Metrics Server Deployment Guide

Overview

metrics-server is a lightweight and scalable resource usage metrics aggregator for Kubernetes. In the Crater platform, it provides real-time CPU and memory usage data for:

• Pod and Node resource monitoring
• Job scheduling decisions
• Frontend visualization (e.g., dashboards, usage stats)

Metrics Server is required by Kubernetes components such as kubectl top, Horizontal Pod Autoscalers (HPA), and Crater's job management UI.

---

Dependencies

The metrics server must be deployed after the Kubernetes cluster is fully operational and can communicate with node components via kubelet.

No external dependencies are required, but make sure:

• The kubelet --read-only-port or --authentication-token-webhook settings are enabled.
• Your Kubernetes version is compatible with the chosen metrics-server version.

---

Crater-Specific Notes

• Crater uses metrics-server as part of its real-time resource display on the UI.
• Job submission and resource prediction modules may also read these values to improve scheduling accuracy.
• The metrics are complementary to Prometheus but provide faster sampling for short-lived pods.

---

Installation

We recommend using the official Helm chart with Crater's values.

📦 Helm values: deployments/metrics-server/values.yaml
📖 Detailed guide: deployments/metrics-server/README.md

Edit on GitHub