In the Mars_McMismatchStudy branch I have created a set of scripts for sending jobs to the ISDC batch systemi (see below for full list).
All of the scripts follow the same style. They are comprised of two files.

FACT.sub_PROGRAM_TYPE.sh which is exicuatable and the one you run.
FACT.qsub_PROGRAM_TYPE.sh which is the template for the script that runs on the batch machines.

PROGRAM = star, calisto, ...
TYPE = data or mc

to find the options on how to run any of the scripts run it with no options:


------
> ./FACT.sub_star_data.sh

FACT.sub_star_data.sh <runlist>

<runlist> should be in the form YEAR MONTH DAY eg:
2013 10 15
------

An example run list for star is provided in the form of: 
Crab.star

The sub script crates a job in the queue directory for each night on the list.
Each job then runs over the sequence files for each day.
To find the command and how it is run take a look inside the qsub file. 
For example star is run, in the script, as follows:

  root -b -q fact/analysis/star.C\(\"$SEQFILE\",4,2.5,\"$INDIR\",\"$OUTDIR\"\)

The output all goes to a directory called something like:
  OUTDIR=$MARSSYS/OUTPUT/star/$DATE
which be created if it does not exist.

Things to do:

The MC chain has not been tested.
The files created by callisto are too large to do a large data set. What might be more
useful for full production in to combine the callisto and star scripts. The callisto 
output can be created and removed on the fly. 

Give it a try and feel free to change/edit what doesn't work. 


Full list:

Crab.star
Crab.ganymed

FACT.qsub_call_mc.sh
FACT.qsub_ceres_mc.sh
FACT.qsub_gan_data.sh
FACT.qsub_gan_mc.sh
FACT.qsub_ped_mc.sh
FACT.qsub_star_data.sh
FACT.qsub_star_mc.sh
FACT.sub_call_mc.sh
FACT.sub_ceres_mc.sh
FACT.sub_gan_data.sh
FACT.sub_gan_mc.sh
FACT.sub_ped_mc.sh
FACT.sub_star_data.sh
FACT.sub_star_mc.sh