Kubernetes Without the YAML Stockholm Syndrome

When You Don't Need K8s

A hard look at whether Kubernetes is the right tool for your team before you spend six months configuring it.

The Cluster You Actually Want

How to design a managed Kubernetes control plane, node pools, and addon stack without over-engineering it from day one.

Workload Identity and IRSA

Why static AWS credentials inside Kubernetes pods are a liability and how IRSA and Workload Identity eliminate them for good.

Networking Primitives That Bite

A practitioner's guide to Services, Ingress, NetworkPolicy, and CNI selection — the four networking layers where Kubernetes clusters routinely fail.

Stateful Workloads

The real cost of running databases and stateful services on Kubernetes — StatefulSets, operators, and when managed cloud services are the right answer.

Autoscaling That Converges

HPA, VPA, and Cluster Autoscaler explained with the real failure modes that keep platform teams awake — and how to avoid them.

Cost Visibility

How to achieve meaningful Kubernetes cost allocation using OpenCost, tag discipline, and showback reports that engineering teams actually act on.

Multi-Tenancy

The real isolation boundaries in Kubernetes — namespaces, dedicated clusters, and vcluster — and what each model actually provides versus what it promises.

Disaster Recovery

What Kubernetes disaster recovery actually requires — Velero, etcd backups, cluster rebuild drills, and the uncomfortable truth about RTO.

Day-2 Operations

The unglamorous reality of long-term Kubernetes operations — upgrades, API deprecations, configuration drift, and the team model that actually sustains it.