Introduction to Helm¶

Overview¶

Explain what Helm solves for Kubernetes teams and define chart, release, and repository in ops language.

Helm is the package manager for Kubernetes. A chart is a versioned bundle of templates and defaults; a release is an installed instance. This course is Helm for Kubernetes Engineers — production charts, not toy demos.

This is a core tutorial in Module 1 · Helm Fundamentals of the REBASH Academy Helm for Kubernetes Engineers series — written for Cloud, DevOps, Platform, and SRE engineers.

Prerequisites¶

Kubernetes — Deployments, Services, kubectl apply basics

Learning Objectives¶

By the end of this tutorial, you will be able to:

State problems Helm solves vs raw YAML
Define chart, release, repository
Contrast imperative apply vs packaged releases
Name when not to use Helm

Architecture¶

This topic’s control points and relationships are shown below.

Theory¶

What it is¶

Helm is the de facto package manager for Kubernetes. Instead of applying loose YAML files by hand, you install a chart — a versioned directory (or .tgz archive) that contains templates, default values, and metadata. An installed instance of a chart is a release: a named deployment in a namespace with history you can upgrade or roll back. Charts are published from repositories (classic HTTP indexes) or OCI registries (container-style references such as oci://…).

Three words you will use every day:

Term	Meaning
Chart	Packaged templates + `values.yaml` + `Chart.yaml`
Release	One installed instance of a chart (name + namespace + revisions)
Repository / OCI	Where teams discover and pull charts

Why it matters¶

Raw kubectl apply works for a single environment. It breaks down when you need the same app shape in dev, staging, and production with different replicas, images, and ingress hosts. Helm gives you parameterisation (values), a shared package teams can reuse, and release history for rollback. Platform teams publish “golden” charts so product squads do not reinvent Deployments and Services. In GitOps workflows, controllers such as Argo CD and Flux treat Helm charts as first-class deployable units.

How it works¶

You (or CI/GitOps) run the Helm CLI against a cluster kubeconfig. Helm fetches the chart, merges default values with your overrides, renders Go templates into plain Kubernetes manifests, and applies them through the API server. Helm 3 stores release metadata in the cluster (Secrets or ConfigMaps in the release namespace). Each upgrade creates a new revision; helm rollback re-applies an earlier revision’s rendered set.

Key concepts and comparisons¶

Pain without Helm	Helm answer
Copy-paste YAML per env	Values overrides (`-f`, `--set`)
No versioned app package	Chart `version` + informational `appVersion`
Manual rollback	`helm rollback` or GitOps revert of values/chart
Hard to share standards	Reusable application charts and library charts

Helm vs Kustomize vs plain YAML: use plain YAML or Kustomize when a small, mostly static set of manifests is enough. Prefer Helm when you need rich parameterisation, dependency composition, and release lifecycle across many environments.

Common pitfalls¶

Helm is not a second control plane — Kubernetes still owns Pods and Deployments after apply.
A chart is not a running app; the release is what you operate day to day.
Helm does not replace GitOps; production teams usually let a controller run Helm, not laptops.
Over-templating every field makes charts unreadable — keep defaults sensible and overrides intentional.

Hands-on Lab¶

Objective¶

Author a minimal Helm chart by hand, lint and render it offline, then install a release and capture evidence that distinguishes the chart (source package) from the release (named cluster instance).

Prerequisites¶

Helm 3.x (helm version)
kubectl configured against kind or minikube (optional for install)
Writable workspace at ~/rebash-helm/module-01

Lab environment¶

Workspace: ~/rebash-helm/module-01 on your workstation; cluster optional until Task 4.

Terminal

mkdir -p ~/rebash-helm/module-01/rebash-app/templates && cd ~/rebash-helm/module-01

Real-world scenario¶

Your platform team ships internal services as small Helm charts. Before merging, you must prove the chart lints cleanly, renders expected Kubernetes kinds, and — when a cluster is available — installs as a named release in an isolated namespace.

