Skip to content

README.md

Latest commit

 

History

History
242 lines (154 loc) · 9.26 KB

README.md

File metadata and controls

242 lines (154 loc) · 9.26 KB

Working with Kubernetes Nodes in Minikube

Overview

This guide provides a beginner-friendly introduction to Kubernetes nodes and how to manage them using a local Minikube cluster. You will learn how to start, stop, and inspect the node that powers your local Kubernetes environment, as well as understand key concepts like labels and taints.

Table of Contents

Project Goals

By the end of this project, you will be able to:

  • Understand the role of a node in a Kubernetes cluster.
  • Use kubectl to inspect node details, including labels, capacity, and taints.
  • Add custom labels to a node for better organization.
  • Confidently perform basic lifecycle operations on your Minikube cluster.

Prerequisites

  • Minikube Installed: You must have a working Minikube installation.
  • Docker Running: The Docker service must be active on your machine.

Core Concepts

What is a Kubernetes Node?

In Kubernetes, a node is a worker machine where your containers are deployed. It can be a virtual or physical machine. Each node is managed by the control plane and contains the services necessary to run Pods, which are the smallest deployable units in Kubernetes. In our Minikube setup, the entire cluster runs on a single node.

Managing the Minikube Cluster and Node

Step 1: Start the Minikube Cluster

If your cluster isn't running, open your terminal and start it:

minikube start

Output of minikube start

Minikube Start

Step 2: View Node Information

Use kubectl to view the nodes in your cluster. You will see a single node for your Minikube setup.

kubectl get nodes

This command lists all nodes, their status (e.g., Ready), roles, age, and Kubernetes version.

Output of kubectl get nodes

kubectl get nodes

Step 3: Inspect a Node in Detail

To get a comprehensive overview of your node, use the describe command. This is crucial for understanding the node's properties.

kubectl describe node minikube

(Replace minikube if your node has a different name).

This command provides detailed information, which we will explore in the next section.

Output of kubectl describe node

kubectl describe node

Understanding Node Details

The describe node command output contains valuable information. Let's break down a few key parts.

Node Labels

Labels are key/value pairs that are attached to Kubernetes objects, such as nodes. They are used to organize and to select subsets of objects. For example, a node could be labeled with disktype=ssd or environment=production. The Kubernetes scheduler can use these labels to decide where to place pods.

In the describe node output, look for the Labels section. You will see several default labels.

Node Resources (Capacity and Allocatable)

This section shows the total resources available on the node (Capacity) and the resources available for pods to consume (Allocatable). The difference between capacity and allocatable is reserved for the operating system and Kubernetes system daemons.

  • Capacity: The total amount of CPU, memory, and storage on the node.
  • Allocatable: The amount of resources that are available for pods.

Understanding this is key to managing application performance and resource quotas.

Taints and Tolerations

  • A Taint is applied to a node to repel a set of pods.
  • A Toleration is applied to a pod to allow it to be scheduled on a node with a matching taint.

This mechanism ensures that pods are not scheduled onto inappropriate nodes. For example, a node with a dedicated GPU might be tainted to only allow pods that require a GPU. In the describe node output, you can see if any Taints are applied.

Working with Node Labels

Labels are a fundamental part of Kubernetes. Let's practice by adding one.

Step 4: Add a Custom Label to a Node

You can add your own labels to a node to categorize it. Let's add a label to identify this node as being for development purposes.

kubectl label nodes minikube environment=development

To verify that the label was added, you can view the labels for the node:

kubectl get nodes --show-labels

You should see your new environment=development label in the list.

Output of kubectl get nodes --show-labels

Node with Custom Label

Node Scaling and Upgrades in Minikube

Understanding Minikube's Single-Node Architecture

By default, Minikube creates a single-node cluster. This is intentional, as Minikube is designed for local development and testing, not for production workloads. This single-node setup means you cannot scale your cluster by adding more nodes as you would in a production environment (e.g., with kubeadm or a cloud provider). The primary goal is to have a lightweight, portable Kubernetes environment that runs on your local machine.

While you can't add more nodes to a single Minikube cluster, you can simulate a multi-node environment by creating multiple Minikube clusters (profiles), as described in the "Managing Multiple Clusters with Profiles" section.

Upgrading Kubernetes Version

Minikube makes it easy to test different versions of Kubernetes. You can create a new cluster with a specific Kubernetes version or upgrade your existing cluster.

To start a cluster with a specific Kubernetes version:

minikube start --kubernetes-version=v1.25.3

You can find available Kubernetes versions on the official Kubernetes release page.

To upgrade the Kubernetes version of an existing Minikube cluster:

You can use the minikube start command again with a newer --kubernetes-version. Minikube will handle the upgrade process. For example, if your cluster is running v1.25.3 and you want to upgrade to v1.26.0, you would run:

minikube start --kubernetes-version=v1.26.0

This approach allows you to keep your development environment in sync with your production environment's Kubernetes version.

Cluster Cleanup

After you are done, you can stop or delete your cluster.

  • Stopping the cluster preserves its state: 
    minikube stop
  • Deleting the cluster removes it completely: 
    minikube delete

Managing Multiple Clusters with Profiles

Yes, you can run more than one Minikube cluster on the same machine. Minikube uses profiles to manage different clusters. By default, you use a profile named minikube, but you can create and manage others.

Creating a New Cluster Profile

To create a new cluster, use the -p (or --profile) flag with the start command.

# Create a new cluster named 'test-cluster'
minikube start -p test-cluster

This will create a completely separate cluster. Your kubectl context will automatically be switched to this new cluster.

Interacting with a Specific Profile

To run any command against a specific cluster, use the -p flag.

# Stop the new cluster
minikube stop -p test-cluster

# Delete the new cluster
minikube delete -p test-cluster

Switching Between Profiles

You can list all your profiles and switch your kubectl context between them.

# List all profiles
minikube profile list

# Switch your kubectl context to the default minikube cluster
minikube profile minikube

This feature is incredibly useful for testing different Kubernetes versions or configurations simultaneously.

Troubleshooting

  • kubectl command not found:

    • Solution: Ensure kubectl is installed and in your system's PATH, or use minikube kubectl -- <command>.

  • Node is not in Ready state:

    • Solution: Run minikube status and minikube logs to diagnose the issue. The problem is often related to resource constraints or Docker issues.

  • Error connecting to the server:

    • Solution: Your cluster may not be running. Use minikube start to activate it.

Evidence for Submission

  • Figure 1: Output of minikube start
  • Figure 2: Output of kubectl get nodes
  • Figure 3: Output of kubectl describe node
  • Figure 4: Output of kubectl get nodes --show-labels