Passing configurations to a Kelvin App

In addition to receiving data, configurations are often required by the application to correctly operate. In this section, configuration injection through the application configuration file will be explained.

Defining configurations

The application configuration section is defined by a dictionary and can take any values. The following highlighed lines contain a configuration exemple with complex values.

YAML

app:
  kelvin:
    inputs:
      - data_type: raw.float32
        name: random_value
        sources:
          - asset_names: [ ]
            workload_names: [ ]
        configuration:
          enabled: true
          inside: 0
          outside: 45
          thresholds:
            min: 20
            max: 30
        language:
          python:
            entry_point: my_app.my_app:App
            requirements: requirements.txt
          type: python
        logging_level: INFO
        version: 4.0.0
    system_packages: [ ]
  type: kelvin
info:
  description: my-app
  name: my-app
  title: my-app
  version: 1.0.0
spec_version: 1.0.0

Handling the configurations in the application code

Once specified, configurations will be available in the code. Configurations are accessible in code through Kelvin App's self.config variable. This instance variable holds the configuration object for quick access.

Python

from kelvin.app import DataApplication
from kelvin.message.raw import Float32

class App(DataApplication):
    """Application."""

    def process(self):
        random_value = self.data.get('random_value')
        min_threshold = self.config.thresholds.min
        max_threshold = self.config.thresholds.max
        if random_value and min_threshold <= random_value <= max_threshold:
            print("The provided random value respects both thresholds")