
I tend to create environments for my labs using containers. These containers would hold the resources I need for a certain job. Here's a guide on how I usually create them from scratch.

## Project Structure

Let’s assume we have a folder named `lab-container` with the following structure:

lab-container/ ├── Dockerfile ├── build │ └── binaries.sh ├── .dockerignore ├── Makefile ├── .github │ └── workflows │ └── ci.yml ├── README.md └── zshrc

We'll walk through each file step by step.

---

## The Dockerfile

Here’s a **multi-stage Dockerfile** that compiles and strips binaries in a builder stage, then copies them into a smaller final image:

```bash
FROM alpine:3.19 AS builder

WORKDIR /build

ARG TERRAFORM_VERSION="1.3.9"
ARG VAULT_VERSION="1.12.4"
ARG HELM_VERSION="3.11.2"
ARG KUBECTL_VERSION="v1.26.1"
ARG KIND_VERSION="v0.17.0"
ARG TRAEFIK_VERSION="2.9.6"
ARG TASK_VERSION="3.19.0"

RUN apk update && apk add --no-cache \
      bash \
      curl \
      wget \
      unzip \
      git \
      tar \
      ca-certificates \
      upx \
      binutils  # `strip` and `upx` to reduce binary sizes

COPY build/binaries.sh /build/binaries.sh
RUN chmod +x /build/binaries.sh && /build/binaries.sh

# Optionally strip and compress
RUN strip --strip-unneeded /usr/local/bin/* || true
RUN upx --best --lzma /usr/local/bin/* || true

################################################################################
# Final image
################################################################################
FROM alpine:3.19

LABEL maintainer="YourName" \
      description="A lightweight DevOps toolbox container."

WORKDIR /labs
ENV PATH="/labs/bin:$PATH"

RUN apk update && apk add --no-cache \
      bash \
      ca-certificates \
      python3 \
      py3-pip \
      aws-cli \
      zsh \
      jq \
      fzf \
      eza

COPY --from=builder /build/bin/ /usr/local/bin/

COPY zshrc /root/.zshrc

SHELL ["/bin/zsh", "-c"]

CMD ["zsh"]
```

### Explanation

Builder Stage:

- Installs dependencies (bash, curl, tar, etc.) needed to fetch and install binaries.

- Copies binaries.sh which downloads and installs Terraform, Vault, Helm, etc.

- Uses strip and upx to reduce final binary sizes.

### Final Stage:

- Starts with a fresh Alpine base.

- Installs additional packages (e.g., AWS CLI, zsh, jq, fzf).

- Copies the compiled binaries from the builder stage.

- Sets up a default shell (zsh) and copies a zshrc config file for a nice shell environment.

## build/binaries.sh

This script downloads and installs each DevOps tool. For example:

```bash
#!/usr/bin/env bash
set -euxo pipefail

# In this script, you can use $TERRAFORM_VERSION, $VAULT_VERSION, etc.
# to download each tool. For instance:
wget https://releases.hashicorp.com/terraform/${TERRAFORM_VERSION}/terraform_${TERRAFORM_VERSION}_linux_amd64.zip
unzip terraform_${TERRAFORM_VERSION}_linux_amd64.zip -d /usr/local/bin
rm terraform_${TERRAFORM_VERSION}_linux_amd64.zip

# Similarly for Vault:
wget https://releases.hashicorp.com/vault/${VAULT_VERSION}/vault_${VAULT_VERSION}_linux_amd64.zip
unzip vault_${VAULT_VERSION}_linux_amd64.zip -d /usr/local/bin
rm vault_${VAULT_VERSION}_linux_amd64.zip

# ... repeat for Helm, kubectl, kind, Traefik, Task, etc.
# e.g. for kubectl:
curl -L "https://dl.k8s.io/release/${KUBECTL_VERSION}/bin/linux/amd64/kubectl" -o /usr/local/bin/kubectl
chmod +x /usr/local/bin/kubectl

# For kind:
curl -Lo /usr/local/bin/kind "https://kind.sigs.k8s.io/dl/${KIND_VERSION}/kind-linux-amd64"
chmod +x /usr/local/bin/kind

# For Task CLI:
curl -L https://github.com/go-task/task/releases/download/v${TASK_VERSION}/task_linux_amd64.tar.gz | tar -zx -C /usr/local/bin
```

Tip: Keep the script idempotent. If you run it twice, it shouldn’t break anything.

## .dockerignore

The .dockerignore helps exclude files you don’t want in your build context (like .git, logs, local environment files, etc.):

```bash
.env
.DS_Store
*.log
.git
.gitignore
README.md
node_modules/
```

Adjust to your needs.

