Pod watcher

Kubernetes has a metrics server that provides CPU and memory usage statistics for pods. By default, these metrics are collected every 60 seconds. This interval can be adjusted using the --metric-resolution flag. However, it is not recommended to set values below 15 seconds, as this is the resolution of metrics calculated by Kubelet (reference1, reference2).

I created this tool to monitor CPU and memory usage and log the peak usage of pods.

Demo

You can view the results similar to kubectl top pod.

NAME     CPU(cores)     MEMORY(bytes)
wren     13m            1450Mi

How to install

• Requirement node 17
• Clone repo
• Install dependencies

  npm install

How to use

npm start

If you want to use this anywhere, you can install it globally.

npm install -g

You can run pod-watcher directly in the terminal from anywhere to start this tool.