Redefinition of Physical and Logical detectors

[valerioda]

The dispatcher get goal_state from MongoConnect.get_wanted_state where there is a loop on physical detectors and the latest_status from MongoConnecy.get_update using the current configuration that is coming from MongoConnect.get_super_detector. Then it loops among the latest_status.keys and checks the goal status, the variable detector is used for both physical and logical detectors, the main goal of the performed changes is to avoid this and have a more clear and separate definition of the two chategories of detectors.

Main changes are listed:

• in MongoConnect.py the logical detector is created with the function get_logical_detector (former get_super_detector) with five possible linking cases that are addressed separately with logical statements, in this way is possible to have proper name for the logical detector (e.g., in case of all physical detectors linked is all_linked instead of tpc). The five cases:

  1. all_linked
  2. tpc_mv + nv
  3. tpc_nv + mv
  4. tpc + mv_nv
  5. tpc + mv + nv

• in dispatcher.py added new loop: 1st loop on logical detectors using latest_status.keys, 2nd loop on physical detectors using latest_status[log_det]['detectors'].keys, the printed status then reports the aggregate status for both logical and physical detectors

• in DAQController.py the input detector explicity called logical, in particular in control_detector and check_timeouts; the first physical detector is used as representative of the logical detector, needed to take goal mode and status and count the missed arm cycles and error stops (also functions handle_timeout, stop_detector_gently and check_run_turnover has been changed in order to use both logical and physical detectors); also in MongoConnect.py modified insert_run_doc and and set_stop_time

• in hypervisor.py changes in linked_nuclear_option: modified loop to check for the status of physical detectors, added debug messages, added check of all hosts (readers and controllers) to see if they are responding, calculated again the aggregate status of the physical detectors and compared with the one from MongoConnect.aggregate_status, if the two status are not matching or the detector is in [ERROR, TIMETOUT, UNKNOWN] the detector is included in timeout_list otherwise in running_list

[mflierm]

I think it's going to complain in this line about detector. Should it be logical?