Personal Kubernetes studies ⚓¶

This repository contains my personal notes and studies on Kubernetes, including various configurations, deployments, and other related topics.

These notes are for personal reference and follow my learning process, but they are also meant to share knowledge with others.

Opinions are my own.

The opinions expressed in these notes are my own.

A foreword on vendor lock-in¶

Kubernetes can be a valuable tool in avoiding or mitigating vendor lock-in, particularly in cloud environments. Its open-source nature and container orchestration capabilities allow for greater flexibility and portability of applications across different cloud providers or even on-premises infrastructure.

How Kubernetes helps avoid vendor lock-in:

Open Source Foundation: Being open-source, Kubernetes is not tied to a single vendor's interests, ensuring a degree of independence and community-driven development.
Abstraction: Kubernetes abstracts away the underlying infrastructure (cloud provider, hardware) allowing you to deploy and manage applications consistently across different environments.
Containerization: Containers package applications and their dependencies, making them portable and able to run consistently on different platforms, further reducing vendor dependency.
Multi-cloud and Hybrid Cloud: Kubernetes facilitates multi-cloud and hybrid cloud strategies, enabling you to distribute workloads across multiple cloud providers or combine on-premises and cloud resources, reducing reliance on a single vendor.
Linux-Centric Platform: Kubernetes is primarily designed to run on Linux, which is itself open-source and widely supported. This focus on Linux ensures broad compatibility, flexibility, and avoids dependencies on proprietary operating systems.

However, to really avoid vendor lock-in, it's essential to learn the standard technologies and tools that are not specific to any cloud provider. In this context, this means focusing on the core Kubernetes concepts and tools like kubectl, kubeadm, Kind or MiniKube, and avoiding over-reliance on cloud-specific services that introduce dependencies on a particular vendor's ecosystem.

For this reason, my notes are focused on the standard Kubernetes technologies and tools, and how to run Kubernetes locally without relying on cloud providers.

This knowledge is anyway a useful foundation for working with managed Kubernetes offerings like Azure Kubernetes Service (AKS), Amazon Elastic Kubernetes Service (EKS), or Google Kubernetes Engine (GKE).

Discouraging as a Service¶

Cloud vendors and big consulting companies regularly brain-wash people into thinking that they need to use their specific services and that they will never manage to achieve anything by themselves.

For instance, when advertising tools like GitHub Codespaces or Azure DevBox, they say that "creating a development environment is too hard", but in reality creating development environments today has never been easier, for most use cases.

Similarly, Cloud providers tend to advertise their managed Kubernetes services, discouraging people from attempting to manage their own Kubernetes clusters on-premises. A defeatist climate is deliberately created to push people to use managed services, which are more expensive and lead to lock-in with the cloud service provider. This annoys me. One of the objectives of my studies is to assess by myself how hard is it to work with Kubernetes locally, and how to create clusters using kubeadm without relying on cloud providers.

David Heinemeier Hansson back in 2023 wrote about how they left the cloud and wrote: "all that work to containerize our applications and prepare them for the cloud actually turned out to make it relatively easy to exit.". There is hope!

Diving deeper section¶

The Diving deeper section contains other topics related to Docker and Kubernetes, including non-technical information and my personal view on events.

I recommend reading my summary on the Docker history before diving into containerization.

Requirements to follow¶

Basic knowledge of the following technologies is required to follow my notes:

Understanding of the web and networking is also helpful, as many of the examples involve deploying web applications in Kubernetes.

The getting started guide includes instructions to prepare a development environment and recommended learning resources for beginners to learn the basics of Docker and Kubernetes.