.0
cd ~/testarea/17.5.0
source ~/scripts/atlas_setup.sh 17.5.0
asetup

# must to asetup before panda_setup.sh!

source ~/scripts/panda_setup.sh
Now "prun" and "pathena" are in your path.

prun

This is for funning arbitrary scrips on the grid. You write the script locally (eg fish.sh in the second example below) and it (along with any other stray small files) are automatically tarred up and sent to the grid node when you submit the job. On the worker node you work entirely with local files, but these get renamed at the end of your job to have extra attributes before being put in the dataset. You don't have to worry about the funny names while the job is running. The files that you list in "outputs" get added to the output dataset. 
source ~/scripts/panda_setup.sh
mkdir ~/testarea/moo13
cd  ~/testarea/moo13
prun --noBuild --exec "echo Hello Dr Lester > lester.txt" --outDS=user.kesterlester.moo13 --express --outputs=lester.txt

# go to pandamon -- wait

dq2-get user.kesterlester.moo13/
or 
source ~/scripts/panda_setup.sh
mkdir ~/testarea/moo14
cd ~/testarea/moo14
cat << EOF > fish.sh
#!/bin/sh
echo Shark > ray.txt
EOF
chmod +x fish.sh
prun --noBuild --exec "./fish.sh" --outDS=user.kesterlester.moo14 --express --outputs=ray.txt

# go to pandamon -- wait

dq2-get user.kesterlester.moo14/
To see how things are going use pandamon:

Sending output histograms to the right file

This job option snippet: 
if not ('outFile' in vars() or 'outFile' in globals()):
  outFile = 'parity_hists.root'
from GaudiSvc.GaudiSvcConf import THistSvc
#THistSvc.OutputLevel = DEBUG
ServiceMgr += THistSvc("THistSvc")
ServiceMgr.THistSvc.Output = ["hists DATAFILE='"+outFile+"' OPT='RECREATE'"]
asks the THistSvc to put our output histograms (which on creation are automatically owned by THistSvc) into the file whose name is contained in outFile. To use THistSvc it is necessary to create a histogram like this: 
#include "GaudiKernel/ITHistSvc.h"
#include "TH1D.h"

class MyAlg:public AthAlgorithm {
... add data member to ananlysis class :

    // Histograms and Trees
    ITHistSvc *m_thistSvc;
...
}

StatusCode MyAlg::initialize(){
...
  // Set up the histogram service
  if (service("THistSvc", m_thistSvc).isFailure()) {
    msg() << MSG::ERROR << "Couldn't get handle to THistSvc" << endreq;
    return StatusCode::FAILURE;
  }
...
  TH1D *p = new TH1D("lesterHist1", "lesterHist1", 11, -0.5, 10.5);
  if (m_thistSvc->regHist("/hists/lesterHist1", p).isFailure()) {
    msg() << MSG::ERROR << "Couldn't register test histogram" << endreq;
    return StatusCode::FAILURE;
  }

pathena (for athena jobs)

NB - this boomerang trick doesn't work: 
mkdir -p ~/testarea/moo15
cd ~/testarea/moo15
source ~/scripts/atlas_setup.sh 17.5.0
asetup 
get_files -jo HelloWorldOptions.py
mv {Hello,Lester}WorldOptions.py
sed -i -e "s/Framework/Dr Lester/g" LesterWorldOptions.py

# NB - pathena tried to figure out what our output files are ... but helloworld doesn't write root files, so here is workaround:
echo Boing > boomerang.txt

source ~/scripts/panda_setup.sh 
# if I would do 
#    athena.py LesterWorldOptions.py 
# then :

pathena LesterWorldOptions.py --outDS user.kesterlester.moo15 --express --noBuild  --extOutFile "boomerang.txt" --extFile boomerang.txt
# --extOutFile "boomerang.txt" means "this is output"
# --extFile "boomerang.txt" means "send this file as input"

ROOT test

root -l /usera/lester/testarea/17.5.0/run/DATA12/./data12_8TeV.00200804.physics_Egamma.merge.NTUP_SUSY.f431_m1111_p981_tid779030_00/NTUP_SUSY.779030._000001.root.1
susy->Draw("EF_2e12Tvh_loose1")
TH1D * hist = new TH1D("lesterHist","lesterHist",10, 2.5, 4.5)
susy->Draw("EF_2e12Tvh_loose1+3 >> lesterHist")
or make lester.C like this 
#include <TFile.h>
#include <TTree.h>

#define deltaPhi(container,i,j) "asin(sin(" #container "[" #i "])*cos(" #container "[" #j "])-cos(" #container "[" #i "])*sin(" #container "[" #j "]))"

void lester() {

  TFile *f = TFile::Open("/usera/lester/testarea/17.5.0/run/DATA12/./data12_8TeV.00200804.physics_Egamma.merge.NTUP_SUSY.f431_m1111_p981_tid779030_00/NTUP_SUSY.779030._000001.root.1","READ");

  TTree * t = static_cast<TTree *>(f->Get("susy"));


  t->Draw( deltaPhi(el_phi,0,1)  ":el_eta[1]-el_eta[0]","el_n>=2","COLZ");

  //t->Draw("el_phi[0]:el_eta[0]","el_n>=2","COLZ");
  //t->Draw("el_phi[1]:el_eta[1]","el_n>=2","COLZ");

  f->Close();
}
and run in root

Athena Hello World

Hello world WorkBookRunAthenaHelloWorld.

