DAQ Interface

  DAQ needs to write to and read from the database. The write process is very similar to the one described above for CMA, except that the command one types to start it is snodb.perl -daq_write and the relevant directory names are $\sim$/write_to_database/daq and $\sim$/write_to_database/daq/todo. The read process for DAQ is constrained by the imperative that the DAQ system must maximize its uptime and should never have to wait for another process to provide it with information.

The read process is started by typing snodb.perl -daq_read, which initiates a process which runs continuously, looking for for new files named *.keys to appear in the directory defined in init.pl. The default is $\sim$/read_from_database/keys_to_dbase. Once a file appears in this directory, it is copied to the $\sim$/read_from_database/keys_to_dbase/todo directory. These ``keys'' files contain information which pinpoints a bank in the database. Here is an example of such a file:

file contents meaning

cst0 bank name

1 bank number

19980527 150705 instant of validity

n (reserved word)

The information in the keys file is read by the $\sim$/$SNODB_VER/exes/sdb_output_titles program which goes to the database, searches for the bank which matches all the keys, and then prints out the bank in titles format onto the file xxxx_n.dat where xxxx is the bank name and n is the bank number. The keys file is renamed to mark it as having been processed and subsequent keys files, if they exist, are then processed.

To stop the DAQ read process, one types snodb.perl -daqr_stop which then creates a file named signal.stop in the appropriate directory and eventually causes the DAQ read process to stop. (Not yet implemented.)