This description refers to 2011 setup, before moving L2 to HLT
The real L2 algo read all peds, status tables from the l2ana01, user=developer, dir=emc_setup (or sth similar) Only the files with the word 'current' in the name are read in. Those are symbolic links pointing to the real files. All other ped, status files, with some date attached are kept only as a record.
Those are all setup files needed by L2 algos: [developer@l2ana01 emc_setup]$ ls *current btowCrateMask.current etowCrateMask.current pedestal.current btowDb.current etowDb.current towerMask.current
Common format: all lines starting with '#' are ignorred as comments. Every file has short description of its contntent. One need to be familiar with EEMC DB to understand the details.
btowCrateMask.current, etowCrateMask.current, btowCrateMask.current etowCrateMask.current are used to mask out whole crates or individual towers. They are editted by hand as needed during the run. Typically all us unmasked at the beggining of each run.
pedestal.current is produced by L2-ped algo under different name (run12058037.l2ped.log) after every run.Once person monitoring L2 pedestal residua decideds pedestals used by L2 drifted too far, the pedestal.current is linked to the fresh runXXX.l2ped.log. Typically twice per week.
btowDb.current, etowDb.current contain full DB mapping for both EMC. Those files are produced once per year, by running $STAR/StRoot/StEEmcDbMaker/StEmcAsciiDbMaker.cxx on the offline STAR DB.If no mapping was changed in the offline DB there is no need to regenerate those 2 files.