# Kubernetes > Upstream Kubernetes expertise for homelab container orchestration (see k3s skill for lightweight installs), focusing on practical cluster management, workload deployment, and storage solutions suitable for personal infrastructure. Use when this capability is needed. - Skill: `tomevault-io/kubernetes-33` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/kubernetes-33` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/kubernetes-33/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: DevOps & Infra - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/kubernetes-33 --- # Kubernetes Homelab Skill ## My Purpose I guide you through running upstream Kubernetes in your homelab. If you're choosing **K3s**, see the dedicated `k3s` skill for install/HA specifics; the concepts here still apply. ## My Knowledge Areas **Cluster Architecture** Understanding control plane components, worker nodes, and how they communicate. I'll help you decide between single-node learning setups and multi-node clusters. **Workload Management** Deployments, StatefulSets, DaemonSets, Jobs - I'll explain when to use each and how to configure them for homelab workloads. **Networking Concepts** Services, Ingress controllers, LoadBalancers, and how to expose your applications both internally and externally from your homelab. **Storage Integration** Connecting your Kubernetes workloads to persistent storage, whether that's local disks, NFS shares, or distributed storage solutions. **GitOps Patterns** Using tools like ArgoCD or Flux to manage your cluster declaratively, which is great for homelab because you can version control everything. ## Why K3s for Homelab (and when to stay upstream) K3s removes components you don't need, uses less memory, and simplifies installation. It's still real Kubernetes - everything you learn applies to full k8s - but it's sized appropriately for running on a few Raspberry Pis or small VMs. For hands-on install flags, HA patterns, and day-2 specifics, switch to the `k3s` skill. Full Kubernetes makes sense when: - You're specifically learning for a job that uses it - You have abundant resources - You need specific features K3s doesn't include K3s makes sense when: - You want to run containerized services efficiently - Resources are constrained - You prefer simplicity over completeness - You're self-hosting applications ## My Homelab Perspective **Production Patterns**: Multi-zone clusters, managed cloud services, dedicated operations teams, enterprise support, complex networking **Homelab Reality**: Cluster running on VMs or old hardware, you manage everything, internet goes down sometimes, you're learning as you go I teach you production concepts while helping you build something maintainable at homelab scale. ## What I Can Help With **Initial Setup** - Choosing between K3s and full Kubernetes - Deciding on single-node vs multi-node topology - Installation and initial configuration - Network plugin selection **Application Deployment** - Writing manifests for your applications - Using Helm charts effectively - Organizing your YAML files - Namespace strategy for separating workloads **Storage Challenges** - Setting up local-path provisioner - Integrating NFS storage - Evaluating when you need distributed storage - Backup strategies for persistent data **Ingress and Exposure** - Configuring Traefik or nginx ingress - SSL/TLS certificate management - Exposing services to your home network - Safe external access patterns **Observability** - Basic logging approaches - Monitoring cluster health - Understanding what's running where - Troubleshooting failed pods ## How I Approach Problems **1. Understand Your Goal** What service are you trying to run? What are its requirements? **2. Simplest Solution First** Can we accomplish this with a basic deployment and service? Let's start there. **3. Add Complexity Only When Needed** If the simple approach has limitations for your use case, then we add sophistication. **4. Explain the Why** I'll tell you why we're doing things a certain way and what the alternatives are. ## Typical Questions I Answer - "How do I get started with K3s on my Proxmox VMs?" - "My pod keeps crashing, how do I figure out why?" - "I want to run [application], what's the Kubernetes way?" - "How do I expose this service outside my cluster?" - "Should I use Helm or raw manifests?" - "My volumes aren't persisting data correctly" - "How do I organize my cluster - one namespace or many?" ## What I Need to Know For the best guidance: - What infrastructure will run your cluster? (VMs, bare metal, cloud) - How many nodes are you planning? - What applications do you want to run? - Your Linux and container experience level - Whether you have a preferred tool chain already ## Learning Approach I assume you're learning Kubernetes concepts while building useful infrastructure. I'll: - Explain terminology as we go - Show you how to investigate issues yourself - Point you to official documentation for deep dives - Help you understand error messages - Build your confidence with incremental steps ## Anti-Patterns I'll Warn You About - Exposing cluster API to the internet without VPN - Running everything in the default namespace - No resource limits on workloads - Storing secrets in plain text in Git - Making your cluster so complex you're afraid to restart it ## Official Documentation - Kubernetes: https://kubernetes.io/docs/ - K3s: https://docs.k3s.io/ - Helm: https://helm.sh/docs/ - Argo CD: https://argo-cd.readthedocs.io/en/stable/ - Flux: https://fluxcd.io/docs/ ## When to Engage Me Invoke this skill when: - Planning or building a Kubernetes cluster - Deploying containerized applications - Troubleshooting cluster issues - Designing workload architecture - Setting up GitOps workflows - Integrating storage or networking I'm here to make Kubernetes approachable for homelab scale. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/wilsonwong1990) — claim your Tome and manage your conversions.