Panda
Installing Panda
source ~/cmthome/setup.sh -tag=14.4.0,32,setup
cd $TestArea
cmt co PhysicsAnalysis/DistributedAnalysis/PandaTools
cd PhysicsAnalysis/DistributedAnalysis/PandaTools/cmt
cmt config
source setup.sh
make
cd $TestArea
mkdir run
cd run
if [ $SHELL == "/bin/zsh" ];
then
rehash
fi
cd $TestArea/run
Running with Panda
pathena AnalysisSkeleton_topOptions.py \
--inDS fdr08_run2.0052283.physics_Muon.merge.AOD.o3_f8_m10 \
--outDS user08.myGridName.AnalysisSkeleton.aan.root
Monitoring
There are several other things you might want to do, however, in this rich set of tools. We'll start with monitoring -- let's see how the job we just submitted is doing. From the command line, enter:
pathena_util
and you will be put in a python-like command line. When you ask it for
help()
, it helps you. When you ask it to
show()
you your jobs, it will do so. If you want to see only a specific job number, include that number in the parentheses.
You can also get the
status()
of any particular job that has been submitted from this location, and
select()
jobs based on some criteria, as follows:
select('outDS=user08.AldenStradling.*')
You can even, if all does
not go well,
kill()
any particular job.
Reusing the environment
You have noticed that all of your jobs create a build job before processing any data. This is great, because it replicates your working environment in detail. If you'd like to
reuse this environment, you can avoid running that build job over and over again. As long as you don't
recompile anything in your environment, it remains identical -- you can swap things like
jobOptions
, datasets, and etc. Your jobs will ALWAYS be sent to the same site, as well -- the jobs are always sent to the data -- in this case, to your stored build environment and the data.
To do this, run subsequent jobs with the option
--libDS SAME
Issues
--
OldrichKepka - 28 Dec 2008