Workloads & Deployment

Helm renders parameterized Kubernetes manifests and tracks releases. This guide installs the client, creates a small chart, renders it locally, installs atomically, upgrades values, inspects history, and rolls back.

Ingress resources do nothing until an Ingress controller is installed. This guide deploys ingress-nginx with its maintained Helm chart, publishes a test Service, and validates routing locally before DNS.

cert-manager automates certificate issuance and renewal through Kubernetes resources. This guide installs it with Helm, creates a staging ACME ClusterIssuer, requests an Ingress certificate, and shows the safe path to production.

ConfigMaps hold non-confidential settings; Secrets hold sensitive bytes but are not encrypted by default unless the cluster enables encryption at rest. This lab consumes both as environment values and read-only files.

Jobs run Pods to completion; CronJobs create Jobs on a schedule. This lab controls retries, deadlines, concurrency, history, manual triggering, log retrieval, and cleanup.

A Pod is Kubernetes’ smallest schedulable unit, but production applications normally use a controller. This lab builds a Pod manifest with labels, environment, resources, probes, and a security context so every field…

Deployments manage ReplicaSets and make stateless rollout history explicit. This lab controls surge and unavailable capacity, watches an update, detects a bad image, rolls back, and scales safely.

A Service gives changing Pods a stable virtual address. This lab creates one Deployment, proves selector-based EndpointSlices, and compares internal ClusterIP with NodePort and provider-backed LoadBalancer exposure.

This command-focused lab teaches a repeatable troubleshooting loop: locate a resource, inspect desired and current state, read events and logs, enter a container only when necessary, change declaratively, and delete by manifest.