Skip to main content
Version: 14.4.0

Configuration

In order to configure the service, a set of environment variables are adopted for describing basic service needs, alongside a main configuration file.

Environment Variables

VariableDescriptionDefault
LOG_LEVELthe maximum log level to emit. Accepted levels are trace|debug|info|warn|errorinfo
HTTP_PORTthe HTTP port on which kubernetes status routes and metrics are exposed3000
CONFIGURATION_FOLDERthe filepath to the folder under which configuration file is located<HOME>/.df/mongezium
OTEL_EXPORTER_OTLP_ENDPOINTspecify the OpenTelemetry OTLP endpoint where traces and metrics should be pushed. When not set, telemetry is not exported
info

Currently <HOME> value is set to /home/mongezium, which is based on how the service image is built.

Configuration File

The application needs a configuration file, named config.json, which respects the following JSON schema configuration.

Loading ....

Starting from version v0.4.3, the Mongezium configuration file supports a connections property. This property allows you to create references to Kafka and MongoDB connections, which can then be reused within the stream and persistence properties, respectively. This feature, already present in other Fast Data services like Stream Processor, Farm Data, and Kango, simplifies the configuration file by eliminating the need to repeat connection details.

warning

Although you can currently define Kafka or MongoDB connections either within the connections property or directly in their respective configuration sections, please be aware that support for the latter is deprecated.

Future versions of Mongezium will exclusively support connection definitions inside the connections property. Therefore, it is highly recommended to adopt this approach as soon as you upgrade to Mongezium v0.4.3 or a later version.

In addition, Kafka configurations and MongoDB persistence properties support secret resolution.

note

The raw JSON schema can also be found here.

Kubernetes

Resources

When the plugin is deployed on Kubernetes, it is advised to set its resources requests and limits. Here are provided which are the recommended ones, although they can be changed according to your needs:

  • requests:
    CPU: 250m 
    Memory: 50MB
  • limits:
    CPU: 1000m 
    Memory: 150MB

Status Probes

The service exposes the liveness and readines status probes as HTTP endpoint, which helps Kubernetes when the service is successfully started and when it may need to be restarted.

The endpoints are:

  • liveness probe: /-/healthz
  • readiness probe: /-/ready