Just to run it: 

get_files -jo HelloWorldOptions.py ;
athena.py HelloWorldOptions.py
To run it and make changes:
cmt co -r UserAnalysis-00-15-08 PhysicsAnalysis/AnalysisCommon/UserAnalysis    #(get the tag from the atlas tag collector)
make -C PhysicsAnalysis/AnalysisCommon/UserAnalysis/cmt
( cd PhysicsAnalysis/AnalysisCommon/UserAnalysis/run ; 
get_files -jo HelloWorldOptions.py ;
athena.py HelloWorldOptions.py
)
(get the tag from the atlas tag collector)

Athena new package

Check out 
svn://pcfs.hep.phy.cam.ac.uk/TRUNK/EmptyPackage
and run the script that modifies its name.

Unix command to grid

source ~/scripts/grid_setup.sh
#voms-proxy-init --voms atlas

#prepare job
mkdir -p /tmp/moo
cd /tmp/moo
cat << EOF > helloworld.jdl
#############Hello World#################
Executable = "/bin/echo";
Arguments = "Hello welcome to the Grid ";
StdOutput = "hello.out";
StdError = "hello.err";
OutputSandbox = {"hello.out","hello.err"};
VirtualOrganisation = "atlas";
#########################################
EOF

#see where it might run:
glite-wms-job-list-match -a helloworld.jdl 

#submit
glite-wms-job-submit -a -o jobIDfile helloworld.jdl

#check status
glite-wms-job-status -i jobIDfile
glite-wms-job-status -i jobIDfile
glite-wms-job-status -i jobIDfile

#cancel if necessary
#glite-wms-job-cancel -i jobIDfile 

#get output -- put it in dir "jo"
glite-wms-job-output -i  jobIDfile --dir jo
Athena on the grid 
source ~/scripts/grid_setup.sh
#voms-proxy-init --voms atlas

#prepare job
cd
rm -rf /tmp/moo
mkdir -p /tmp/moo
cd /tmp/moo

cat << EOF > hello.sh
#!/bin/bash
############# hello.sh ###############
# Script to run AthenaHelloWorld on the Grid

# Use this for releases before 16.0.0
#source $VO_ATLAS_SW_DIR/software/17.0.6/setup.sh
#source $SITEROOT/AtlasOffline/17.0.6/AtlasOfflineRunTime/cmt/setup.sh

# Use this for release 16.0.0 and above
RELEASE=17.0.6
source $VO_ATLAS_SW_DIR/software/$RELEASE/cmtsite/asetup.sh AtlasOffline $RELEASE

athena.py HelloWorldOptions.py
######################################
EOF

cat << EOF > helloworld.jdl
############# Athena #################
Executable = "hello.sh";
StdOutput = "hello.out";
StdError = "hello.err";
InputSandbox = {"hello.sh","LesterWorldOptions.py"}; 
OutputSandbox = {"hello.out","hello.err", "CLIDDBout.txt"};
Requirements = Member("VO-atlas-offline-17.0.6-i686-slc4-gcc34-opt", other.GlueHostApplicationSoftwareRunTimeEnvironment);
######################################
EOF
... see requirements specifications ... 
get_files -jo HelloWorldOptions.py
mv {Hello,Lester}WorldOptions.py
sed -i -e "s/Framework/Dr Lester/g" LesterWorldOptions.py

#see where it might run:
glite-wms-job-list-match -a helloworld.jdl 

#submit
glite-wms-job-submit -a -o jobIDfile helloworld.jdl

#check status
glite-wms-job-status -i jobIDfile
glite-wms-job-status -i jobIDfile
glite-wms-job-status -i jobIDfile

#cancel if necessary
#glite-wms-job-cancel -i jobIDfile 

#get output -- put it in dir "jo"
glite-wms-job-output -i  jobIDfile --dir jo
Relocatable rpm installation 
create file ~/.rpmmacros and add following line
       %_rpmlock_path lib/rpm/__db.000

Initialize the database by running command:

  rpm --initdb \
    --root $HOME/rpm-local/ \
    --dbpath $HOME/rpm-local/lib/rpm

Check the dependence

  rpm --root $HOME/rpm-local/ \
    --dbpath $HOME/rpm-local/lib/rpm \
    -ivh package.rpm

Install package

  rpm --root $HOME/rpm-local \
    --dbpath $HOME/rpm-local/lib/rpm \ 
    --relocate /usr=$HOME/rpm-local \
    --nodeps \
    -ivh package.rpm

Good Runs Lists GoodRunsLists

The GRL generator says it is only for experts, and that mortals should look here, not only for getting them, but for lots of info on how to use them and tools and combine them.
In particular, the luminosity associated with a GRL can be calculated with this tool. I wonder if 
iLumiCalc.exe
as mentioned here is ever useful.

Nice selection example

here