Architecture · Networking · Error Runbook · Storage · Observability
Battle-tested technical guides covering every layer of Kubernetes — from control plane internals to real-world production debugging. Written from the Platform Ops trenches at 6D Technologies.
Five deep-dives covering the full K8s stack — architecture, networking, error runbook, storage, and observability.
Complete structural guide — Storage Nodes, Worker Nodes, Pods, Control Plane, CSI, PVs, QoS classes, health probes, and scheduling controls with interactive diagrams.
Six architecture diagrams — Pod-to-Pod networking, CNI internals, Ingress, Gateway API, NetworkPolicy deep-dive, CoreDNS resolution, and eBPF observability.
32 error types across 5 layers — Pod, Worker Node, Cluster, API Server, and OpenShift. Root causes, diagnostic commands, exit code table, and production fixes.
PV/PVC lifecycle, StorageClasses & dynamic provisioning, CSI driver internals, StatefulSets, and a backup/DR strategy with VolumeSnapshots and Velero.
Metrics (Prometheus + Grafana), logs (Loki/ELK), traces (Jaeger/OpenTelemetry), SLO burn-rate alerting, and how a platform like OpsCore ties every signal together.
The same map as the Kubernetes category hub, and the same links — every green block is covered by a published issue, so click one to go straight to it.
Every article covers multiple layers — click the cards above to read the full deep-dive.
New to Kubernetes? Follow this sequence to build a solid mental model before diving into production ops.
Start with the big picture — how control plane, worker nodes, and storage interact. Understand the reconciliation loop before touching anything else.
Read Issue #048 → K8s ArchitectureArchitecture introduced Storage Nodes and CSI at a glance — this is where PV/PVC binding, StorageClasses, and StatefulSets get the full treatment.
Read Issue #050 → K8s StorageOnce you understand the architecture, layer in networking. CNI plugins, CoreDNS, Ingress, and NetworkPolicy are where most production headaches live.
Read Issue #047 → K8s NetworkingNow that you understand architecture, storage, and networking, learn how to diagnose every failure — 32 error types across 5 layers with production-tested fixes and diagnostic commands.
Read Issue #049 → K8s Error RunbookPrometheus + Grafana, Loki/ELK, and Jaeger/OpenTelemetry — plus how a platform like OpsCore ties every signal together into one pane of glass.
Read Issue #051 → K8s ObservabilityRBAC deep-dive, admission controllers, and multi-cluster federation are on the roadmap. Looking for GitOps? That's now its own deep-dive in the CI/CD series.
Read CI/CD & GitOps → Issue #052