DAMN are Data Analysis Mask Numbers banks. These hold the collaboration-wide masks. The data is divided into sub-run windows, with the bank number as the run number for the first mask (mask number 0). If all 32 bits of the bit mask are used up then further masks starting with mask number 1 will be added. These will have bank numbers i*10**8 + r where i is the mask number and r the run number. The sub-run number is stored in the database header ``Task Type'' field.
These banks will be hold a very large volume of data as there will be a mask for every trigger. It estimated to be3Gbytes/year.
The banks are held in their own database and updates distributed along with other database updates. Should it be necessary to reissue the banks then the plan would be to recreated the database files from scratch and ship out completed ones to replace the previous versions. So we give up the ability to go back to a previous version as the price for keeping the database files as small as possible. The banks will have version numbers to ensure that the distribution system is working correctly.
Originally DAMN banks were organised by run, rather than by sub-run so the system supports two types of DAMN bank: single sub-run and multiple sub-run. When creating new banks the single sub-run form will be created where possible. For run numbers less than 10614 the sub-run number was not coded in the ZDAB bank and so DAMN banks have to be multiple sub-run for these early runs. When updating exsisting banks, the format is never changed.