Operations daemon

opsd is the top-level controller for robotic observatory operation.

ops is a commandline utility that controls the operations daemon.

Configuration

Configuration is read from a json file that is installed by default to /etc/opsd. A configuration file is specified when launching the server, and the ops frontend will search this location when launched.

{
  "daemon": "sting_operations", # Run the server as this daemon. Daemon types are registered in `rockit.common.daemons`.
  "log_name": "opsd@sting", # The name to use when writing messages to the observatory log.
  "control_machines": ["STINGTCS"], # Machine names that are allowed to control (rather than just query) state. Machine names are registered in `rockit.common.IP`.
  "pipeline_machines": ["STINGTCS"], # Machine names that are allowed to notify pipelined frame metadata.
  "actions_module": "rockit.operations.actions.sting", # Python module to search for actions for this telescope.
  "loop_delay": 10, # Delay between loop updates.
  "site_latitude": "28.76022 N", # Telescope latitude used for scheduling calculations.
  "site_longitude": "17.87928 W", # Telescope longitude used for scheduling calculations.
  "site_elevation": 2350, # Telescope elevation used for scheduling calculations.
  "dome": {
    "module": "rockit.operations.dome.simulated", # Python module defining the dome interface logic.
    "open_delay": 60, # Module-specific configuration.
    "close_delay": 120
  },
  "environment_daemon": "observatory_environment", # Daemon to query environment state from. Daemon types are registered in `rockit.common.daemons`.
  "environment_conditions": [
    { # Each condition type can contain multiple sensors, pulled from the environment data dictionary.
      "label": "Wind", # Human readable label for this condition type (visible in ops output and web dashboard).
      "sensors": [ # A condition is considered unsafe if ANY sensor returns unsafe or if ALL sensors are unavailable.
        {
          "label": "W1m", # Human readable label for this condition sensor.
          "sensor": "w1m_vaisala", # Sensor name in environmentd data dictionary.
          "parameter": "wind_speed" # Sensor parameter name in environmentd data dictionary.
        }
        # Additional sensors can be defined
      ]
    }
    # Additional conditions can be defined
  }
}

Individual nightly plans are scheduled using ops schedule plan.json where plan.json takes the format

{
  "night": "2021-07-13", # Start of night date for this plan.
  "dome": { # Times to open and close the dome. Can be omitted if opsd is not to control the dome.
    "open": "2021-07-13T21:00:00Z",
    "close": "2021-07-13T21:10:00Z"
  },
  "actions": [
    { # List of actions that are executed in order.
      "type": "Wait", # Investigate the actions python module referenced in the telescope config for a list of actions and their parameters.
      "delay": 120
    }
  ]
}

The dome and telescope should be in automatic mode before trying to schedule a plan.

Initial Installation

The automated packaging scripts will push a collection of RPM packages to the observatory package repository:

Package	Description
rockit-operations-server	Contains the opsd server and systemd service file.
rockit-operations-client	Contains the ops commandline utility for controlling the operations server.
rockit-operations-data-	Contains the telescope-specific configuration.
python3-rockit-operations	Contains the python module with shared code for all telescopes.
python3-rockit-operations-	Contains the python module with telescope-specific code (actions, dome control).

After installing packages, the systemd service should be enabled:

sudo systemctl enable --now opsd@<config>

where config is the name of the json file for the appropriate telescope.

Now open a port in the firewall:

sudo firewall-cmd --zone=public --add-port=<port>/tcp --permanent
sudo firewall-cmd --reload

where port is the port defined in rockit.common.daemons for the daemon specified in the ops config.

Upgrading Installation

New RPM packages are automatically created and pushed to the package repository for each push to the master branch. These can be upgraded locally using the standard system update procedure:

sudo yum clean expire-cache
sudo yum update

The daemon should then be restarted to use the newly installed code:

sudo systemctl restart opsd@<config>

Testing Locally

The ops server and client can be run directly from a git clone:

./opsd sting.json
OPSD_CONFIG_PATH=sting.json ./ops status