#kubernetes

Keep configuration out of the image: ConfigMaps for plain settings, Secrets for credentials, injected as environment variables or mounted files.

A top-down routine for any "my pod is not working" moment: status, events, logs, inside, network, resources. Stop at the first thing that looks wrong.

Run a stateless app with 3 replicas, health probes and resource limits, expose it inside the cluster with a Service, and test it with port-forward.

Install the kubectl CLI on Linux or macOS, enable shell completion and a k alias, and verify the client version.

The mental model behind every kubectl command: control plane vs nodes, pods, deployments, services, config, storage and namespaces, and how labels tie them together.

The order to work through the Kubernetes notes: concepts, local setup, daily kubectl, first deployment, config, networking, storage, operations and debugging.

A pre-merge checklist for Deployments and Services: health, resources, security, config and release safety.

Spin up a disposable Kubernetes cluster on your laptop inside Docker with kind (fast, multi-node) or minikube (dashboard and addons).

Bootstrap a real multi-node cluster on Ubuntu VMs with kubeadm and containerd: prepare nodes, init the control plane, install a CNI, join workers.

Namespaces split one cluster into logical areas; contexts in ~/.kube/config switch between clusters, users and default namespaces.

Pods are ephemeral. A PersistentVolumeClaim requests durable storage and the StorageClass provisions the actual volume.

The container starts, exits, and Kubernetes backs off before restarting it. The pod is a symptom the cause is almost always in the previous container log or the probe config.