Kubernetes

Kubernetes is the standard platform for running containerized workloads at scale. These posts cover how it works, how to operate it, and how to extend it, from the scheduler to the Operator pattern.


Architecture

  • Architecture: How Kubernetes Works
    The control plane, worker nodes, and the reconciliation loop that holds it all together. A complete walkthrough of every component and how a request moves through the system end to end.

Scheduling

  • Explained: Taints and Tolerations in Kubernetes
    How nodes repel pods by default, and how pods opt in with a matching toleration. The foundation for dedicated node pools and workload isolation.
  • Explained: Node Affinity in Kubernetes
    Expressing scheduling requirements and preferences using label selectors. Hard requirements that block placement, soft preferences that guide it, and how the two combine with taints for precise control.

Networking

  • Explained: CoreDNS
    How Kubernetes resolves service names inside the cluster. The plugin chain, the Corefile, the ndots:5 behavior that generates unexpected query volume, and how to debug DNS failures methodically.

Operators

  • Explained: Kubernetes Operators
    What Operators are and why they exist. Custom Resource Definitions, the reconciliation loop, and how domain-specific operational knowledge gets encoded as code that runs inside the cluster.
  • How to Write a Kubernetes Operator
    A complete end-to-end walkthrough using Kubebuilder and Go. Define a CRD, write a reconciler, set up RBAC, and deploy a working Operator, with the full code and the reasoning behind each decision.

Posts are added to this page as they are published.

Written by Phee Jay.



×