Workshop

Kubernetes promises high availability and resilience, but how does it recover from failures?

In this hands-on session, you will test Kubernetes’ fault tolerance by building a cluster from scratch and then methodically breaking it to observe its behaviour. You will gain deep insights into core components, including the control plane and the event-driven architecture that powers Kubernetes.

Target Audience

DevOps engineers, SREs, platform engineers, and Kubernetes administrators who want to move beyond theoretical knowledge and gain deeper insights into Kubernetes’ resilience features.

Workshop Overview

You’ll examine how kubelet maintains the desired state and verify the “no single point of failure” design principle. By taking down the cluster one node at a time, you will witness Kubernetes’ recovery mechanisms and develop a practical understanding of designing robust, production-ready deployments.

This session is ideal for DevOps practitioners, platform engineers, and anyone looking to strengthen their Kubernetes troubleshooting skills beyond theoretical knowledge.

Key Takeaways

• The process of bootstrapping a Kubernetes cluster without relying on managed services
• The inner workings of etcd, API server, scheduler, and kubelet and how they interact
• How Kubernetes detects and responds to node failures in real-time
• Why the system takes 5 minutes to reschedule pods after a node failure
• Critical insights about workload distribution when nodes rejoin the cluster

Benefits for Operations Teams, responsible for managing infrastructure, ensuring reliability, and optimizing performance (key concerns in DevOps practices), By the end of the session, you will:

• Have a deep understanding of Kubernetes’ fault-tolerance mechanisms
• Understand how to evaluate cluster resilience through controlled experiments
• Learn troubleshooting approaches for common failure scenarios
• Develop practical insights for designing robust production deployments

Prerequisites for Participants

• Basic understanding of containers and docker experience
• A laptop with your favourite IDE, Docker and the docker-compose CLI installed.
• A GitHub account

Useful Links

1. Troubleshooting Kubernetes deployments
2. Kubernetes instance calculator
3. Learn Kubernetes Weekly newsletter