Nebari Plugin - Metrics Server

---

Overview

This plugin integrates Metrics Server into the Nebari platform, allowing container resource metrics for Kubernetes built-in autoscaling pipelines within Nebari. Utilizing Python, Terraform, and Helm, the plugin provides a configurable deployment.

Design and Architecture

The plugin follows a modular design, leveraging Terraform to define the deployment of Metrics Server within a Kubernetes cluster. Key component is: Terraform Configuration: Defines variables and Helm release for deployment.

Installation Instructions

pip install nebari-plugin-metrics-server

Usage Instructions

Configurations: Various configurations are available, including name, namespace, and affinity settings.

Configuration Details

Public

Configuration of the Metrics Server plugin is controlled through the metrics_server section of the nebari-config.yaml for the environment.

metrics_server:
  # helm release name - default metrics-server
  name: metrics-server
  # target namespace - default kube-system
  namespace: kube-system
  # configure default affinity/selector for chart components
  affinity:
    enabled: true # default
    selector: general # default
    # -- or --
    selector:
      default: general
      worker: worker
      db: general
      auth: general
  # helm chart values overrides
  values: {}

Internal

The following configuration values apply to the internally managed terraform module and are indirectly controlled through related values in nebari-config.yaml.

• name: Chart name for Helm release.
• namespace: Kubernetes namespace configuration.
• affinity: Affinity configuration for Helm release.
• overrides: Map for overriding default configurations.

License

nebari-plugin-metrics-server is distributed under the terms of the Apache license.