## Makefile

A simple Makefile can streamline common Docker tasks:

```makefile
IMAGE_NAME = your-dockerhub-username/devops-toolbox
VERSION = v1.0.0
LATEST = latest

.PHONY: build
build:
	docker build -t $(IMAGE_NAME):$(LATEST) .

.PHONY: run
run:
	docker run --rm -it $(IMAGE_NAME):$(LATEST) zsh

.PHONY: test
test:
	docker run --rm $(IMAGE_NAME):$(LATEST) sh -c ' \
		echo "Checking installed tools..."; \
		terraform --version && \
		vault --version && \
		aws --version && \
		kubectl version --client && \
		helm version && \
		kind version && \
		traefik version && \
		task --version && \
		docker version \
	'

.PHONY: push
push:
	docker tag $(IMAGE_NAME):$(LATEST) $(IMAGE_NAME):$(VERSION)
	docker push $(IMAGE_NAME):$(VERSION)
	docker push $(IMAGE_NAME):$(LATEST)

.PHONY: clean
clean:
	docker image prune -f

.PHONY: deploy-k8s
deploy-k8s:
	kubectl run toolbox --image=$(IMAGE_NAME):$(LATEST) -- sleep infinity

.PHONY: exec-k8s
exec-k8s:
	kubectl exec -it toolbox -- zsh

```

## Explanation

build: Builds the Docker image locally with the latest tag.

run: Launches the container in an interactive shell.

test: Runs each tool’s --version to confirm they’re installed.

push: Tags and pushes the image to Docker Hub.

deploy-k8s: Runs the container in a Kubernetes pod.

exec-k8s: Opens a shell into that running pod.

## zshrc

For a better shell experience, add your Zsh config:

```bash
export PATH="/usr/local/bin:$PATH"
export EDITOR=vim

# Enable autocompletion
autoload -Uz compinit
compinit

# Enable fzf for fuzzy search
[ -f ~/.fzf.zsh ] && source ~/.fzf.zsh

# Some aliases
alias ls="eza --icons=always"
alias la="eza -al --icons"

# Theme or prompt settings (e.g. powerlevel10k)
ZSH_THEME="powerlevel10k/powerlevel10k"
[[ ! -f ~/.p10k.zsh ]] || source ~/.p10k.zsh
```

More aliases for kubectl, terraform, etc. as desired

## GitHub Actions (Optional)
 
If you want automatic builds when you push changes, create .github/workflows/ci.yml:

```yaml
name: devops-toolbox

on:
  push:
    paths:
      - .github/workflows/ci.yml
      - Dockerfile
      - build/binaries.sh

jobs:
  build-and-publish:
    runs-on: ubuntu-latest
    permissions:
      id-token: write
      contents: write

    steps:
    - name: Check out
      uses: actions/checkout@v2

    - name: Log in to Docker Hub
      uses: docker/login-action@v1
      with:
        username: ${{ secrets.DOCKER_USERNAME }}
        password: ${{ secrets.DOCKERHUB_TOKEN }}

    - name: Build and Publish
      run: |
        registry="your-dockerhub-username"
        sha=$(git rev-parse --short HEAD)
        primary_tag="${registry}/devops-toolbox:${sha}"

        docker build -t $primary_tag -f Dockerfile .
        docker push $primary_tag

        current_branch=$(git rev-parse --abbrev-ref HEAD)
        if [ "$current_branch" = "main" ]; then
          retention_tag="${registry}/devops-toolbox:rc-${sha}"
          docker tag $primary_tag $retention_tag
          docker push $retention_tag
        else
          retention_tag="${registry}/devops-toolbox:alpha-${sha}"
          docker tag $primary_tag $retention_tag
          docker push $retention_tag
        fi
```

When you push changes to GitHub, this workflow will build your Docker image and push it to your Docker Hub account.

## Building and Testing Locally

Build:

```bash
make build
```

`docker build -t your-dockerhub-username/devops-toolbox:latest .`

Test:

```bash
make test
```

This checks each installed tool’s version.

Run:
```bash
make run
```

You should see a Zsh prompt with Terraform, Vault, AWS CLI, Helm, kubectl, etc. available.

## Pushing to Docker Hub
Assuming you’re logged in to Docker Hub (docker login), run:

`make push`

This tags the image with v1.0.0 and latest, then pushes both to your Docker Hub repository.

## Deploying in Kubernetes

If you want to spin up this container in a Kubernetes cluster:

`make deploy-k8s`

It creates a pod named toolbox that sleeps indefinitely. To open a shell inside that pod:

`make exec-k8s`

You’ll then see the same environment as the local container—handy for on-cluster troubleshooting or dev tasks.

## Conclusion

By following these steps, you’ve built a DevOps Toolbox container from scratch. You learned how to:

- Use a multi-stage Dockerfile to install and strip binaries.

- Write a script (binaries.sh) for easy updates to Terraform, Vault, Helm, etc.

- Set up a Zsh environment with helpful aliases.

- Automate builds with Makefile and optionally GitHub Actions.

- Push your image to Docker Hub and deploy it on Kubernetes.

This container simplifies DevOps tasks, letting you quickly spin up a consistent environment anywhere Docker or Kubernetes runs. 


An end-to-end guide for creating a lab container for DevOps training.

Posts tagged with: container