Run Control Instructions for shift takers

     

     


    This is a "User Guide" for  KLOE Run Control.
    The main functions of the Run Control are described in the following:
     
    • Start Run Control
    • Configure DAQ
    • Init DAQ
    • Start run
    • Pause and Resume run
    • Stop run
    • Reset DAQ System

    •  
    For a general description of  the programs, see Run Control system description



    Start Run Control programs,  runctl and operator:
     
    • >> log in KRUNC  (KLOE Run Control node) as user daq
    • >> type start_operator
    Now you can use the graphic user interface  to control the DAQ system.



    Configure DAQ processes
    (i.e. start cmdsrv on each DAQ node and send to each cmdsrv Config command, to start all the processes) :
    ConfigDAQ must be done everytime Run type is changed and after every Reset.
     
      >> Select a type in the Run type menu (default is N - Global physics run)

      >> Select ConfigDAQ  in Config&Reset menu

      >> Wait for the message "CONFIGDAQ DONE"!

      If Configdaq  fails :

        try once more
        try to localize the problem from the log messages.
        if panic... Reset All!


    Each run type corresponds to a configuration file in $RUNCTLCONF directory.
    To check the contents of the file, you can open it from File menu: Select "Open Config nodefile" .
    Don't change the file contents unless you are SURE of what you are doing. If you must make any change, write it in the logbook  and tell the following shifter.
     



    Initialize DAQ system:
    INIT must be done after ConfigDAQ and everytime the run ends with errors.
     
      >> Click on INIT button
      >> Wait for the message "INIT DONE!"

      If INIT fails:

        try once more
        then try to localize the error from the log messages.
        Soft Reset can help  if you suspect the processes are messed up.

    Start the run:
     
      If it's a physics run:
        >> Select disk in the Storage menu
        >> Click on DC On to switch DC HV on.
      >> Click on START button
      >> Wait for the message "START DONE!"
              >>  Check that Event # and Trigger # are increasing. If not, expect some error message from l2trigger  or l2collector.



    Pause and Resume the run
       
      While running, you can temporary suspend data taking by clicking on PAUSE button.
      Click on RESUME button to continue.
       

    Stop the run:
     
      >> Click on END button
      >> Wait for the message "END DONE!"
      >> If needed switch DC Standby

      After a successful END you can just make START again when needed.
      If runtype is changed you need a ConfigDAQ.
      If you get errors at END,  try INIT;  if INIT fails Soft Reset.

       


    Reset the system in case of SOFTWARE unrecoverable errors.

    After any Reset, a ConfigDAQ must be done to restart the reset processes.

    Different kinds of Reset are available:
     

      Reset Single Node:
      executes Reset_local_kloe on the selected node.
      cmdsrv and all DAQ programs are killed and the shared memories are deleted.
      Use Reset Single Node when you localize problems (at Config or Init) in a single chain.

      Soft Reset:
      DAQ processes on all nodes in the current configuration are reset, but cmdsrv, and the shared memories are not deleted.
      Soft Reset is fast and it works in most cases (INIT failed in more then one chain, START failed or crash during the run) if some DAQ process is stacked or crashed.
      Always try Soft Reset before Reset All.

      Reset All:
      executes a complete reset on all nodes in the current configuration, deleting the shared memories.
      Use Reset All when ... everything is messed up and you don't know what to try!
      If after a Reset All + ConfigDAQ you still get errors, look for hardware problems (some crate switched off?) or errors in the maps.