Minikube and kubectl

Understanding Minikube and Kubernetes Cluster Architecture

• Kubernetes Cluster in Production Environments

  • Typically involves multiple nodes:

    • At least two master nodes

    • Multiple worker nodes

  • Each node (virtual or physical) runs independently and fulfills specific responsibilities

• Challenges with Local Testing

  • Full production cluster setup is resource-intensive:

    • Requires significant CPU, memory, and network resources

    • Infeasible for quick, local testing

• Minikube – Local Kubernetes Solution

  • A one-node Kubernetes cluster suitable for local development

    • Combines master and worker processes on a single node

    • Utilizes a virtual machine (VM) via hypervisor (e.g., VirtualBox, HyperKit)

  • Comes with:

    • Docker runtime pre-installed to run containers

  • Runs within a virtualized environment (e.g., VirtualBox, HyperKit)

  • Designed for:

    • Local development

    • Quick testing of applications or components

  Example:

  • Instead of deploying a complex multi-node setup, a developer can use Minikube to simulate a Kubernetes cluster for testing on a laptop.

---

kubectl – The Kubernetes Command-Line Tool

• Purpose and Functionality

  • CLI tool used to interact with Kubernetes clusters

    • Supports actions like creating pods, services, and deleting resources

  • Communicates with Kubernetes API server (part of the master processes)

• API Server – Central Entry Point

  • Accepts requests from:

    • kubectl (CLI)

    • Kubernetes Dashboard (UI)

    • Direct API calls

  • kubectl is the most powerful and comprehensive interface among the three

• Execution Flow:

  • kubectl sends requests to the API server

  • API server coordinates with worker processes to:

    • Execute commands

    • Manage containers, services, deployments, etc.

• Compatibility:

  • Not limited to Minikube

    • Works with cloud-based and hybrid clusters as well

---

Installation Requirements and Process

General Requirements

• Minikube Requires a Hypervisor

  • To run VM locally

  • Options include:

    • VirtualBox

    • HyperKit (used in the example)

• Operating System Support:

  • Works across platforms

  • Installation commands vary per OS (Mac, Windows, Linux)

Installation on macOS (Example)

• Step 1: Install Hypervisor

  • Example used: HyperKit

    • Required to create VM for Minikube

• Step 2: Install Minikube

  • Also installs kubectl as a dependency

    • No need to install kubectl separately

  • Installer pulls down:

    • Kubernetes CLI (i.e., kubectl)

    • Any additional dependencies

• Validation of Installation

  • Run kubectl to confirm CLI is operational

    • Displays command list

  • Run minikube to confirm Minikube CLI is accessible

---

Starting and Managing a Minikube Cluster

Starting the Cluster

• Command Example:

  minikube start --vm-driver=hyperkit
  

  • -vm-driver: Specifies the hypervisor to use (e.g., HyperKit)

• Setup Highlights:

  • Downloads necessary resources if not cached

  • Starts a VM with both master and worker processes

  • Docker runtime is pre-installed inside the node VM

    • Works even if Docker is not installed on the host machine

Post-Setup Checks

• Verify Cluster Status

  • Run:

    • kubectl get nodes → Shows node readiness

    • minikube status → Confirms services (e.g., kubelet, container runtime) are running

  • Example output:

    • Node is ready

    • Role: master

• Version Check

  • kubectl version

    • Displays client and server versions

    • Example: Kubernetes v1.17

---

Debugging and Troubleshooting Minikube

• Deleting and Restarting the Cluster

  • To reset setup:

    minikube delete

• Starting in Debug Mode

  • For detailed logs and error tracking:

    minikube start --vm-driver=hyperkit --alsologtostderr -v=7

  • -alsologtostderr and v=7: Enable verbose logging to console

  • Helpful when:

    • Minikube fails to start

    • Dependencies or VM drivers fail