Step-by-step tasks¶

Task 1 – Chart metadata and defaults¶

Create rebash-app/Chart.yaml:

Chart.yaml

apiVersion: v2
name: rebash-app
description: Minimal REBASH introduction chart
type: application
version: 0.1.0
appVersion: "1.27"

Create rebash-app/values.yaml:

values.yaml

replicaCount: 1
image:
  repository: nginxinc/nginx-unprivileged
  tag: "1.27-alpine"
  pullPolicy: IfNotPresent
service:
  type: ClusterIP
  port: 8080

Task 2 – Deployment and Service templates¶

Create rebash-app/templates/_helpers.tpl:

{{- define "rebash-app.name" -}}
{{- default .Chart.Name .Values.nameOverride | trunc 63 | trimSuffix "-" }}
{{- end }}

{{- define "rebash-app.fullname" -}}
{{- printf "%s-%s" .Release.Name .Chart.Name | trunc 63 | trimSuffix "-" }}
{{- end }}

Create rebash-app/templates/deployment.yaml:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: {{ include "rebash-app.fullname" . }}
  labels:
    app.kubernetes.io/name: {{ include "rebash-app.name" . }}
    app.kubernetes.io/instance: {{ .Release.Name }}
spec:
  replicas: {{ .Values.replicaCount }}
  selector:
    matchLabels:
      app.kubernetes.io/name: {{ include "rebash-app.name" . }}
      app.kubernetes.io/instance: {{ .Release.Name }}
  template:
    metadata:
      labels:
        app.kubernetes.io/name: {{ include "rebash-app.name" . }}
        app.kubernetes.io/instance: {{ .Release.Name }}
    spec:
      containers:
        - name: web
          image: "{{ .Values.image.repository }}:{{ .Values.image.tag }}"
          imagePullPolicy: {{ .Values.image.pullPolicy }}
          ports:
            - containerPort: {{ .Values.service.port }}

Create rebash-app/templates/service.yaml:

apiVersion: v1
kind: Service
metadata:
  name: {{ include "rebash-app.fullname" . }}
spec:
  type: {{ .Values.service.type }}
  ports:
    - port: {{ .Values.service.port }}
      targetPort: {{ .Values.service.port }}
  selector:
    app.kubernetes.io/name: {{ include "rebash-app.name" . }}
    app.kubernetes.io/instance: {{ .Release.Name }}

Task 3 – Lint and render (offline)¶

Terminal

cd ~/rebash-helm/module-01
helm lint rebash-app | tee lint-m01.txt
helm template demo rebash-app --namespace rebash-helm-m01 | tee render-m01.yaml
grep -E '^kind:' render-m01.yaml | sort | uniq -c | tee kinds-m01.txt
grep -q 'kind: Deployment' render-m01.yaml
grep -q 'kind: Service' render-m01.yaml
grep -q 'nginxinc/nginx-unprivileged:1.27-alpine' render-m01.yaml

Expected output

lint-m01.txt shows 0 chart(s) failed; kinds-m01.txt lists Deployment and Service; rendered image uses the pinned tag.

Task 4 – Install release and prove chart vs release¶

Create namespace.yaml:

namespace.yaml

apiVersion: v1
kind: Namespace
metadata:
  name: rebash-helm-m01

Terminal

cd ~/rebash-helm/module-01
if command -v helm >/dev/null && kubectl cluster-info >/dev/null 2>&1; then
  kubectl apply -f namespace.yaml
  helm upgrade --install demo rebash-app -n rebash-helm-m01 --wait --timeout 120s | tee install-m01.txt
  helm list -n rebash-helm-m01 | tee list-m01.txt
  helm get metadata demo -n rebash-helm-m01 | tee metadata-m01.txt
  grep -E '^name:|^version:|^app_version:' metadata-m01.txt | tee chart-vs-release-m01.txt
  kubectl get deploy,svc -n rebash-helm-m01 | tee objects-m01.txt
else
  echo "Skipping install — helm or cluster unavailable; offline lint/template is sufficient" | tee install-m01.txt
fi

Expected output

list-m01.txt shows release name demo with chart rebash-app-0.1.0; chart-vs-release-m01.txt shows chart metadata (name, version, app_version) distinct from the release name demo.

Validation steps¶

Chart directory contains Chart.yaml, values.yaml, and templates
helm lint passes with no failures
helm template renders Deployment and Service with pinned image
Evidence files distinguish chart package version from release name (when install runs)
Optional install completes in namespace rebash-helm-m01

Common errors and fixes¶

Error	Cause	Fix
`helm lint` template undefined	Missing `_helpers.tpl`	Add `rebash-app.name` and `rebash-app.fullname` helpers
Rendered YAML invalid	Bad indentation in template	Re-run `helm template` and inspect `render-m01.yaml`
Install watch timeout	Image pull or readiness	`kubectl describe pod -n rebash-helm-m01`
Confusing chart and release	Same string used for both	Release name is `demo`; chart name is `rebash-app` — compare `helm list` vs `Chart.yaml`

Challenge exercise¶

Bump replicaCount to 2 in a separate values-scale.yaml file, re-run helm template demo rebash-app -f values-scale.yaml, and assert replicas: 2 appears in the rendered Deployment.

Learning outcomes¶

Built a minimal chart without helm create scaffolding noise
Ran offline validation with helm lint and helm template
Explained the difference between chart (package) and release (installed instance)
Installed a pinned-image release into an isolated namespace when a cluster is available

Cleanup¶

Terminal

helm uninstall demo -n rebash-helm-m01 2>/dev/null || true
kubectl delete namespace rebash-helm-m01 --ignore-not-found

Validation¶

Lab commands run under ~/rebash-helm/module-01/
You can explain each Theory section in your own words
You used modern tooling where it applies to this topic
You can describe one production failure mode for this topic

Code Walkthrough¶

Production practice for Introduction to Helm always combines:

Inspect before you change (status, plan, logs, dry-run)
Prefer reversible, documented changes (Git, IaC, drop-ins, version pins)
Capture evidence (command output, pipeline logs) for handovers
Prefer current tools and APIs over legacy shortcuts
Least privilege — escalate credentials only when required

Keep runbooks short enough to follow under pressure. Automate checks; keep humans for judgement.

Security Considerations¶

Treat credentials and tokens for helm as privileged — never commit them
Prefer short-lived auth (OIDC, roles, SSO) over long-lived keys
Validate blast radius before apply/deploy/delete operations
Restrict who can approve production changes
Collect audit logs; limit who can read sensitive traces

Common Mistakes¶

Helm is not a second control plane — Kubernetes still owns Pods and Deployments after appl

Validate assumptions against the Theory section and official docs before changing production.

A chart is not a running app; the release is what you operate day to day.

Lab shortcuts (open security groups, admin roles, skip approvals) must not ship unchanged.

Changing production without a rollback path

Always know how to revert (previous artefact, prior release, state rollback, DNS failback).

Best Practices¶

Encode Introduction to Helm changes as code and review them in pull requests
Pin versions (images, modules, actions, provider plugins)
Separate environments with clear promotion gates
Alert on symptoms with runbooks attached
Destroy lab resources; tag everything with owner and expiry where possible

Troubleshooting¶

Symptom	Likely cause	Fix
Auth / permission denied	Wrong identity, policy, or scope	Check caller identity, roles, and least-privilege policies
Timeout / no route	Network, DNS, security group, or endpoint	Trace path, DNS, and allow-lists before retrying
Drift / unexpected plan	Manual change or wrong state/workspace	Reconcile desired vs actual; avoid click-ops on managed resources
Pipeline/job red	Flaky step, cache, or missing secret	Read failing step logs; bisect recent workflow/config changes
Cost spike	Idle load balancer, NAT, oversized compute	Inventory billable resources; stop/delete labs promptly

Summary¶

Introduction to Helm is essential for Cloud and DevOps engineers working with helm. Practise the lab until the inspection and change path is muscle memory, then continue the track.

Interview Questions¶

What problem does Helm solve for Kubernetes packaging?
What is a release in Helm 3?
How do charts differ from raw manifests?
What risks come from installing untrusted charts?
What is the difference between helm template and helm install?

Sample answer — question 2

A release is a named instance of a chart running in a cluster (with revision history). Helm tracks it via release metadata stored as Secrets (or ConfigMaps) in the namespace.

Sample answer — question 4

Untrusted charts can create privileged workloads, ClusterRoles, or exfiltrate secrets. Always render and review, pin versions, and install into least-privilege namespaces.

References¶

Helm docs — introduction

Introduction to Helm¶

Overview¶

Prerequisites¶

Learning Objectives¶

Architecture¶

Theory¶

What it is¶

Why it matters¶

How it works¶

Key concepts and comparisons¶

Common pitfalls¶

Hands-on Lab¶

Objective¶

Prerequisites¶

Lab environment¶

Real-world scenario¶

Step-by-step tasks¶

Task 1 – Chart metadata and defaults¶

Task 2 – Deployment and Service templates¶

Task 3 – Lint and render (offline)¶

Task 4 – Install release and prove chart vs release¶

Validation steps¶

Common errors and fixes¶

Challenge exercise¶

Learning outcomes¶

Cleanup¶

Validation¶

Code Walkthrough¶

Security Considerations¶

Common Mistakes¶

Best Practices¶

Troubleshooting¶

Summary¶

Interview Questions¶

Related Tutorials¶

References¶